Naukratis: Greek Diversity in Egypt: Studies On East Greek Pottery and Exchange in The Eastern Mediterranean [PDF]

Zitiervorschau

Naukratis: Greek Diversity in Egypt Studies on East Greek Pottery and Exchange in the Eastern Mediterranean

Edited by Alexandra Villing and Udo Schlotzhauer

The British Museum Research Publication Number 162 Publishers The British Museum Great Russell Street London WC1B 3DG Series Editor Dr Josephine Turquet Distributors The British Museum Press 46 Bloomsbury Street London WC1B 3QQ

Naukratis: Greek Diversity in Egypt Studies on East Greek Pottery and Exchange in the Eastern Mediterranean Edited by Alexandra Villing and Udo Schlotzhauer Front cover: Fragment of North Ionian black-figure amphora (?) from Naukratis. British Museum GR 1886.4-1.1282 (Vase B 102.33) ISBN-13 978-086159-162-6 ISBN-10 086159-162-3 ISSN 0142 4815 © The Trustees of the British Museum 2006

Note: the British Museum Occasional Papers series is now entitled British Museum Research Publications.The OP series runs from 1 to 150, and the RP series, keeping the same ISSN and ISBN preliminary numbers, begins at number 151. For a complete catalogue of the full range of OPs and RPs see the series website: www/the britishmuseum.ac.uk/researchpublications or write to: Oxbow Books, Park End Place Oxford OX1 1HN, UK Tel: (+44) (0) 1865 241249 e mail [email protected] website www.oxbowbooks.com or The David Brown Book Co PO Box 511, Oakville CT 06779, USA Tel: (+1) 860 945 9329;Toll free 1 800 791 9354 e mail [email protected]

Printed and bound in UK by Latimer Trend & Co. Ltd.

Contents

Contributors Preface Naukratis and the Eastern Mediterranean: Past, Present and Future Alexandra Villing and Udo Schlotzhauer

v vii 1

I NAUKRATIS: THE SITE, ITS CULTS AND ITS POTTERY The Hellenion at Naukratis: Questions and Observations Ursula Höckmann and Astrid Möller

11

The Delta: From Gamma to Zeta Alan Johnston

23

‘Drab Bowls’ for Apollo: The Mortaria of Naukratis and Exchange in the Archaic Eastern Mediterranean Alexandra Villing

31

Carian Mercenaries at Naukratis? Dyfri Williams and Alexandra Villing

47

II EAST GREEK POTTERY AND ITS PRODUCTION CENTRES: ARCHAEOLOGY AND SCIENCE The Study of East Greek Pottery John Boardman

49

East Greek Pottery from Naukratis: The Current State of Research Udo Schlotzhauer and Alexandra Villing

53

Neutron Activation Analysis of Pottery from Naukratis and other Related Vessels Hans Mommsen with M.R. Cowell, Ph. Fletcher, D. Hook, U. Schlotzhauer, A. Villing, S. Weber and D. Williams

69

Naukratis: Les importations grecques orientales archaiques. Classification et détermination d’origine en laboratoire Pierre Dupont and Annie Thomas

77

Archaic Greek Plates from the Apollo Sanctuary at Emecik, Knidia. Results and Questions Concerning Dorian Pottery Production Regina Attula

85

The Non-Figured Wares from the Anglo-Turkish Excavations at Old Smyrna. Points of Contact with Naukratis Stavros Paspalas

93

Chemical Provenance Determination of Pottery: The Example of the Aiolian Pottery Group G Hans Mommsen and Michael Kerschner

105

On the Provenance of Aiolian Pottery Michael Kerschner

109

The Chian Pottery from Naukratis Dyfri Williams

127

Some Observations on Milesian Pottery Udo Schlotzhauer with contributions by P. Herrmann (†) and S. Weber

133

East Greek ‘Situlae‘ from Egypt Sabine Weber with an Appendix: Neutron Activation Analysis Results by H. Mommsen, A. Schwedt, S. Weber and M.R. Cowell

145

The Apries Amphora – Another Cartouche Donald Bailey

155

III EAST GREEK POTTERY AND THE EASTERN MEDITERRANEAN: CONTACT, EXCHANGE AND IDENTITY The Greeks in Berezan and Naukratis: A Similar Story? Richard Posamentir

159

Some Ceramic Inscriptions Istrian Sanctuaries: The Naukratis Approach Iulian Bîrzescu

169

Naukratis and Archaic Pottery Finds from Cyrene’s Extramural Sanctuary of Demeter Gerald Schaus

175

Imported Greek Pottery in Archaic Cyrene: The Excavations in the Casa del Propileo Ivan D’Angelo

181

Etruscan and Italic Finds in North Africa, 7th–2nd century BC Alessandro Naso

187

Identity in the Making: Greeks in the Eastern Mediterranean during the Iron Age Alexander Fantalkin

199

Bibliography

209

Contributors

Regina Attula Ernst-Moritz-Arndt-Universität Greifswald Institut für Altertumswissenschaften Rudolf-Petershagen-Allee 1 17487 Greifswald Germany [email protected]

Alexander Fantalkin Tel Aviv University Department of Archaeology and Ancient Near Eastern Civilizations Ramat Aviv, Tel Aviv 69978 Israel [email protected]

Donald Bailey The British Museum Greek and Roman Department Great Russell Street London WC1B 3DG United Kingdom [email protected]

Ursula Höckmann Taunusstr. 39 55118 Mainz Germany [email protected]

Iulian Bîrzescu Institute for Archaeology ‘Vasile Pârvan’ of the Romanian Academy Str. Henri Coanda, nr. 11, sector 1 010667 Bucharest Romania [email protected] John Boardman Ashmolean Museum Beaumont Street Oxford OX1 2PH United Kingdom [email protected] Ivan D'Angelo Università di Napoli ‘L'Orientale’ Dipartimento Mondo Classico e Mediterraneo Antico Palazzo Corigliano Piazza S. Domenico Maggiore 80138 Naples Italy [email protected] Pierre Dupont and Annie Thomas CNRS-UMR 5138, Archéométrie – Archéologie Université Lyon 2 7, Rue Raulin 69365 Lyon CEDEX 7 France [email protected]

Alan Johnston Institute of Archaeology University College London 31–34 Gordon Square London WC1H 0PY United Kingdom [email protected] Michael Kerschner Österreichisches Archäologisches Institut, ÖAI Franz-Klein-Gasse 1 1190 Vienna Austria [email protected] Astrid Möller Albert-Ludwigs-Universität Seminar für Alte Geschichte Kollegiengebäude 1 Werthmannplatz 79098 Freiburg i. Br. Germany [email protected] Hans Mommsen Rheinische Friedrich-Wilhelms-Universität Bonn Helmholtz-Institut für Strahlen- und Kernphysik Nussallee 14–16 53115 Bonn Germany [email protected]

Naukratis: Greek Diversity in Egypt | v

Contributors

Alessandro Naso Università degli Studi del Molise Dipartimento di Scienze Storiche, Umane e Sociali Via G. de Sanctis, snc 86100 Campobasso Italy [email protected]

Udo Schlotzhauer Deutsches Archäologisches Institut, DAI Eurasien-Abteilung Im Dol 2-6, Haus II 14195 Berlin Germany [email protected]

Stavros Paspalas Australian Archaeological Institute at Athens Zacharitsa 23 Koukaki 11741 Athens Greece [email protected]

Alexandra Villing The British Museum Greek and Roman Department Great Russell Street London WC1B 3DG United Kingdom [email protected]

Richard Posamentir Deutsches Archäologisches Institut, DAI Abteilung Istanbul Gümüssuyu/Ayapasa Camii Sok. 48 34437 Istanbul Turkey [email protected]

Sabine Weber Walkmühlstr. 6 65195 Wiesbaden Germany [email protected]

Gerry Schaus Wilfrid Laurier University Department of Archaeology and Classical Studies 75 University Avenue West Waterloo, Ontario N2L 3C5 Canada [email protected]

vi | Naukratis: Greek Diversity in Egypt

Dyfri Williams The British Museum Greek and Roman Department Great Russell Street London WC1B 3DG United Kingdom [email protected]

Preface

This volume has its origin in a workshop on Naukratis and East Greek pottery held at the British Museum in December 2004 as the 28th British Museum Classical Colloquium, the result of a collaboration between the British Museum and members of the Naukratis Project of SFB 295 at the Gutenberg-Universität Mainz. Made possible by the generosity of the Gerda-HenkelStiftung and the Caryatid Group of the British Museum’s Greek and Roman Department, to whom we extend our gratitude, the workshop brought together archaeologists, historians and scientists with the aim of generating a fruitful discussion and exchange of ideas and knowledge to further our understanding of the site of Naukratis in its wider, Eastern Mediterranean context. As it emerged, the scientific analysis of pottery samples taken both at the British Museum and elsewhere proved particular vital for many results presented here. To a large extent this was made possible by subsidies from the Deutsche Forschungsgemeinschaft, by the personal interest of Professor Hans Mommsen of the Helmholtz-Institut, Friedrich-WilhelmUniversität Bonn and the various other individuals, excavations and institutions that allowed material in their care to be analysed, and by the generous help of the staff of the Department of Conservation, Documentation and Science of the British Museum, notably Mike Cowell and Duncan Hook. As editors, we have greatly enjoyed working with such knowledgeable, reliable and responsive colleagues as have come together for the present volume. The collaborative spirit that pervades the volume has its roots in the stimulating discussion and collaborative ambience of the workshop, which led to further exchanges well beyond the confines of the actual gathering. We are grateful to all participants, who made it such an exceptionally productive experience. The contributions assembled in this volume reflect this ongoing research and discourse, which has helped the volume to be, we hope, not just a gathering of individual papers but more a thematically linked whole. Many people have contributed to making the workshop, the related research and this volume possible. On the Mainz side, we would like to thank in particular Sabine Weber (Mainz) for her vital input in the workshop and related research, and Ursula Höckmann and Detlev Kreikenbom (Mainz), Naukratis project

leaders within SFB 295 – Kreikenbom for his support in organising the financing of the workshop, and Höckmann for much help and constant openness to discussions. On the British Museum side, we would like to thank in particular Dyfri Williams, Keeper of the Greek and Roman Department, for making the workshop possible and for his unfailing support throughout; all colleagues in Greek and Roman Department and the Educational AV unit for help with organising the workshop; colleagues in the Department of Ancient Egypt and Sudan, especially Jeffrey Spencer and Neal Spencer, as well as in the Middle East Department, for helpful discussions and access to objects; Lesley Fitton, Susan Woodford, Mira Hudson, Bárbara Freitas, Sara Cambeta and Sotiria Papastavrou for help with proof-reading; Kate Morton for producing two wonderful new maps and several profile drawings; the British Museum’s Photography and Imaging Department, especially Dudley Hubbard, for producing new photographs of objects at short notice; Lindy Crewe for help with image editing; John Boardman for encouragement and the donation of his invaluable Naukratis archive to the British Museum; and last but not least Josephine Turquet for producing the volume sympathetically and efficiently as ever. Editorial note For Greek names a Greek spelling has been retained wherever it was deemed not too unusual for the eye, which invariably means there will be considerable inconsistencies (such as Klazomenai and Aiolis but Cyrene and Laconia). A joint bibliography can be found at the end of the volume. Journals have been abbreviated after the guidelines of the American Journal of Archaeology. Some additional abbreviations are used, such as NAA for neutron activation analysis. Stylistic phases in the development of East Greek pottery from various regions have been abbreviated (e.g. as NiA I = North Ionian Archaic I; MileA II = Milesian Archaic II) according to the new system set out in Kerschner and Schlotzhauer 2005. The order in which the contributions are arranged was in part determined by the practical necessities of printing the colour sections.

Naukratis: Greek Diversity in Egypt | vii

Naukratis and the Eastern Mediterranean: Past, Present and Future Alexandra Villing and Udo Schlotzhauer The Greek trading port of Naukratis in the Egyptian Nile Delta would have been a bustling harbour town in the Archaic period, the Shanghai of ancient Egypt, as Thomas Brown once put it.1 Greek ships docked here to sell Greek silver, wine and oil to Egyptians in exchange for linen, papyrus, grain, natron, and other goods. Greek traders deposited gifts in the local sanctuaries and stopped over with the local hetairai, whose famous beauty must have turned the head of many a sailor, not just, notoriously, that of the wine trader Charaxos from Lesbos,

brother of Sappho (Hdt. 2.135). The people of Archaic Naukratis, their cults and their trade, their relations with Egypt and their links with Greece, Cyprus and Phoenicia, and particularly their pottery – its use, its production centres in the East Greek world, and its distribution – are at the heart of the present volume, which arose out of a conference/workshop held at the British Museum late in 2004. This focus is reflected in the division of the volume into three main parts: the site itself and its cults; the pottery of Naukratis

Figure 1a The eastern Mediterranean and the Black Sea

Naukratis: Greek Diversity in Egypt | 1

Villing and Schlotzhauer

Figure 1b East Greece

Figure 2 Naukratis from the late 7th to 3rd centuries BC

and its production centres; and the position of Naukratis in the wider context of trade and exchange in the Archaic Mediterranean. Revisiting old material, publishing recent fieldwork in East Greece, North Africa and the Black Sea, and presenting the latest research and analyses, the contributions assembled here make clear what advances have been made in all those areas over the past few decades. This introductory essay aims to set the scene for the volume.2 It is not a summary of the chapters it contains (abstracts prefacing each contribution give easy access to the main topics and results of each article), but rather introduces, connects and considers some of the key questions relating to the site of Naukratis and its position in the Eastern Mediterranean web of contacts during the Archaic period; in doing so it draws on, expands and links up in different ways the evidence and insights provided by the various contributions. The more specific and specialist insights relating to the pottery from Naukratis and its production centres are summarised and contextualised in greater detail in an overview essay at the beginning of section II.

Both East Greek and Carian mercenaries played a significant role in the Egyptian army of the 26th Dynasty,5 having been first dispatched presumably following the alliance between Psammetichos I and the Lydian king Gyges in 662/1 bc, some even reaching advanced levels of command within their own ‘foreigners’’ branch of the army and navy. Integration into Egyptian society can be witnessed particularly in the region of Memphis, where intermarriage and adoption of Egyptian names and burial customs are recorded.6 In return, East Greek sanctuaries received ‘diplomatic’ gifts from the Egyptian pharaohs, and Egyptian goods and influence began to infiltrate Greece and the wider Mediterranean world. Naukratis at this time was one of the main intersection points between the Greek and Egyptian worlds(Fig. 1a). According to Herodotus (2.178-9), it had been established at the instigation of the Pharaoh Amasis by 12 Greek cities, mostly located in East Greece (Fig. 1b), to act as a gateway for trade between Greece and Egypt:

Naukratis, 120 years after Petrie Relations between Greece and Egypt go back a long time. In the Bronze Age contacts between the Minoan Cretans and Egypt are amply attested,3 and the Minoans and Myceneans who had settled on the Eastern shores of the Mediterranean, at sites such as Miletos, also attracted the Egyptians’ attention: Ionians are for the first time depicted among subject states at time of Amenophis III (1403–1364 bc).4 But after a long break it was only in the 7th century bc that significant contacts again developed. 2 | Naukratis: Greek Diversity in Egypt

Amasis favoured the Greeks and granted them a number of privileges, of which the chief was the gift of Naukratis as a commercial headquarters for any who wished to settle in the country. He also made grants of land upon which Greek traders, who did not want to live permanently in Egypt, might erect altars and temples. Of these latter the best known and most used – and also the largest – is the Hellenion; it was built by the joint efforts of the Ionians of Chios, Teos, Phokaia, and Klazomenai, of the Dorians of Rhodes, Knidos, Halikarnassus, and Phaselis, and of the Aiolians of Mytilene. It is to these states that the temple belongs, and it is they who have the right of appointing the officers in charge of the port. Other cities which claim a share in the Hellenion do so without justification; the Aiginetans, however, did build a temple of Zeus

Naukratis and the Eastern Mediterranean: Past, Present and Future

Figure 3 Petrie’s excavations at Naukratis; a) Sir Flinders Petrie, c. 1886; b) the mound of Naukratis during Petrie’s excavations; c) finds of Greek architectural fragments

separately, the Samians one in honour of Hera, and the Milesians another in honour of Apollo. In old days Naukratis was the only port in Egypt, and anyone who brought a ship into any of the other mouths of the Nile was bound to state on oath that he did so of necessity and then proceed to the Canopic mouth; should contrary winds prevent him from doing so, he had to carry his freight to Naukratis in barges all round the Delta, which shows the exclusive privilege the port enjoyed. (tr. A. de Sélincourt)

Over 120 years ago, in 1884, Sir William Flinders Petrie discovered the remains of ancient Naukratis (Fig. 2) in the Western Nile Delta on the Canopic branch of the Nile and identified it correctly as the site mentioned by Herodotus. Petrie’s first excavation campaign in 1884/5 (Figs 3a–d) at once uncovered rich remains relating to the Greek presence at the site; the sanctuaries of Apollo, Hera (originally identified as a palaistra), and of the Dioskouroi, along with the Scarab Factory and the Great Temenos (believed by Petrie to be the Hellenion) were excavated. Even if quite advanced for their time, excavations were by modern standards somewhat chaotic, conducted under difficult circumstances and in a constant race against the sebakhin, locals digging up soil for use as fertilisers on fields (cf. Fig. 4e).7 Work was continued in 1885/6 by Ernest A. Gardner on behalf of Petrie. Gardner further excavated the sanctuaries of Apollo, Hera, and the Dioskouroi, and discovered the sanctuary of Aphrodite. Some years later, David Hogarth of the British School at Athens, in 1899 and 1903, concentrated on the Hellenion and the Great Temenos (Figs 4a–d).8 More recently, an American expedition led by W.D.E. Coulson und A. Leonard in the 1970s and 80s set out to re-study Naukratis. Unfortunately, although their surveys and excavations shed much light on the post-Archaic site, its gradual destruction and the great lake that now covers the site of early Naukratis made it impossible to follow truly in the footsteps of Petrie.9

Research on Archaic Naukratis thus has to remain based on the finds of the old excavations. The four seasons of fieldwork produced much material evidence for the history of the site, while additional material was collected by travellers. Altogether these finds are vital evidence for the history of the Greek diaspora around the Mediterranean, for relations between Greeks and Egyptians, and for contacts and trade in the Eastern Mediteranean in general. Yet in spite of the considerable importance of the site and the continuing scholarly interest it has attracted, no comprehensive publication of the surviving material from Naukratis has ever been attempted, a fact that has severely hampered scholarly study of the site. Petrie’s and Gardner’s publications of finds were a model of speed, and the results of Hogarth’s excavations were also soon put into print.10 But neither were comprehensive, according to Petrie’s famous motto ‘half a loaf is better than no bread’.11 This situation would be less of a problem were it not for a further complicating factor: as the earlier excavation project was funded through subscriptions to the Egypt Exploration Fund, the material from the site was distributed among subscribers; material from Hogarth’s excavations, too, was spread among various collections, while further material was collected by private individuals. As a result, the finds are now shared between some 40 museums and collections all over the world – even though the largest part of this, some 50%, is held by the British Museum.12 An additional handicap is the skewed nature of the preserved sample of material and the uncertainty about what was discarded already on site, an issue addressed in more detail by Schlotzhauer and Villing, this volume. Notwithstanding these difficulties, the site, and in particular its pottery, has attracted much scholarly attention over the past 120 years: one only needs to mention E.R. Price’s study and Naukratis: Greek Diversity in Egypt | 3

Villing and Schlotzhauer

Figure 4 Hogarth’s excavations at Naukratis: a) Hogarth on the excavation; b) excavation and village; c) ‘Pavement of Artemis shrine?’; d) ‘Pedestal vase in situ, Edgar holding the pieces’; e) Sebakhin at Naukratis

Marjory Venit’s work on the Greek pottery from Naukratis, or Bernand’s catalogue of the pottery inscriptions in Le delta égypties d’après les textes grecs. From a wider perspective, several authors attempted an evaluation of the history and significance of Naukratis, such as von Bissing in 1951 and John Boardman in his seminal work on The Greeks Overseas, indispensable still for the study of early Greco-Egyptian relations.13 More recently, a crucial analysis of Naukratis and its role as a trading port was published in 2000 by Astrid Möller, spawning ongoing fruitful discussions, such as in a colloquium at Lyons dedicated to Naukratis.14 A year after Möller’s study, the proceedings of the 1999 Naukratis Colloquium at the Johannes GutenbergUniversität in Mainz, appeared, adding further aspects to our understanding of the history and material culture of the Greeks

4 | Naukratis: Greek Diversity in Egypt

and Carians in Egypt. The colloquium had been held under the auspices of the Mainz Naukratis Project, led by Ursula Höckmann, which between 1997 and 2003 made much progress in cataloguing and studying various types of Greek material at Naukratis with a focus on acculturation phenomena. Numerous subsequent articles and three forthcoming volumes present further results of the project.15 Still ongoing is the work on the database originally set up by the Mainz Naukratis Project and continued by the British Museum, which will eventually allow an overview of the material held by different museums and collections. The present volume, too, is a result of the British Museum’s collaboration with the Mainz group, and a starting point for future research into the extraordinary trading port that was Naukratis.

Naukratis and the Eastern Mediterranean: Past, Present and Future Life at Naukratis: Greeks and Egyptians It has long been recognised that the material evidence from Naukratis dates back to the latter part of the 7th century bc, the time of the reign of Psammetichos I. Greek objects (notably pottery) first appear in Egypt around the middle or even the last third of the of the 7th century bc,16 and the earliest finds at Naukratis of Greek pottery – Corinthian, Attic, East Greek and Carian (cf. Williams and Villing, this volume) – seem to confirm this. It remains difficult, however, to reconcile this material evidence with the account in Herodotus which appears to ascribe the foundation of Naukratis to Amasis. Even if there is no unanimous consensus, it is agreed by most scholars17 that Naukratis must have been founded during Psammetichos’ reign, presumably under the leadership of the Ionian city of Miletos (and perhaps as one of several trading posts in the Delta), while a re-organisation under Amasis concentrated Greek trade just on Naukratis and gave specific status to the other Greek cities involved in the venture, an interpretation which seems quite compatible with the wording of Herotodus’ passage quoted above.18 They were allowed to establish the Hellenion and were granted the administration of the site through the prostatai tou emporíou, thus, perhaps, marginalising Miletos – a city which had been an important supporter of Amasis’ adversary Apries. While the privileging of Naukratis at the expense of other trading posts can be seen as granting the site a special favour, it was also a way of keeping tight control of foreign traders entering the country, an aspect that may have gained importance particularly with the nationalist backlash that followed Apries’ reign.19 As has been realised to its full extent only very recently, Naukratis was in fact not necessarily the first and only point of contact for ships entering Egypt in the region of Sais. The harbour town of Hone (Thonis-Herakleion) guarded the mouth of the Canopic branch of the Nile as it entered the ‘sea of the Greeks’, i.e. the Mediterranean, and seems to have been the very first port of call where trade goods were taxed on behalf of the Egyptian state. This was certainly the situation in the Classical period, when the stelai20 erected by pharaoh Nektanebos I in Naukratis and Thonis-Herakleion specify that one tenth of taxes on imports passing through Thonis-Herakleion and on all transactions and local production of goods at Naukratis (Piemrôye) should be given to Neith of Sais. Yet it may have applied already to the late 7th century or 6th centuries bc, as finds from Thonis-Herakleion date back to at least the 6th century bc,21 thus raising new questions concerning the status of both sites and their relationship. What seems clear, however, is that Naukratis and ThonisHerakleion must have had a close relationship at least from an Egyptian point of view, and both must have been guarded closely by Egyptian officials, like any other point of intersection with foreign lands. During the Saite period, officials known as ‘Overseer of the gate of the Foreign Lands of the Great Green’ (i.e. the Mediterranean) would have been in charge of securing the borders as well as – as suggested by Posener – administrating trade taxes,22 and it seems likely that the administration of both sites reported to them, even if under the prostatai tou emporíou Naukratis can be assumed to have had autonomy at least in some regards. Just how strong an Egyptian presence would have been at Naukratis in the Archaic period remains uncertain.23 The

early excavators reported that in the southern part of the site only Egyptian objects were found, but it seems that these were mostly of Hellenistic date. Similarly, the Great Mound within the Great Temenos (identified originally by Petrie as a stronghold and storage building identical with the Hellenion) is now predominantly considered a ‘high temple’ in a temenos built in the 4th century bc under Nektanebos I for Amun of Batet,24 although the possibility that it was a much earlier fort for an Egyptian garrison established by Psammetichos I (and restored by Ptolemy) is still maintained by some.25 Others have suggested that Naukratis was originally an Egyptian settlement,26 whose name Pi-emroye (or Pr-mryt, ‘the Harbour/Port House’), used for Naukratis on the stelai erected by Nektanebos I in Naukratis and Heraklion/Thonis27 as well as in several other hieroglyphic and demotic inscriptions,28 was in fact the town’s original name. Aristagoras of Miletos29 even mentions an Egyptian settlement on the opposite side of the river to Naukratis at the time of its foundation, but no archaeological trace of this has been located to date. Nevertheless, we surely must assume at least some administrative and policing staff as well as interpreters (cf. Hdt. 2.154). Archaic Egyptian inscriptions (of unknown provenance), one referring to the renewal of a donation connected with the temple of Amon-Re Batet (assumed to be in Naukratis) and the other to a man from Naukratis,30 indeed seem to point to resident Egyptians at Naukratis,31 and there may have been Egyptians involved in the local scarab workshop, too.32 Fragments of 6th-century bc Egyptian pottery (Schlotzhauer and Villing Fig. 41),33 even if rare in the known extant record, may well have belonged to such Egyptian residents. Only further study of the Egyptian remains from the site may ultimately shed more light on the question of Egyptians at Naukratis and on the level of direct interaction between Greeks and Egyptians at Naukratis itself.34 The presence of Greeks in the Archaic period, by contrast, is amply attested, at least in terms of pottery, if not in architectural remains (cf. Fig. 3c),35 Of course, the evidence is largely confined to Greek sanctuaries, with the temenos of Apollo and the sanctuary of Aphrodite going back to the earliest period of the site, and the Hellenion to the time of Amasis,36 and it thus remains unclear what proportion of Greeks actually permanently lived at the site compared to the proportion of traders who only passed through the port and deposited their votives along the way. The excavated cemetery at the site seems to cover merely the Classical and Hellenistic periods (although a dinos stand [?] perhaps of 6th century bc date is also said to have been found there).37 Similarly, no firm evidence has been recorded for Archaic living quarters, apart from Petrie’s record of some some Archaic finds in the area of the houses, even though these must have existed, both on the evidence of Herodotus talking of Greeks settling down and living permanently in Naukratis, and the existence in Archaic times of presumably not just seasonal workshops. The latter produced scarabs and faience, perhaps also terracotta figurines, alabastra, floral garlands and some sculpture, as well as, as is more fully discussed by Schlotzhauer and Villing in this volume, pottery in an East Greek style, at least from the time of Amasis onwards. As has been remarked by many scholars before, the profile of the Greek pottery finds in Naukratis is well matched to the literary account of the founding cities of the emporion, with Naukratis: Greek Diversity in Egypt | 5

Villing and Schlotzhauer pottery from Ionia, Aiolis and the East Dorian region much in evidence, a picture reflected also in various contributions in the present volume (for a summary see Schlotzhauer and Villing, this volume). The presence of Greek pottery from elsewhere – Attica, Laconia, Corinth – matches the profile at Archaic East Greek sites and reflects the general pattern of pottery trade in the Archaic Mediterranean. The study of the pottery inscriptions, too, currently undertaken by Alan Johnston, essentially confirms this picture; once completed, it will provide a more complete understanding of the relative chronology and ritual life of the various sanctuaries. The study has already significantly expanded the range of dedicators and the number of pieces inscribed by each, in ceramic texts totalling well over 2500, yet it remains true that only a few visitors to the sanctuaries seem to have come from further afield: some possibly Lydian and Carian names can now be added to the already-known single Phoenician graffito (Schlotzhauer and Villing Fig. 24) and the two Cypriot graffiti of Classical date (Höckmann and Möller Fig. 6).38 In general, the cultic life of Archaic Naukratis presents itself as similar to that in the (East) Greek home cities, from where most cults were filiated and most pottery imported. East Greek decorated plates, for example, are much in evidence and presumably served as display pieces, similar to votive pinakes,39 even if – as is the case with East Dorian plates with marine motifs discussed here by R. Attula (Attula Figs 6-11) – not all types of such plates are present. A particularly close connection with the homeland can be witnessed in the import of specially made crockery from the Samian homeland for sacred meals: mugs and cups with dipinti to Hera (Villing and Schlotzhauer Figs 14-17) of exactly the same type as have been found in large numbers in the Samian Heraion appear in the sanctuary of Hera at Naukratis, a Samian foundation (Hdt 2.178),40 and clay analysis by Hans Mommsen shows that they were produced with the same clay as the numerous examples found on Samos. Other instances of commissioning of pottery from back home specifically for use and/or dedication in a specific sanctuary at Naukratis are also attested. A dipinto on a North Ionian LWG large cup, for example, designates it specifically for ‘Aphrodite at Naukratis’; it may have functioned as a mixing bowl in communal drinking rites.41 Chian chalices, too, carry bespoke votive dipinti: those by Aigyptis and Mikis (or –mikis) have been taken by Dyfri Williams to have been commissioned (presumably through intermediate traders or travelling acquaintances) by some of the famous hetairai resident at Naukratis,42 while the Chian/Aiginetan pair of traders Aristophantos and Damonidas43 presumably brought their chalices to Naukratis in person (Johnston Fig. 1). The actual presence of the dedicant at the sanctuary is also indicated by an interesting fragment (Johnston Fig. 9) that shows that transport across the seas might occasionally result in damage: a large Chian chalice with a painted pre-firing dedication by a …]mides has the mu incised at a point where the slip had peeled away, suggesting that the dedicator must have repaired the damage on the spot. Much of this inscribed as well as most of the uninscribed decorated pottery consists of drinking vessels and mixing bowls and – as in most Greek sanctuaries – must have been used in communal rites in the sanctuaries. Even undecorated coarse bowls, mortaria, presumably used for the preparation of sacred 6 | Naukratis: Greek Diversity in Egypt

meals in Apollo’s cult, frequently bear votive inscriptions (cf. Villing, this volume). It is not difficult to imagine that communal ritual meals must have been of particular importance in a place like Naukratis, where Greeks were gathered in a foreign environment and where cult was one way of re-enforcing a communal spirit and identity, and where the gods were, moreover, vital in ensuring the success of voyages and trade ventures. Ritual dining to further social and political cohesion is perhaps most prominently associated with feasting for Apollo Komaios, who was honoured by a symposium in the prytaneion.44 The prytaneion may have been located inside the Hellenion (Höckmann and Möller Fig. 2), the common sanctuary set up by the joint efforts of nine poleis, presumably following the reorganisation of Naukratis by the pharaoh Amasis around 570 bc. Here, as Höckmann and Möller conclude in the present volume, all three ethnic groups of Hellenes together worshipped the Greek gods and organized the administration for their emporion – a statement of their Hellenic, East Greek identity in the face of a foreign, Egyptian environment. Naukratis and trade in the Archaic Mediterranean Naukratis was, of course, not the only Greek emporion situated in a foreign environment, but one trading post among many in the Archaic Mediterranean, with manifold connections to other sites and with many of its features being paralleled, to some extent, at other sites. For Cyrene, further west along the Mediterranean coast of Africa, Gerry Schaus and Ivan D’Angelo in their contributions to the present volume note that, as at Naukratis, finds include little 7th and much 6th century bc pottery, notably of North Ionian and Chian, as well as South Ionian, provenance. Schaus suggests that Chian traders first came to Naukratis and then went on to Cyrene, and that Fikellura pottery, too, reached Cyrene on the back of trade to Naukratis, alongside, possibly, faience, scarabs and alabaster alabastra of Naukratite production. All this would presuppose the existence of a coastal trade route connecting Naukratis with Cyrene, in addition to the well-known sea route via Crete, a possibility also raised by Ivan D’Angelo in his study of pottery from domestic contexts in Cyrene, which complements the picture of the sanctuary pottery discussed by Schaus. There are, however, also distinctive differences between the pottery profiles of Naukratis and Cyrene. For example, no early Attic pottery has as yet been found in Cyrene – unlike at Naukratis, which yielded some of the earliest exported Attic material.45 Could this be explained by the involvement of Aigina in the foundation of Naukratis? Conversely, the Theran (D’Angelo Fig. 6), Cycladic and Cretan pottery at Cyrene demonstrates continuation of contact between colonists and their Aegean homeland. No pottery of these islands has been identified at Naukratis, yet the phenomenon of an on-going link with the mother cities is exactly the same, extending to otherwise little-exported pottery fabrics such as grey wares from the Aiolian and Trojan/Lesbian region (Kerschner Fig. 10; Schlotzhauer and Villing Figs 11–13). Also in the North, on the shores of the Black Sea, Milesian colonies (with a trading-post element) such as Istros (Histria) and Berezan – discussed in the present volume by Iulian Bîrzescu and Richard Posamentir – mirror the strong East Greek profile in pottery finds that is found at Naukratis. At Berezan, for example, from about 630 bc onwards, North Ionian, Chian,

Naukratis and the Eastern Mediterranean: Past, Present and Future South Ionian and Aiolian pottery suitable for drinking parties as well as decorated plates is much in evidence. In the 7th century bc Milesian, or South Ionian, pottery is predominant, while the picture changes dramatically in the first half of the 6th century bc in favour of North Ionian products (Posamentir Figs 3–4) – the same pattern as has recently been established by Michael Kerschner for Western Greek colonies.46 Also at Naukratis the large amount of 6th century bc North Ionian pottery is remarkable; unlike at Berezan (Posamentir Fig. 11), however, bird, rosette and other hemispherical bowls seem dwarfed in numbers by South Ionian cups with everted rim (‘Knickrandschalen’; Schlotzhauer and Villing Figs 21, 23, 27–29) – although we cannot be sure if this might not be due to their owners being keener to inscribe them, and thus making them more attractive for the excavators to keep. Finally, unlike at Naukratis, where a local pottery workshop has now been established with some certainty (Schlotzhauer and Villing, this volume), Posamentir’s research suggests that unusual pieces of pottery from Berezan more likely stem from workshops established not at Berezan itself but located in another Milesian colony in the Hellespont area. 47 At Naukratis, as in other Greek sites abroad, a characteristic mixture of pottery produced in the home cities and elsewhere can thus be observed. Who brought it here? The mariners who peddled those wares, or themselves dedicated them in the sanctuaries, clearly were not always of the same origin as their cargo. The wide distribution of Athenian and Corinthian pottery, for example, must be due in no small measure to the activities of Aiginetan traders, even though evidence such as the Corinthian dedication on a Corinthian louterion from Chios (Johnston Fig. 8) also points to the involvement of Corinthians themselves.48 The distribution of Laconian pottery is presumably due largely to traders from Aigina and Samos. Similarly, as is argued in this volume by Michael Kerschner, Aiolian pottery produced in Kyme (and Larisa?) may well have been traded by Phokaians. As regards Cypriot mortaria found at Naukratis, as Alexandra Villing points out in the present volume, they may have been traded not merely by Cypriots or Phoenicians but also by Greeks. Unfortunately, the scarce evidence for trade amphorae among the extant pottery from Naukratis prohibits a reliable profile of this type of trade to be established; among the inscribed pieces that were kept by the excavators are several Cypriot and Chian as well as some Samian, Klazomenian, other North-Greek, and Corinthian amphorae (e.g. Johnston Figs 14, 21); in addition, amphorae of Phoenician type were found (cf. Johnston, this volume, and Villing, this volume). The trading connections of Naukratis thus extended eastwards beyond the borders of Greece, towards Cyprus and Phoenicia, and westwards towards Cyrene. As Alessandro Naso demonstrates in his contribution to the volume, they even reached as far as Italy, from where several pieces of Etruscan bucchero pottery reached Naukratis. Again, this does not necessarily suggest the actual presence of Etruscans, but might be due to mediation by East Greeks or Aiginetans; a sizable number of bucchero sherds has, after all, been found in Archaic Miletos and other East Greek sites as well as on Aigina. Nevertheless, some degree of contact or trade is attested between Etruria and Southern Italy and Egypt from the mid-8th century bc onwards,49 though often probably through Greek and Phoenician/Cypro-Phoenician or Carthaginian merchants.50

Phoenician and Cypro-Phoenician traders were important players in the Archaic Mediterranean in general. As Alexander Fantalkin argues in the present volume, alongside Cypriots and Euboeans they were instrumental in the renewal of contacts between Greece and the East in the 10th to 8th centuries bc, encouraged in their ventures by the Assyrian empire, while from the 7th century onwards East Greek trade and expansion gained in importance, supported by Lydian imperial policy. Archaic East Greece was naturally more a part of the East than the West, but was also a mediator between the two, while mainland Greece remained on the margins (a situation, as Fantalkin points out, that paradoxically turned out to be instrumental in its unique development towards the ‘Greek miracle’ in the Classical period). That Phoenician traders played a role in Egypt, too, alongside the Greeks, is suggested by Diodorus (1.68.8), who points out that Greeks and Phoenicians were the main traders admitted into Egypt since the time of Psammetichos I. Phoenicians are attested notably in the Eastern part of the Delta and in the region around Memphis.51 Did they also come to Naukratis, as some have suggested?52 A single Phoenician inscription on a cup of East Dorian (Knidian?) production (Schlotzhauer and Villing Fig. 24),53 Phoenician-type amphorae,54 a Phoenician dipinto on a trade amphora,55 and Classical or later amphorae of Phoenician type with Greek dipinti56 hardly provide sufficient evidence to assume that Phoenician traders regularly frequented the port of Naukratis in the Archaic period, even if, of course, we need to remember that we do not possess the complete archaeological picture of the site. The situation is thus somewhat similar to the question of the presence of Cypriots in Naukratis, where, as is suggested by Villing in this volume, Archaic Cypriot sculpture, terracotta figurines, some pottery and few (Classical) inscriptions hardly suffice to postulate a thriving Cypriot community, even if, as Schlotzhauer57 once pointed out, occasional visits or even a handful of residents are not inconceivable. Archaic Naukratis, in its function as a primarily East Greek trading post in Egypt, was thus one of several vital points of contact between the main players of the ancient Mediterranean and their wide network of connections – a complex web of trade routes that linked the whole Mediterranean in the Archaic period, from East Greece to the Phoenician coast, Cyprus, mainland Greece, the Nile Delta, North Africa, Sardinia, Etruria, and Spain. More specifically, it connected the two great civilisations of Greece and Egypt. What impact did this role have on the Greeks at Naukratis, on the Greeks back home, and on the Egyptians? Greece and Egypt: Naukratis as cultural crossroads As far as can be judged from the limited research done to date, in spite of the influx of numerous Greeks into Saite Egypt, the Egyptian adoption of Greek elements of culture in the Archaic period seems to have remained rather limited.58 This is exemplified by the relative lack of interest in painted Greek symposion pottery, so popular in many other regions of the ancient world but only rarely found in Egyptian contexts.59 Only transport amphorae were valued not only for their original contents but also as convenient containers for re-use (Hdt. 3.5-7 – note also the Chian amphora from Tell Defenneh with sealings of Amasis: Johnston Fig. 18). Naukratis: Greek Diversity in Egypt | 7

Villing and Schlotzhauer In the other direction, the case was different. Life in Egypt certainly did not fail to make an impression on the Ionian and Carian soldiers in the pay of the pharaoh, and a considerable degree of acculturation is manifested by the adoption of Egyptian motifs and their mixing with Greek/Carian traditions on the grave stelai of Saqquara.60 Carians and Ionians might marry Egyptian women, adopt Egyptian names, be involved in Egyptian cults61 and adopt Egyptian burial customs. Inhumation in a completely Egyptian style is attested, for example, in the late 7th century bc for the son of Alexikles and Zonodote at Tell elNebesheh(?); he even adopted an Egyptian name, Wah-ib-Reem-ahet.62 Returning from Egypt to their home cities (be it permanently or for a visit), both mercenaries and traders, as well as possibly craftsmen, not only brought with them Egyptian goods to dedicate in the local sanctuaries63 (most conspicuous is the Egyptian statue dedicated by Pedon at Priene in the late 7th century bc)64 but also tales of the grandeur of Egyptian temples, Egyptian painting, Egyptian cult and ideas of the afterlife that were to leave a profound influence on those who heard them. For example, as Bilge Hürmüzlü has established recently, the Egyptian idea of the preservation of the body for the afterlife may well be responsible for the introduction at Archaic Klazomenai of inhumation in general and of Egyptian-style sarcophagi in particular, a phenomenon that seems paralleled also in Archaic Samos.65 Such a change is a fundamental transformation of beliefs, not a mere superficial fad, and testifies to the profundity of Egyptian influence on Eastern Greeks. Equally significantly, Egyptian architecture and technology proved fundamental for the development of (East) Greek monumental architecture, such as it is found at Didyma or on Samos,66 and sculpture, such as the monumental lions of Egyptian type at Didyma67 – part of a shared culture of monumentalisation, used not least for political ends.68 Perhaps the most successful of these developments was of course the kouros and kore motif.69 Beyond the realm of art, we also find Egyptian ideas in cosmology or philosophy.70 Phenomena such as the popularity of Egyptian amulets – scarabs and faience71 – further demonstrate the appeal exerted by Egyptian ideas of divine protection on the wider Mediterranean world, which at the time amost seems to have been in the grip of some ‘Egyptomania’.72 That the deep impression made by Egyptian ideas also extended to the medium of Greek pottery is suggested, for example, by the Laconian Arkesilas cup (Schaus Fig. 1); as is set out by Schaus in the present volume, its depiction of the king of Cyrene overseeing the weighing of goods seems to have been influenced by the Egyptian iconography of the weighing of hearts (souls) on entry in the afterlife (Schaus Figs 2–3). In East Greek vase-painting, too, Egyptian motifs appear: we find them on a Fikellura (MileA II) fragment from Naukratis depicting the mythical Egyptian king Bousiris;73 on the amphora from Saqqara featuring a typically Egyptian way of representing bull’s horns mentioned by John Boardman in this volume; in the falcon on the nb basket and the stick-fighters on the situlae from Tell Defenneh discussed in this volume by Sabine Weber (Weber Figs 16–17); and, perhaps most obviously, in the band of cartouches on the Apries amphora from Thebes examined in this volume by Donald Bailey (Bailey Figs 1-5). Representations of black Africans, such as on the (North-Ionian?) fragment from 8 | Naukratis: Greek Diversity in Egypt

Naukratis (book cover), must also ultimately derive from contacts with Africa. With most of these representations having been found in Egypt, it is tempting to suspect that they were locally produced by Greeks in Egypt. Yet as will emerge from the various discussions and analyses in the present volume, on balance and on present evidence it seems more likely that most of these pieces were produced in various East Greek centres. If so, they were clearly produced with Egypt in mind, quite possibly commissioned, even though for what client and what precise purpose remains unclear: a symposium, a dedication in a sanctuary, a prize, a gift?74 What will also emerge, however, is that there was indeed some local production of East Greek style pottery in Naukratis (Schlotzhauer and Villing Figs 30–40). What is surprisingly at first glance, however, is that there is nothing at all Egyptian about this pottery, beyond the use of the local clay. Shapes and decoration are all Ionian, even if rather idiosyncratic – no cartouches, here, or Egyptian symbols.75 This contrasts sharply with the adoption of Egyptian funerary ideas by Carians and Ionians at Saqqara, and also with the influence of such ideas in East Greek cities and the adoption of Egyptian iconography in some East Greek pottery, but also, in Naukratis itself, with the mixture of Egyptian and Greek motifs in at least some products of the Naukratis scarab workshop. It thus seems that the (East) Greek inhabitants of Naukratis admitted Egyptian influence only selectively into their material culture, and, at least in their pottery production, were more intent on expressing and re-enforcing their Greek identity, similar to the way a certain common (East) Greek administrative and cultic identity was shaped in the Hellenion (Höckmann and Möller, this volume). Rather than uniformly encourage acculturation and exchange, the special position of Naukratis, an (East) Greek enclave closely controlled by Egyptians but not integrated into Egyptian society, in fact seems to have encouraged a drawing together and re-enforcement of a Hellenic identity precisely in opposition to the surrounding Egyptian environment. Frequently paralleled in expatriate communities in the ancient and modern worlds, this should hardly surprise, yet the lack of enthusiasm for things Egyptian nevertheless strikes one as paradoxical at a place that was the very heart of Greek contact with Egypt and that radiated out Egyptian influence all across the Greek Mediterranean and beyond. Illustration credits

Fig. 1 drawing Kate Morton; Fig. 2 drawing Marion Cox, after Möller 2000, fig. 1; Fig. 3a © Egypt Exploration Society; Figs 3b-c © Petrie Museum of Egyptian Archaeology, University College London, PMAN 2698, 2683; Figs 4a-e courtesy of John Boardman.

Notes 1 2

3 4 5 6

Braun 1982, 38. We are grateful to D. Williams, S. Ebbinghaus, S. Woodford, V. Smallwood and F. Wascheck for comments on the manuscript of this introduction, to the contributors to the volume for providing information on various questions; to J. Boardman for kindly supplying photographs from Hogarth’s excavations, and to S. Quirke and P. Spencer for identifying and supplying images from Petrie’s excavations. Karetsou 2000. Sourouzian and Stadelmann 2005, 82-3, fig. 6. Cf. especially the extensive discussions by Haider 1988, 1996, 2001, and Kaplan 2002; see also Williams and Villing, this volume. Cf. Haider 1988, 1996, 2001. For a critical view of Haider’s assessment of foreigners in the Egyptian army see Pressl 1998.

Naukratis and the Eastern Mediterranean: Past, Present and Future 7 8 9 10 11 12

13 14 15

16 17 18

19 20

Cf. for example Petrie 1886b, 13, on the destruction of a column fragment from the temple of Apollo, or Gardner 1888, 12-5. On Petrie and his work, see Drower 1985. The history of the excavations as been summarized most recently by Leonard 1997, 1-35; cf. also Möller 2000a, 90-2. Only a very brief account is therefore given here. Coulson 1988, 1996; Coulson and Leonard 1977/8, 1979, 1981a, 1981b, 1982a, 1982b; Coulson et al. 1982; Leonard 1997, 2001. Petrie 1886b; Gardner 1888; Hogarth 1898/9, 1905. Petrie 1888, V. On the history of excavations and distribution of finds, see Cook 1954, 60-1; Bernand 1970, 634-6; Schlotzhauer 2001, 112-13; Höckmann 2001, V-VI; Kerschner 2001a, 72-4. Research into the whereabouts of Naukratis material are still ongoing; collections so far identified are: (in Britain): Bath (Royal Literary and Scientific Institution); Birmingham; Bolton; Bristol (City Museum and Art Gallery); Cambridge (Fitzwilliam Museum and Museum of Classical Archaeology); Dundee (McManus Galleries); Edinburgh; Glasgow (Art Gallery and Museum, Kelvingrove, and Hunterian Museum and Art Gallery); Greenock (McLean Museum and Art Gallery); Liverpool (Liverpool Museum); Harrow School; London (British Museum, UCL and Petrie-Museum); Macclesfield; Manchester (Manchester Museum); Newbury (West Berkshire Museum); Newcastle upon Tyne (Hancock Museum); Nottingham (Brewhouse Yard, Museum of Nottingham Life); Oxford (Ashmolean Museum); Reading (Ure Museum); St Helens (The World of Glass); Southport (Atkinson Art Gallery); (elswhere): Dublin (Department of Classics); Amsterdam (Allard Pierson Museum – from Coll. v. Bissing); Den Haag; Leiden (Rijksmuseum – from Coll. v. Bissing); Brussels (Musées Royaux – from Coll. J. de Mot); Paris (Louvre, from Coll. Seymour de Ricci); Compiègne; Berlin (Antikensammlung); Bonn (Akademisches Kunstmuseum – from Coll. v. Bissing); Heidelberg (Antikenmuseum – from Coll. O. Rubensohn and P. Gardner); Hildesheim; Karlsruhe; Leipzig (from Cambridge, Fitzwilliam); Munich (Antikensammlung – from Coll. v. Bissing); Palermo; Syracuse; Alexandria; Cairo; Athens (BSA; L. Benaki); Moscow; Boston; Brooklyn; Bryn Mawr; Chautauqua; Chicago; Clinton/NY (Hamilton College); New York (Metropolitan Museum of Art – from Coll. E. Price); Philadelphia; San Francisco; Vermont; Toronto; Sydney (Nicholson Museum). We are grateful to U. Höckmann (Mainz Naukratis project) and M. Marée (British Museum, Department of Ancient Egyt and Sudan) for contributing to this listing. Price 1924; Venit 1982, 1988; Bernand 1970; Bissing 1951; Boardman 1999. Möller 2000a; several contributions in TOPOI 12/13, 2005. Höckmann and Kreikenbom 2001. Ursula Höckmann examined the Kouroi of limestone and alabaster, Gabriele Nick the small scale sculpture, Wolfgang Koenigs the remains of architecture from the sanctuaries at Naukratis, and Sabine Weber und Udo Schlotzhauer the Archaic Greek pottery from Naukratis and the rest of Egypt. The results are published in Nick (forthcoming); Höckmann and Koenigs (forthcoming); Schlotzhauer and Weber (forthcoming). The fragment of a sub-geometric oinochoe from Memphis is generally acknowledged to be the earliest preserved fragment: Weber 2001, 136, pl. 21.1. Pace James 2003 and 2005. On Mediterranean chronologies, see most recently Nijboer 2005; Tsetskhladze 2006; as well as Fantalkin, this volume, ns 35, 43, 81. Astrid Möller (2000a, 2001, 2005) in particular has studied the role Naukratis played as a trading emporion in Egypt and has established the way it functioned as a port of trade at the intersection between Egypt and the Mediterranean; her findings do not need be repeated here (note, however, that she argues against a prominent role of Miletos: Möller 2001). On the nature of early Greek trade, see most recently Reed 2004. As Reed argues, early voyaging aristocrats – such as Sappho’s brother Charaxos, known to have sailed to Naukratis with a load of Lesbian wine – are unlikely to have engaged in trade as a regular activity but might have used it as a means of financing ‘sightseeing’ voyages – like the Athenian Solon, said to have travelled to Egypt ‘both on business and to see the country’ (Arist. [Ath. Pol.]). The growth of dedicated trade, by independent and agent traders, from the last third of the 7th century bc onwards, may well be reflected in the developments at Naukratis as attested by Herodotus. Cf. Pébarthe 2005, 172; Bresson 2005; Carrez-Maratray 2005. Leonard 1997, 13; J. Yoyotte in Goddio and Clauss 2006, 316-23.

21 Goddio and Clauss 2006, 92-9; J. Yoyotte in Goddio and Clauss 2006, 316-23; D. Fabre in Goddio and Clauss 2006, 289-303. Finds suggest that the town existed from the 26th Dynasty onwards; they include East Greek trade amphorae as well as East Greek and Corinthian fine-ware pottery: C. Grataloup in Goddio and Clauss 2006, 332-49. 22 Cf. Pressl 1998, 70-3; Posener 1947; cf. also Austin 1970, 27-8; CarrezMaratray 2005, 202-3. The post is attested from the time of Psammetichos II; under Amasis it was filled by Nachthorheb, whose statue is preserved (Vittmann 2003, 220-1, fig. 111). It was complemented by an ‘Overseer of the gate to the Foreign Countries in the North’, who seems to have been in charge of the Eastern Delta region frequented by Phoenicians, and an ‘Overseer of the gates to the Foreign Lands of the Temeh’, i.e. Libyans. The interpretation of the ‘Great Green’ (ouadj our – W3d-wr) as the Mediterranean is still dominant, in spite of a recent re-interpretation as the Nile Delta (Vandersleyen 1999). 23 The situation is not helped by the fact that, to date, the Egyptian finds from Naukratis have not been systematically collected and studied. 24 Cf. Möller 2000a, 108-13. 25 Hogarth 1898/9, 41-3, 45-6, an interpretation considered likely also by Spencer 1996, 1999, and Smoláriková 2000. Just how problematic the archaeological evidence for the site is, is indicated by the fact that in 1903 Hogarth was not able to find the Great Temenos that Petrie had recorded in his excavations: Hogarth 1905, 111-12. 26 Discussed most recently by Möller 2001, 5-11. 27 Leonard 1997, 13; J. Yoyotte in Goddio and Clauss 2006, 316-23. 28 Yoyotte 1982/3; 1992. 29 FGrH 608 F 8. 30 Berlin 7780 dating from the reign of Apries (589-70 bc) and St Petersburg, Hermitage 8499, dating from 554 bc; cf. Yoyotte 1992. An ‘Egyptian from Naukratis’ is also mentioned in the later Lindos decree, cf. Bresson 2005; Möller 2005. 31 As pointed out by Möller 2001. 32 Gorton 1996, 92. 33 We are grateful to Jeffrey Spencer for his identification of this piece. For Egyptian pottery at Naukratis, see also Edgar 1905. 34 There is no reason to assume that the law forbidding Naukratites intermarriage with Eyptians, dating from Hadrianic times, goes back to this early phase; intermarriage is certainly attested for Carians and Greeks elsewhere in Egypt, and Amasis himself is known to have married a Greek princess from Cyrene. The very fact that such a law was needed later on may, in fact, point to intermarriage as a common practice in an earlier period; cf. Braun 1982, 43 35 Cf. Koenigs in Höckmann and Koenigs (forthcoming). 36 Cf. Möller 2000a, 94-113; 2001, with Kerschner 2001, 70; cf. also Höckmann and Möller, this volume. 37 Gardner 1888, 21-9; Höckmann 2001b, 217 n. 2. The dinos stand (sample Nauk 21; Fairbanks 1928, 116 no. 336, pl.37) has parallels in vessels from the Archaic cemetery of Klazomenai; we are grateful to Bilge Hürmüzlü for this information. An Archaic bowl produced by a Greek potter at Naukratis with a votive inscription to Aphrodite (Schlotzhauer and Villing Fig. 40) may also, surprisingly, come from the cemetery, since it bears a modern graffito ‘CEM’ written by the excavators. 38 To date no inscription in Carian script has been identified, although there are some Carian sherds (Williams and Villing Figs 1–2) that presumably were brought by Carians. Whether the few examples of Etruscan bucchero (Naso Figs 3–4) were brought by Etruscans is uncertain. 39 Cf. Paspalas, Attula, Höckmann and Möller, all this volume. 40 Discussed in detail by Schlotzhauer 2005 and 2006, 294-301, and Kron 1984, 1988; cf. also Villing, this volume, on pottery for ritual meals at Naukratis. 41 BM GR 1888.6-1.531: Gardner 1888, 64-5 and pl. 21 (inscr. no. 768); Möller 2000a, 178 no. 4. 42 Cf. Williams 1983a, 185; Williams 1999, 138 and fig. 52 d. 43 Cf. Williams 1983a, 184-6; Johnston, this volume. 44 For an extensive discussion, see Herda (forthcoming b). 45 Venit 1984. 46 Kerschner 2000, 487. 47 Posamentir and Solovyov 2006. 48 As also suggested by Schaus, this volume. 49 Hölbl 1979, 368-73. 50 Cf. also Bellelli and Botto 2002. 51 On Phoenicians in Egypt, see Kaplan 2003, 8-9; Vittmann 2003, 4483; Docter 1997. A Phoenician community at Naukratis has again Naukratis: Greek Diversity in Egypt | 9

Villing and Schlotzhauer

52 53 54 55

56 57 58

59 60 61

been suggested by James 2003, 256-8 (cf. also Yoyotte 1994), going back to ideas of Hogarth and Edgar. Phoenicians are thus credited with the production of carved Tridacna shells, faience and scarabs at Naukratis. Against this supposition, the scarabs produced at Naukratis from the late 7th century bc onwards until the mid-6th century bc and widely exported (cf. Gorton 1996, 91-131; Hölbl 2005) are considered by many experts to have been produced primarily by Greek craftsmen (Hölbl 1979, 141, 207-9), perhaps with Egyptian help (Gorton 1996, 92). As regards tridacna shells, the plain tridacna shells from the site (Petrie 1886b, 35, pl. 20.16,16a; Edgar 1898/9, 49) do not need to have been destined for carving, as undecorated shells were also found deposited in graves in cemetery of Naukratis (Gardner 1888, 29) and are common also in many other sites (Möller 2000a, 163-6). The timber and worked wood mentioned in the stele of Nectanebos I as imports to Egypt passing through the port of Hone, of course, may well stem from Phoenicia or Cyprus; but this only applies to a later date. Cf. e.g. Braun 1982, 41. Schlotzhauer 2006, 301-7, 316 figs 4-6. Torpedo-shaped amphora Petrie 1886b, pl. 16.3. Oxford, Ashmolean Museum G.124, from Hogarth’s excavation in the Hellenion 1903, presumably the piece mentioned by Hogarth 1905, 118, even though he describes the letter as a shin while the Ashmolean fragment seems to show a mem. Hogarth 1905, 124 fig. 3. Schlotzhauer 2006, 305. This is not a topic to which much research has been devoted; however, one gets the impression that Herodotus is generally right in his assessment (2.91) that ‘the Egyptians shun the use of Greek customs’, even if he himself then goes on to mention an example to the contrary, namely the Greek-style athletic games at the Egyptian city of Chemmis. Cf. most recently Weber (forthcoming). We are grateful to the author for supplying a copy of her article before publication. Höckmann 2001b; Kammerzell 2001. Grallert 2001; Höckmann 2001b; Kammerzell 2001.

10 | Naukratis: Greek Diversity in Egypt

62 Grallert 2001. 63 Ebbinghaus 2006. 64 For Pedon, see e.g. Boardman 1999a, 281 fig. 324; Vittmann 2003, 203-6, fig. 103; Höckmann and Vittmann 2005, 100 fig. 2; cf. also Kourou 2004 for Egyptian statuettes dedicated in East Greek sanctuaries. 65 Hürmüzlu 2004b. Note also the fact that Aiolian Larisa seemst to have been home to Egyptian troops retired from service for Cyrus, so that continued contact with Egyptians existed even in the homeland: Xen. Hell. 3.1.7. 66 See e.g. Bietak 2001. 67 Höckmann 2005. 68 See. e.g. Tanner 2003. 69 See e.g. Kyrieleis 1996, 68-86, 108-27. 70 For a detailed discussion, see Haider 2004. 71 Gorton 1996; Hölbl 1979; Webb 1978; James 2003, 251-6. 72 Scarabs seem to have been produced from the late 7th century bc onwards until the mid-6th century bc and were distributed across the Aegean and as far as Italy, Spain, Carthage and the Black Sea region; cf. Gorton 1996, 91-131, and most recently Hölbl 2005. On ‘Egyptomania’, see Ebbinghaus 2006, 201. 73 Discussed in detail by Schlotzhauer and Weber 2005. 74 Dedications especially by Hellenomemphitai and Caromemphitai, Ionians and Carians at Memphis, into Egyptian sanctuaries are certainly attested; cf. Braun 1982, 46-7 fig. 4; Höckmann 2001b. East Greek painted pottery is not normally encountered in Egyptian sanctuaries, but an exception is Saïs: cf. P. Wilson, Saïs Report, March-April, 2003, http://www.dur.ac.uk/penelope.wilson/ 3g2003a.html (27 June 2006); Weber (forthcoming). Other instances of Greek painted pottery in connection with Egyptian towns and burials are cited by Weber 2006. The possibility of ‘prize vases’, raised by Herodotus’ mention (2.91) of Greek-style gymnastic contests at Chemmis in the district of Thebes, is discussed most recently by Decker 2003. 75 The same conclusion (labelled with the term ‘Beharren’) is also reached also by Schlotzhauer in Schlotzhauer and Weber 2005, 80-1.

The Hellenion at Naukratis: Questions and Observations Ursula Höckmann and Astrid Möller Abstract The Hellenion at Naukratis still poses questions as to its development, its function, the gods being venerated there and what kind of sanctuary the Hellenion might have been. This contribution wishes to make some observations towards solving some of these questions.* In his description of Naukratis, Herodotus places special emphasis on a sanctuary named the Hellenion. To Herodotus, it was the largest, the best known and the most visited of all sanctuaries at Naukratis. It was set up by the joint efforts of nine poleis: the Ionian poleis of Chios, Teos, Phokaia, and Klazomenai, the Dorian poleis of Rhodes, Knidos, Halikarnassos, and Phaselis, and the Aiolian polis of Mytilene (Fig. 1).1 Herodotus alludes to a conflict between those and other poleis that claimed an unjustified share in the Hellenion and the right to appoint the prostatai tou emporiou, the chief officials of the emporion.2

In 1899, David G. Hogarth started excavating a temenos in the north-east of Naukratis which he identified with the Hellenion mentioned by Herodotus.3 It proved rather difficult to come to clear conclusions as a result of the high water table and the level of destruction discovered during the excavation. The whole temenos might have covered an area measuring about 150 x 100m, surrounded by a massive mud brick wall, but further interpretation involves a lot of guess-work. Both the literary and the archaeological evidence are meagre, causing many hypotheses and arguments to flourish. Indeed, doubts as to whether the Hellenion has been identified correctly have not ceased.4 The archaeological evidence, architecture and finds alike, poses questions as to the kind of sanctuary the Hellenion was. Apart from dedications to the ‘Gods of the Hellenes’ in their entirety, the dedications to different Greek gods at the Hellenion might indicate cults branched off from particular cults celebrated in the various East Greek poleis participating in the Hellenion. Figure 1 The nine ‘founder cities’ of the Hellenion

Naukratis: Greek Diversity in Egypt | 11

Höckmann and Möller Figure 2 The Hellenion

1. The Archaeology of the Hellenion (A.M.) The excavation and the later survey at the site of the Hellenion point to three phases of development. Remnants of sandstone structures at 7, 9, 16 and 17, all located underneath the later chambers, should be attributed to the oldest phase (Fig. 2).5 Their orientation is parallel to the walls of chambers 3, 4, and 5; chamber 3 is dated to the oldest phase by a North Ionian LWG vessel fragment of the second quarter of the 6th century bc.6 In a large earthenware basin (18) which is now suggested to have been a basin for ritual cleansing7 lay the limestone relief of a hoplite warrior.8 The relief, dated to the end of the 6th century bc, seems to be of the same material as the ashlars with dedications reported by Hogarth.9 The basin, however, was deposited at its later find spot only secondarily and the relief was placed in it later, so that it does not permit any secure connection with the earliest structures of the Hellenion of about 570 bc. The next phase is indicated by chambers 10, 11, 14, 15, 19, and 20, forming a long row of rooms. The walls were made of mudbricks. On what remained of the floors, terracotta figurines and red-figure sherds from the late 5th century bc were found. The orientation of the chambers slightly changed, indicating a reconstruction sometime before the end of the 5th century bc. Chambers 12, 13, 14, and 14a should be seen in connection with these chambers; some finds made in this area point to the 4th 12 | Naukratis: Greek Diversity in Egypt

century bc. Chambers 57, 59, 61, 64, and 65 follow the same orientation, and black- and red-figured sherds were found here. In chamber 64 was discovered the Stesichoros kylix, attributed to Onesimos.10 The stone foundations found under the Ptolemaic chamber 63 belong to this phase. These rested on about 60cm of older objects, including Chian sherds. In the Ptolemaic chamber 58, too, an older floor was discovered. Hogarth11 maintained that the structures of the Hellenion ‘were restored practically from the foundation in the first half of the 5th century’, although he did not mention signs of destruction. Instead he quoted Gardiner, who reported traces of a ‘calamity’ at the Aphrodite shrine that befell Naukratis at the time of Cambyses’ conquest in 525 bc, and deduced from the lack of early red-figured vases that something similar must have happened to the Hellenion as well. If the first sandstone Hellenion was destroyed in 525 bc, then, in the 440s or 430s bc, when Herodotus visited Naukratis,12 he must have admired the mudbrick building, the second Hellenion, which probably was set up sometime during the first half of the 5th century bc.13 The Ptolemaic Hellenion was constructed on a sand-bed of between 60cm to 2m thickness. This sand deposit could be seen especially well in the north-east of the Hellenion. The complex of chambers 24, 25, 26, 27 adjoining the long passageway 28 belongs to the Hellenistic period. The pottery finds in these rooms cover the 3rd century bc to Roman times.

The Hellenion at Naukratis: Questions and Observations

Figure 3 Fragment of the handle plate of a Late Corinthian column krater, Cambridge, Fitzwilliam Museum GR.31.1899

2. Dedications found at the Hellenion (U.H.) According to the pottery finds, the Hellenion was in use from the 6th century bc down into the Ptolemaic or even the Roman period.14 The earliest finds date from the late second quarter of the 6th century bc. This may be connected with Herodotus’ statement (2.178.1) that the pharaoh Amasis (570–525 bc) gave the Greeks at Naukratis the right to set up sanctuaries to their gods. The other sanctuaries display older pottery finds, making it likely that Herodotus did not distinguish between different pharaohs, thus crediting only Amasis with benefactions to the Greeks. The Hellenion, however, seems to have been established following a possible reorganisation of the emporion by the pharaoh Amasis. Among the earliest items in the Hellenion there is a fragment of the handle plate of a Late Corinthian column krater (Fig. 3).15 Its exact find-spot is not known. Hogarth, however, states that ‘Corinthian pottery was comparatively scarce, and most of the fragments belonged to large craters, chiefly of the red-clay variety. Several crater-handles, […], were obtained, with representations of male and female heads, sirens, &c.’16 We should add the North Ionian LWG fragment of the early second quarter of the 6th century bc with dedicatory inscription by a Chian to Apollo.17 In 1903, Hogarth found a rare Chian chalice, now in Oxford, to the west of spot 66.18 Eleanor Price states that ‘the violet ground with white spots is unique’.19 Anna Lemos suggests a date in the late second quarter of the 6th century bc.20 The two antithetical riders with their spears seem to be a very rare motif in Chian pottery. They possibly represent the Dioskouroi.21 For a clearer picture of the Hellenion and its dedicators in the Archaic period it would be interesting to know the percentages of South and North Ionian as well as Attic pottery, a wish that can never be truly fulfilled, even though we might get a better idea once the database of Naukratis pottery is completed. Besides drinking vessels and kraters of the Archaic period, there are abundant finds from the 5th century bc and later, Late Archaic and Classical Attic pottery, and terracotta protomes and statuettes mostly connected with the worship of Aphrodite and, in several cases, of Herakles.22 An Attic black-glaze cup of 5thcentury date was dedicated to Aphrodite by Teleson of Rhodes.23

Figure 5 Inscription on the limestone base Figure 4

Archaic small-size sculptures of limestone or alabaster of Cypriot style have not been found in the Hellenion.24 But there are two classical limestone statuettes of the 5th and 4th centuries bc. The limestone base of a statuette of Aphrodite dedicated by Deinomachos from Mytilene which was found near spot 3 is apparently lost.25 The base of a Herakles statuette dedicated to him by Aristion,26 a work of the sculptor Sikon of Cyprus, which was found broken in two parts, was discovered between spots 7 and 9 (Figs 4–5). The only two Cypro-syllabic inscriptions from Naukratis, on Attic black-glaze cups of the late 5th or early 4th century bc, have been found in the same area as the Herakles base (Fig 6).27 Among numerous Hellenistic terracotta statuettes some Graeco-Egyptian ones may be mentioned.28 Their exact find-spot is, however, unknown. As far as I know, no Egyptian bronze statuettes have been found in the Hellenion.29 Some limestone and sandstone ashlar blocks with dedications may be supposed to have been excavated by Hogarth in the Hellenion.30 Since none was found in situ, their original setting is unknown, nor do we know their current whereabouts. 3. The dedicatory inscriptions (A.M.)31 In the Hellenion only a few dedicators can be identified by name. Apart from the above-mentioned dedications by Deinomachos of Mytilene and Teleson of Rhodes, there are some more personal names, but usually without an ethnikon.32 From other sanctuaries at Naukratis we have dedications by Chians, Teians, Phokaians, Cnidians, and Mytilenians.33 It seems that dedicators from poleis that later participated in the Hellenion first dedicated to gods venerated in the other sanctuaries. A clearer picture of the dedicators’ behaviour towards the different sanctuaries will hopefully be gained from the catalogue of inscribed pottery being prepared by Alan Johnston as a part of the Naukratis pottery database. In the area that we now call the Hellenion, Hogarth found several vase fragments dedicated to the ‘Gods of the Hellenes’. This prompted him to identify the north-eastern temenos with the Hellenion mentioned by Herodotus.34 To restore the standard formula for these otherwise unique dedications is hampered by the fact that we only have tiny fragments and not a

Figure 6 Bottom of a bowl or plate (inside and outside and profile drawing); London BM GR 1900.214.17 Figure 4 Limestone base of Classical statuette of Herakles, dedicated by Aristion, signed by Sikon of Cyprus, London BM GR 1900.2-14.22

Naukratis: Greek Diversity in Egypt | 13

Höckmann and Möller

Figure 7 Pottery fragments with dedications to the ‘Gods of the Hellenes’

single inscription that gives us the full wording. Neither does there seem to be epigraphic material from other sanctuaries with which to compare these dedications.35 There are three possibilities as to the restoration of the dedicatory formula: (a) toi=j qeoi=j toi=j (Ellh/nwn (Fig. 7)36 (b) toi=j (Ellh/nwn qeoi=si 37 14 | Naukratis: Greek Diversity in Egypt

(c) tw=n qew=n tw=n (Ellh/nwn 38 Some inscriptions could theoretically be restored as toi=j qeoi=j toi=j (Ellhni/oij – ‘To the Hellenic Gods’,39 but this is not substantiated by the evidence, as not a single fragment among the finds known today shows this dative ending. We have, however, the genitive ending –wn.40 Since in all cases the article is lacking, we cannot tell whether this could possibly be part of

The Hellenion at Naukratis: Questions and Observations the formula tw=n qew=n tw=n (Ellh/nwn – ‘Of the Gods of the Hellenes’, of which we have two fragments.41 The restoration toi=j qeoi=j toi=j tw=n (Ellh/nwn,42 however, cannot be correct, because (Ellh/nwn being an attributive genitive requires, if following the noun, the repetition of the article. That it is unlikely to be toi=j qeoi=j toi=j tw=n (Ellh/nwn hinges on the fact that (/Ellhnej is an ethnic name which is usually given without the article.43 We need to consider, however, that in dedicatory formula such as these short graffiti on vases, a formula might be used which might not be grammatically correct. The semantic use was important. Therefore, the restoration favoured here is the reading toi=j qeoi=j toi=j (Ellh/nwn – ‘To the Gods those of the Hellenes (I mean)’, giving Hellenes as attributive genitive, thus placing special emphasis on the fact that these are the Greek gods and no others. In a Greek sanctuary in Egypt this is particularly apt. The formula toi=j qeoi=j toi=j (Ellhni/oij, though not substantiated by the epigraphic evidence at Naukratis, is, however, the formula used in Herodotus. When Herodotus (5.49.3) relates the story of Aristagoras, the tyrant of Miletus, coming to Sparta and trying to persuade their king Kleomenes to fight the Persians, he has Aristagoras imploring the Spartan king to free the Ionians from slavery, especially since they are their blood-brothers, by ‘the gods of the Hellenes’ (oi( qeoi\ oi( (Ellh/nioi). Later in Herodotus (5.92g.1), Sosikles, an envoy from Corinth, speaks against the Spartan plan to re-install the former tyrant Hippias at Athens. He beseeches the Spartans not to give tyrants to the Greek poleis, calling as witnesses again ‘the gods of the Hellenes’ (oi( qeoi\ oi( (Ellh/nioi). Both contexts make it quite clear that in Herodotus’ usage the gods of the Hellenes were evoked when different Greek tribes should be acting together. Hugh Bowden, who argues against the identification of the north-eastern temenos with the Hellenion, favours the restoration toi=j qeoi=j toi=j (Ellhni/oij, although he admits that toi=j qeoi=j toi=j (Ellh/nwn is far more likely. He considers (Ellh/nioj as a formal cult title which should have been used in dedications.44 He knows, however, only one case where (Ellh/nioj is used as a cult title and this is the cult for Zeus Hellanios on Aigina.45 Such a cult title at Naukratis could, however, only be connected to the Hellenion, relating to a plurality of gods belonging to all Greeks. toi=j qeoi=j toi=j (Ellh/nwn – ‘To the Gods those of the Hellenes’ Bowden considers as a more general phrase which ‘would seem to be a convenient grouping for a dedication at any sanctuary, by any dedicator, not the indicator of a specific cult’.46 With this line of argument he wants to support his disagreement with the identification of the Hellenion, which he suggests was either not found or was situated where one usually locates the

Figure 8 Foot of Ionian vessel; London BM GR 1900.2-14.5

Figure 9 Profile drawing of London BM GR 1900.2-14.5

Figure 10 Attic black-glaze kylix dedicated to Herakles by Artemon; London BM GR 1900.2-14.16

Figure 11 Profile drawing of London BM GR 1900.2-14.16

Figure 12 Fragment of Milesian kylix dedicated to Herakles; Oxford G 141.58

Hera sanctuary.47 But it is exactly this formulation of the dedicatory inscriptions, which distinguishes the Gods of the Hellenes from those of the Egyptians, and the discovery of about two dozen such dedicatory inscriptions to the Gods of the Hellenes within this temenos, that makes it its identification with the Hellenion so likely.48 Unfortunately, we cannot tell who dedicated these vases to ‘The Gods of the Hellenes’. All identified inscriptions were written on cups or bowls, all drinking vessels, the earliest on an Ionian cup that dates from the 6th century, but the majority date from the 5th century bc. 4. Deities and cult filiations (U.H.) As far as can be judged by the dedicatory inscriptions scratched or painted on pottery found in the sanctuaries of Naukratis, most cults seem to have been transferred from East Greek cities. This is clear for Samos49 and Miletus.50 This topic, however, has not yet been thoroughly investigated and my ideas expressed in the following pages remain only tentative.51 Among the dedicatory inscriptions on pottery from the Hellenion only a few deities are mentioned. There are several dedications to the Dioskuroi52 and the above-mentioned ‘Gods of the Hellenes’. Dedications to Apollo53 – in the context of the Hellenion never with an epiclesis54 – to Herakles (cf. infra), to Artemis55 and to Poseidon, are rarer.56 There are Late Archaic dedications to Aphrodite, one of them on an Ionian Late Archaic vessel foot to Aphrodite py…57 (Figs 8–9), one on the rim of an Attic Late Archaic red-figured volute krater (Figs 13–14)58 to Aphrodite Pandemos. The latter was found near the south wall of the Hellenion at 39.59 Two more Late Archaic Attic cups dedicated to Aphrodite Pandemos have been found in the sanctuary of Aphrodite in the south of Naukratis (Figs 15–18).60 Aphrodite and the Dioskuroi were probably worshipped, too, as seafarers’ deities.61 The epiclesis Pandemos refers to another aspect of Aphrodite. As Andrew Scholtz has demonstrated, this Aphrodite had a ‘broad-based appeal in connection with economic activity’62 and the famous hetairai of Naukratis.63 I think that the two dedications to Aphrodite Pandemos found in her sanctuary in the south were addressed to the same deity as the Aphrodite Pandemos in the Hellenion. The same can be supposed for the Dioskuroi, who are worshipped in a little temenos in the northwest of Naukratis64 as well as in the Hellenion. Naukratis: Greek Diversity in Egypt | 15

Höckmann and Möller

Figure 13 Two rim fragments of Attic red-figured volute-krater; London BM GR 1900.2-14.6 and Bonn,Akademisches Kunstmuseum, inv. no. 697.90

Figure 14 Profile drawings of BM GR 1900.2-14.6 and Bonn,Akademisches Kunstmuseum inv. no. 697.90

Herakles seems to have been venerated at the Hellenion at Naukratis since the mid-6th century bc; his cult is likely to go back to contacts with East Greece. The Busiris adventure can be taken as having been invented by East Greek mercenaries in Egypt.65 This becomes apparent not only from the Caeretan hydria in Vienna,66 but also from Attic vase-painters who represent the gigantomachy with Herakles stepping on the pole of his chariot in imitation of reliefs of Rameses III.67 Such emphasis being given to the aspect of the victorious hero would have been well understood among mercenaries. The earliest evidence for a cult of Herakles at Naukratis are two kylix fragments68 with dedicatory inscriptions, one of them Attic (Figs 10–11),69 the other from a Milesian kylix dated to about the mid6th century bc (Fig. 12).70 A dedication from the 5th century bc at Naukratis is reported on a limestone ashlar block.71 The 4th century bc is represented by the limestone base mentioned above (Figs 4–5). It is usually referred to as a Cypriot work of the 6th century bc,72 although F.H. Marshall as early as in 1916 dated the inscription to the 4th century bc. The remaining parts of the statuette show that Herakles stood relaxed with his left foot slightly forward of the right, weight-bearing, leg, his club resting on the ground beside his right foot. The statuette can be imagined as having been similar to a bronze statuette of the 4th century bc in the Louvre73 or a Roman statue in the Palazzo Pitti in Florence,74 with the apples of the Hesperides in the left hand, alluding to the apotheosis of the hero. The inscription names a certain Sikon as the sculptor (Fig. 5).75 The third word being reconstructed as Kyprios, ‘from Cyprus’, presents a problem, since no such statuette in the Classical Greek posture is known from Cyprus, only a different type.76 Since limestone was used in Egypt, the Classical Sikon base might go back to a Classical Naukratite work by a sculptor from Cyprus. Leaving these considerations aside, it attests a cult of Herakles at the Hellenion in the 4th century bc, perhaps its earlier half.77 This cult is further confirmed by some Hellenistic terracotta statuettes of Herakles from the Hellenion.78 All this points to a cult of Herakles in the

Figure 15 Fragment of Attic bowl; London BM GR 1888.6-1.211

Figure 16 Profile drawing of London BM GR 1888.6-1.211

16 | Naukratis: Greek Diversity in Egypt

Figure 17 Fragment of Attic small type C cup; London BM GR 1888.6-1.212

Figure 18 Profile drawing of London BM GR 1888.6-1.212

Hellenion from about the middle of the 6th century bc until Hellenistic times. As yet, it is impossible to tell which city instituted the cult of Herakles in Archaic times in the Hellenion. Erythrai with its old and famous cult of this hero might be a likely candidate,79 although the polis is not explicitly mentioned by Herodotus. Further investigations will perhaps show continuity of cults for the other deities venerated in the Hellenion as well, including Aphrodite.80 5. Aphrodite Pandemos at Naukratis (A.M.) There is one dedication81 to Aphrodite Pandemos from the Hellenion showing the following inscription (Figs 13–14): 82 )Afrod]i/thi : Pandh/m[wi on a rim fragment of an Athenian redfigured volute krater of around 500 bc. The text is incised in an almost lapidary manner which seems, however, to have been applied after firing. D. Williams noticed a rim fragment at Bonn with the inscription – ]AMM[ – that belongs to this krater.83 In the Aphrodite temenos in the south of Naukratis, two dedications to Aphrodite Pandemos have been found: (1) )Afrodi/thi] Pandh/mwi on an Attic type C cup of 500–480 bc (Figs 15–16).84 Gardner observed that the dedication was probably incised after the vase was broken. (2) )Afrodi/thi P] andh/m[wi on an Attic type C cup of the first half of the 5th century bc (Figs 17–18).85 As to the interpretation of the presence of Aphrodite Pandemos, two main lines of argument can be distinguished: the political and the erotic.86 Recently, Andrew Scholtz,87 being unsatisfied with the political interpretation88 in the context of Naukratis, has argued that at Naukratis, we are not dealing with an Aphrodite ‘of the whole demos’, but with an Aphrodite ‘for all people’, a kind of ‘general access’ goddess concerned with trade and prostitution. His line of argument runs against an early civic character of Naukratis and emphasizes its cosmopolitan traits.89 Although I do sympathize with Andrew Scholtz’s results, I should like to follow a different line of argument here, which might be called a ‘history of ideas approach’, resulting in an interpretation related to cultural identity. Ever since Plato in his Symposion (180D–181C) distinguished between Aphrodite Ourania and Aphrodite Pandemos, there have been two interpretations concerning the character of Aphrodite Pandemos. Plato distinguished an older goddess, the motherless daughter of Ouranos, called Ourania, from a younger one, daughter of Zeus and Di0 ¯ ne¯, called Pandemos. Pandemos’ Eros is therefore called Eros Pandemos or ‘vulgar love’; he is responsible for love among the common and uneducated (fau=loi) of human beings.90 Xenophon91 followed Plato in this distinction. He considers the possibility that there might only be one Aphrodite, because Zeus has several epithets (e)pwnumi/ai), too. But since each of the

The Hellenion at Naukratis: Questions and Observations two Aphrodites has her own altar and temple and Pandemos receives less pure sacrifices than Ourania, there should be two goddesses. Pandemos sends us, Xenophon goes on to reason, bodily love, Ourania on the contrary the love of the soul, the love of friendship, and the love of good deeds. Taking up these philosophical reflexions, Aphrodite Pandemos appears closely connected with bodily love, which matches very well the presence of the renowned hetairai at Naukratis. The distinction drawn up by Plato, however, is not necessarily an original distinction, but rather the result of moral reflexion. The other interpretation of Aphrodite’s epiclesis, the political one, also goes back to antiquity. Theseus is said to have founded the cult of Aphrodite Pandemos and Peitho at the time of the Athenian synoikism.92 Athenaios93 relates the story that Solon introduced brothels to Athens and from the profit was said to have founded the sanctuary of Aphrodite Pandemos. This story clearly draws on an understanding of Aphrodite Pandemos as providing erotic entertainment for all people. The connection between Aphrodite Pandemos and the synoikism gave way to the political interpretation. The goddess is made responsible for civic unity, sponsoring synoikism, and political structure. But even here, the erotic aspect slips in and makes her responsible for the procreation of the oikoi united in the polis.94 Although the cult of Aphrodite Pandemos is mythologically connected with Theseus, the evidence is not early. Her sanctuary is located on the slopes of the Acropolis, directly below the Nike temple.95 The earliest inscription found there, which does not, however, show the epiclesis, dates to the beginning of the 5th century bc; from around 287/6 bc, we have a lex sacra from the sanctuary of Aphrodite Pandemos.96 According to Pausanias, the sanctuary at Athens belonged both to Aphrodite Pandemos and Peitho.97 Peitho is the goddess of persuasion and in particular that of erotic persuasion. Therefore, Peitho appears also as an epiclesis of Aphrodite. It seems as if Pausanias, being the only one making the cult connection between Pandemos and Peitho for the Athenian sanctuary, has divorced the epiclesis Peitho from Aphrodite, thus referring to two goddesses. This would indicate that Aphrodite was worshipped under two aspects at Athens, the erotic aspect perhaps being much stronger than that of civic unity. The political interpretation of Pandemos is closely connected to Athens and no wonder if one considers the Athenians’ civic ideology. The Athenians, however, did not play a big role at Naukratis, as far as we know. There were closer connections with the Greeks from North Ionia, where Aphrodite Pandemos had a temple at Erythrai.98 Since, however, the inscription on the Athenian red-figured krater (Figs 13–14) certainly looks very special, one might be tempted to consider whether it had a special status. Would it be possible that it was dedicated to the Hellenion at Naukratis during the process of cult filiation? If one takes the different meanings of the adjective pa/ndhmoj or the adverb pandhmei=, collected by Scholtz in his article,99 it is obvious that in civic as well as non-civic and extracivic contexts, the meaning of pa/ndhmoj extends beyond one single demos. The prefix pan- takes on an inclusive meaning, integrating Greeks from different poleis. Interestingly enough, Harpocration interprets Aphrodite’s epithet as pa/gkoinoj –

‘common to all’. This word is already found in Pindar where it is used in the Olympic context as pa/gkoinoj xw/ra – the common land of Olympia.100 Scholtz is probably right in reminding us not to reduce Pandemos to some pan-Hellenic essence.101 Of course, if the dedicators at Naukratis knew the Athenian myth of Theseus and gave Pandemos a political, cohesive meaning, they might well have played with a pan-Hellenic idea. But if the inscriptions are early Classical at the latest, this does not seem very likely – unless one were to assume that the special background of Naukratis itself fostered a particular sense of unity among the Hellenes. The Hellenion, if we follow Herodotus, was a place where all three ethnic groups of Hellenes worshipped together and organized some form of administration for the emporion. This is not too unusual in itself, as any joint venture between Greek states was normally carried out via a sanctuary. Whether we should compare this organisation of the Hellenion with an amphictyony, as suggested by Bowden,102 remains to be discussed. The Hellenion at Naukratis, then, clearly evokes interpretations of a pan-Hellenic cult or even of panHellenism.103 The Hellenion was a joint venture of very different Greek poleis, but what about Pan-Hellenism? 6. Pan-Hellenism (A.M.) In discussing pan-Hellenic sanctuaries and pan-Hellenism more generally, Naukratis seems to provide an attractive case. In her article ‘The origins of pan-Hellenism’, Catherine Morgan starts from the idea of pan-Hellenic sanctuaries like Olympia and Delphi to look for the beginnings of panHellenism.104 She points to the confusion the archaeological application of the term pan-Hellenic causes, as the appearance of votives from a number of different areas does not automatically constitute a pan-Hellenic sanctuary. It is necessary to identify the interests of the dedicators and their social position and relationship to the community that controlled the sanctuary. Her idea of pan-Hellenic sanctuaries is modelled on the examples of Olympia and Delphi. A panHellenic sanctuary surely involves being a centre of interstate communication. In 2003, Irad Malkin published an article on ‘Pan-Hellenism and the Greeks of Naukratis’105 in which he maintains the articulation of a pan-Hellenic identity at Naukratis. He holds Naukratis to reveal the Egyptian view of foreigners as Greeks (i.e. the outside view of ethnic identity), an accommodated articulation of Greek identity among Greeks of varying origins (i.e. the inside view), and he assumes also a generalized, selfreferential Greek identity in relation to Egyptians (i.e. the identification of oneself in difference to others). Even if the details of the argument are not entirely convincing, this seems to be the right approach to the question of Hellenic identity, though not an assumed pan-Hellenic one. At the APA-meeting of 2004, Denise Demetriou gave a paper106 about ‘Negotiating Identity: Group-Definition in Naukratis.’ She takes Naukratis as an example of a polyvalent mode of self-definition. The Hellenion and its name show that the Greeks of Naukratis ultimately formed a cohesive Hellenic identity and collectively opposed themselves to the Egyptians. There remains, however, the question whether Naukratis, as Naukratis: Greek Diversity in Egypt | 17

Höckmann and Möller Malkin claims, points to an explicit concept of pan-Hellenism? What then is pan-Hellenism? First of all, this term was coined by modern scholars to describe the various attempts made by intellectuals of the late 5th and early 4th centuries bc to promote Hellenic unity and to prevail over differences between the poleis in attacking the enemy that was Persia. These appeals to greater Hellenic unity were hardly a goal in themselves, but a tool of political propaganda serving the hegemonial and imperialistic aims of Athens.107 In the 4th century bc, the definition of Greekness was rather Athenocentric. Xenophon and Isokrates, the champions of Pan-Hellenism, constantly evoked the relationship between Hellenes – under Athenian leadership – and non-Hellenes, or barbarians, i.e. the Persians. The term Panhellenes, however, appears early in Homer (Il. 2.530), which might be an interpolation, but there are also references in Hesiod (Op. 528) and Archilochos (fr. 102 West). It was used as a term for Greeks.108 Pan-Hellenes at this time does not necessarily imply an ethnic concept of Hellenes: it probably designated all who settled in Hellas. To sum up: Pan-Hellenism is a modern term for a concept introduced around 400 bc and connected to Athens’ renewed imperialistic ambitions. The pan-Hellenic sanctuaries such as Olympia and Delphi provide a different case. It does not matter here just from which time on we are allowed to call them panHellenic. At Naukratis, the Hellenion can hardly be attributed to pan-Hellenic ambitions or a pan-Hellenic status such as Olympia or Delphi. The temenos was proudly called Hellenion109 and people from different backgrounds dedicated to the gods of the Hellenes – a sign of something like a cohesive identity in the face of the surrounding Egyptians. 7. The Hellenion as a place for ritual dining (U.H.) The vase shapes found at the Hellenion, such as bowls, chalices, skyphoi and kraters, all belong to the category of drinking vessels and mixing bowls of the Archaic and Classical periods. These vases were typically used during sacrificial meals.110 Perhaps plates and the mortaria studied by Alexandra Villing111 could be added to this collection of ritual ware. Fine painted pottery, as for instance the Chian chalices, was actually used for sacrificial meals or ritual dining, but it was also dedicated as a thing of beauty.112 The architectural remains of the Hellenion include some chambers aligned in a kind of row (Fig. 2). They unfortunately can only be discussed with some reservation since Hogarth’s plans are not very detailed. As to the chambers, Hogarth held that ‘distinct groups of chambers were devoted to distinct deities’.113 Other scholars interpreted the chambers as small treasuries,114 living quarters of priests,115 magazines,116 or administrative offices.117 There might be a chance that further scrutiny of the excavation reports could contribute details essential for solving these problems. For the moment, I should like to add a vague suggestion: Could some of these chambers be banquet-rooms like the rooms in the south-east building II 2 in the sanctuary of Aphaia at Aigina,118 or the chambers in the sanctuary of Demeter and Kore at Corinth?119 At Naukratis no reliable evidence exists on which such an interpretation of the Hellenion’s chambers could be firmly based. Moreover, all ‘chambers’ date from the Classical and especially from the Hellenistic periods; it is unknown whether Archaic ones had existed. There are no direct traces of dining 18 | Naukratis: Greek Diversity in Egypt

couches. Hogarth describes a number of floors of ‘hammered mud’,120 to some of which a layer of concrete had been added.121 Such floors are characteristic of banquet rooms.122 Finally, remains of coloured stucco in chambers 19 and 20123 call for attention. All these details relate to rooms of the Classical and especially of the Hellenistic periods. Hogarth saw similarities between the chambers in the Hellenion of Classical and Hellenistic times and rooms in Egyptian temples of the Fayum and took their arrangement and architectural structure for Egyptian.124 The irregular layout of the small chambers, however, does not show any similarity to Egyptian temples, which also holds true in respect of the late temples at Karanis and Dionysias as cited by Hogarth.125 The assumption of ritual dining taking place at the Hellenion at Naukratis seems to be supported by traces of sacrifices, namely remains of ashes,126 ‘burnt stuff’ in spot 2,127 and wood, for instance in the southern part,128 west of the great red granite door-jamb 1, in well 35.129 Furthermore, altars130 can be postulated, although no finds can confidently be connected with an altar. Hogarth, however, believed that an altar may have stood above a square depression in the ‘basal mud’, filled with sand, in spot 6.131 The depression measured c. 2m x 1.70m. The stratum, unfortunately, was badly disturbed. The basal mud favours an Archaic dating of that feature. Moreover, there seem to have been basins for the essential ritual cleansing.132 All this evidence, taken together, suggests that ritual dining at the Hellenion in Archaic and Classical times is more than likely. 8. The Hellenistic Hellenion as a place for public dining and political administration (A.M.) During the Ptolemaic period the Hellenion saw major reconstructions, the pattern of its rooms, however, does not seem to have been changed.133 If ritual dining took place here during the Archaic and Classical periods, the same would hold true for the Hellenistic period. During Hellenistic times, the citizens of Greek poleis were provided with banquets by wealthy benefactors.134 It is from the Hellenistic period that we have the first evidence for a prytaneion at Naukratis. Athenaios (4.149D-150B) reports what he found in Hermeias’ second book on Apollo Gryneios: ‘The Naukratitai feast together in the prytaneion on the birthday of Hestia Prytanitis, the Dionysia, and at the festival of Apollo Komaios. They all attend in white garments called prytanikai esthêtes – garments for the Prytaneion. […] No women are allowed in the prytaneion except the aulétria, the female flute player.’ Athenaios quotes a lot about restrictions in the consumption of food and wine, too, but this is of no interest here.135 The author Hermeias136 is only known from this quotation, and Athenaios, of around 200 ad, is not a valuable terminus ante quem for the Hellenistic or even the Classical period. Of course, Athenaios would have picked out any work mentioning customs at Naukratis, his home town. Hermeias mentioned a cult of Apollo Komaios,137 of which we unfortunately do not have a trace among the inscriptions. Fritz Graf138 derives his epiclesis either from kômos – the procession of revellers in honour of a god, often connected to Dionysus, or kômê – the village or district which makes Apollo Komaios the god of the kômê and thus an eminent political deity. Graf admits that we do not hear of Apollo Komaios in the

The Hellenion at Naukratis: Questions and Observations context of spatial or political divisions. At Naukratis, however, he is honoured by a symposium in the prytaneion which surely has political connotations, even if we believe the prytaneion to be inside the Hellenion and the symposion a ritual banquet – the function of these feasts is the social and political cohesion of the group. All festivities mentioned in Athenaios’ passage are connected to the prytaneion. Since no prytaneion has been found at Naukratis, however, we are not able to date this civic building with any certainty. From Hellenistic times on, when Naukratis had gained the status of a Greek polis within the Ptolemaic kingdom, such a public building would certainly have been required. Herodotus, however, does not mention a prytaneion, which suggests that before the second half of the 5th century bc such a building is unlikely. What Naukratis needed as an emporion concerning administration was probably dealt with by the prostatai of the emporion having their office possibly in the

Hellenion.139 Only in Hellenistic times do we have evidence for civic institutions such as bouleutai140 or timouchoi.141 9. Summary A closer scrutiny of the dedicatory inscriptions to the ‘Gods of the Hellenes’ found by Hogarth in the north-eastern temenos confirms its identification as the Hellenion, the place where the Greeks from different poleis venerated different gods such as Herakles and Aphrodite and the Greek gods as a collective. The dedicatory formula seems to particularly emphasize that the Gods of the Hellenes were invoked and not the Egyptian gods. It shows a consciousness of belonging to a common Greek culture, but does not involve ideas of pan-Hellenism. Thus, the Hellenion seems to be the place where a Greek identity could grow142 in contrast to the Egyptian culture, which nevertheless provided inspiration and stimulation to the Greek culture.

Appendix (A.M.) Vases dedicated to the ‘Gods of the Hellenes’ found in the Hellenion No.

1 2 3 4 5

A: Hogarth 1898/9 B: Hogarth 1905 C: Bernand 1970, 2 A 71/C 594 A 81/C 604 A 95/C 618 A 19/C 541 C 350

6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

A 14/C 536 A 15/C 537 A 16/C 538 A 17/C 539 A 97/C 620 A 72/C 595 A 18/C 540 A 79/C 602 A 80/C 603 A 64/C 587 A 73/C 596 B 4/C 647 A 74/C 597 A 75/C 598 A 76/C 599 B 3/C 646 B 1/C 644

23 24 25 26 27

inv. no.

Cambridge, Fitzwilliam GR.103.1899 Oxford G 141.57 BM GR 1911.6-6.13+36 BM GR 1911.6-6.39 Petrie 1886, 62 no. 690 pl. 35: found between the temenos of Apollo and that of the Dioskouroi BM GR 1911.6-6.22

BM GR 1911.6-6.28 Oxford G 141.50 BM GR 1911.6-6.40 Oxford G 141.56

Shape (compare Sparkes and Talcott 1970 [S-T] no.) Ath. black-glazed cup Ath. black-glazed cup Ath. cup (S-T 576-7) Ath. cup type C (S-T 409-10)

Ionian cup Ionian cup Ionian cup Ionian cup Ath. cup-skyphos (S-T 578) Ath. black-glazed cup Ionian cup Ath. cup type C (S-T 398-413) Ath. black-glazed cup Ath. black-glazed cup Ath. cup

Date BC

Inscription

—] toi=j (Ell[— —]oi=j (Ellh/[— early 5th c. —] qe[oi] si to(i)j E ( [— 500-450 —]ac to[— —]i=j qeoi=[—

550-500

c. 480

500-450

Cambridge Fitzwilliam GR.105.1899 Oxford G 141.14 Cambridge Fitzwilliam GR.338.1899 BM GR 1911.6-6.34 Cambridge Fitzwilliam GR.104.1899 Oxford G 141.31 Oxford G 141.1

Ath. black-glazed cup Ath. black-glazed stemless cup (S-T 446) c. 480 Ath. black-glazed cup Ath. cup Ath. black-glazed cup

A 106/C 629 A 77/C 600 A 78/C 601 A 96/C 619

BM GR 1900.2-14.8 Oxford G 141.51 Cambridge Fitzwilliam Gr.337.1899 BM GR 1911.6-6.14

Ath. cup Bloesch 1940, pl. 11.2-3 Ath. black-glazed cup Ath. black-glazed cup Ath. cup-skyphos(S-T 565)

B 2/C 645

Oxford G 141.36

c. 500

c. 520 or 490-80

—] qeoi=si [— —]eoi=si[— —] qe[— —]eoi=s[— —]oi=si [— —]oi=si t[— —] (El(l)h/[— — (E]ll[h/nwn —] (Ell[h/nwn — (Ellh/]nwn Aqh— (E]llh/nwn — (E]llh/nwn — (Ellh/]nw[n — (Ell]h/nw[n — (El]lh/nwn — (Ellh/n]wn toi=j q[e]oi=si [toi=j (Ellh/] I nwn HN[- I me a)[ne/qhken — (El]lh/nwn : q[e]oi=si tw=n q]ew=n tw=n [ 9Ellh/nwn tw=n q]ew=n [tw=n (Ellh/nwn —ac qeo[i=si — —]qe(ke) qeoi=[si —

Note: Only those sherds in the British Museum, London, were inspected for the data in the above table.

Naukratis: Greek Diversity in Egypt | 19

Höckmann and Möller Illustration credits

Fig. 1 after Kerschner 2001, 73 fig. 1; Fig. 2 after Möller 2000a fig. 5; Fig. 3 photo Fitzwilliam Museum, Cambridge; Figs 4, 15, 17 photos the British Museum; Fig. 5 after Hogarth 1898/9, 32, photo the British Museum; Figs 6–9, 11, 14, 16, 18 photos the British Museum, drawings O. Höckmann; Fig. 10 photo the British Museum; Fig. 12 photo Ashmolean Museum, Oxford; Fig. 13 photo the British Museum; Bonn, Akademisches Kunstmuseum inv.no. 697.90, photo U.Höckmann.

Notes *

1

2 3 4 5 6 7 8 9 10 11 12 13

14 15 16 17

18 19 20 21 22 23

24 25 26 27

We would both like to thank the organisers for inviting us to this inspiring meeting of the Naukratis community. D. Williams, A. Villing, and A.W. Johnston helped with invaluable advice in questions of pottery and inscriptions. E. Böhr, U. Schlotzhauer and S. Weber contributed to the dating of pottery and B.B. Shefton to the subject of Castulo cups. U.H. wishes to extend her special thanks to O. Höckmann who drew the profiles and to A.M. who helped to translate her manuscript which was also read by S. Weber. A.M. owes special thanks to M. Nafissi and A. Arenz for discussion of the dedicatory inscriptions. We like to thank the following institutions for providing photographs: Akademisches Kunstmuseum Bonn (W. Geominy), Fitzwilliam Museum Cambridge (L. Burn), British Museum (A. Villing), Ashmolean Museum Oxford (M. Vickers). The map was drawn after Kerschner 2001, 73 fig. 1. Herodotus does not distinguish between Ialysos, Kameiros, and Lindos, the three Rhodian poleis before the synoikism. Apparently, Rhodes acted as one polis in external matters already before the synoikism of 408/7 bc (Diod. 13.75.1; Strab. 14.2.10); cf. Bresson 2000, 37-40; Nielsen and Gabrielsen 2004, 1196-7, for more evidence. Hdt. 2.178; cf. Möller 2000a, 23, 192-6. Hogarth 1898/9. Bowden 1996, 24-5. Cf. Möller 2000a, 105-8. Hogarth 1898/9, 31, 55 no. 51; cf. n. 17. Cf. infra ns 129, 132. Hogarth 1898/9, 33, 65-7 pl. 9; for the date see Höckmann and Koenigs (forthcoming). Cf. infra n. 30. 2 Oxford G 138.3+5+11; ARV 326.93. Onesimos has been dated by Williams 1993b, 15 to between 505 and 485 bc at the latest. Hogarth 1905, 109. Lloyd 1993/4, 61-8. To connect Herodotus and Naukratis in this way cannot be taken as an invitation to date Naukratis on the basis of Herodotus’ travels or his chronology. The evidence regarding Herodotus is even shakier than that for Naukratis. Coulsen, Leonard and Wilkie 1982, 79-80. Cambridge, Fitzwilliam Museum GR.31.1899; Hogarth 1898/9, 62 pl. 8.6; Lorber 1979, 50 no. 60 pl. 14; Amyx 1988, 508 n. 272; Möller 2000a, 220 no. 54. Hogarth 1898/9, 62. Fragment of a large vessel found in chamber 3, its whereabouts is unknown: Hogarth 1898/9, 31, 55 no. 51 pl. 4, who describes the letters of the dedicatory inscription as running ‘along the back and down the tail of a bull’; Bernand 1970.2, 700 no. 574; Möller 2000a, 106 n. 122; 168 no. 2c; 244 no. 21. The dating was suggested by U. Schlotzhauer; cf. supra n. 6 and infra n. 53. Hogarth 1905, 114 pl. 5.1: found ‘in a small patch of undisturbed deposit, just west of 66, [...] at a height of 10 inches above the basal mud’. Price 1931, 82 pl. 396.28 a.b. Lemos 1991, 118-22 fig. 62, 185, 285 no. 810, pl. 113; cf. also WalterKarydi 1973, 69, 140 no. 781 pl. 95.781; Boardman 1967, 169 n. 9. On the Dioskuroi in general cf. Koehne 1998. Gutch 1898/9, 67-97 pl. 10.13; Hogarth 1905, 115, 131-2. Oxford, Ashmolean Museum G 141.13; Bernand 1970.2-3, 707 no. 659 pl. 27.1 (first row, left sherd); Möller 2000a, 170. It is perhaps not from a Castulo cup, as I first supposed, see Shefton 1996, 178. I thank B.B. Shefton for his kind information. They come to an end by the mid-6th century bc, Jenkins 2001, 165; Nick 2001a, 210-1; Höckmann and Koenigs (forthcoming). Hogarth 1898/9, 38; Bernand 1970.2, 745-6 no. 7 (Classical), found near spot 3. Hogarth 1898/9, 32 pl. 14. 9; infra n. 69. London, BM GR 1900.2-14.17: Hogarth 1898/9, 33 (found in the

20 | Naukratis: Greek Diversity in Egypt

28 29 30 31

32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

55

56 57 58 59 60 61 62 63

surface debris beyond spot 54), 56 no. 144, pl. 5; Masson 1971, 33 (three syllabic signs); Johnston 1978, no. 17; Masson 1983, 354 no. 370: ka-wa-?-[-; Möller 2000a, 162 n. 593, 238 no. 2. The bottom fragment presumably belongs to a shallow (?) bowl or plate; on the inside impressed linked palmettes surround the lost middle ornament, on the outside the inscription is incised in the dark band; cf. Sparkes and Talcott 1970, 107-8, 128-47 pl. 36.1052; pp. 22-32 pls 53.560, 59.826; cf. bowls from tombs in Marion, Masson 1983, 354 n. 3. – Oxford, Ashmolean Museum G 141.29: Hogarth 1905, 117 no. 38: mo-ta-to-? Find-spot not indicated, but presumably found in the same region as the above- mentioned fragment London, BM GR 1900.2-14.17 or the Herakles base; Bernand 1970.2-3, 709 no. 681, pl. 27.1 (third row, first sherd left); Masson 1971, 33 no. 370a (four syllabic signs); Möller 2000a, 162 n. 593, 239 no. 11. Gutch 1898/9, 85-97 pls 12-3. Katja Weitz (Mainz) identified a find of some 70 Egyptian bronze statuettes. They came to light in a house in the south of Naukratis, cf. Petrie 1886b, 41-2. Most of them seem to be Hellenistic. Bernand 1970.2, 759 no. 26; 764-5 nos 36, 37, 38, 39, 40, 41; 767-8 nos 45, 48. Since it was impossible to inspect the inscriptions on pieces at Oxford which are currently packed into boxes, new observations might change the picture. I wish to thank U.H. for checking the pieces in the BM. Edgar 1898/9, 53-7. Möller 2000a, 166-81. Hogarth 1898/9, 44. Bowden 1996, 23, does not know of evidence from other sanctuaries. Cf. Appendix nos 1-22, Fig. 7. Cf. Appendix no. 23, Fig. 7. Cf. Appendix nos 24-25. Cf. Appendix nos 1-3 and theoretically nos 4-14, too; cf. Fig. 7. Cf. Appendix nos 15-23, Fig. 7. Cf. Appendix nos 24-5. Cf. Bowden 1996, 23. Schwyzer 1959, 2: 24. LSJ s.v. (Ellh/nioj refers to the Hellenion with temples for the qeoi\ (Ellh/nioi. Bowden 1996, 23-4. Bowden 1996, 23. Bowden 1996, 24. Cf. Chaniotis 1997. Schlotzhauer 2006. Nick (forthcoming); Ehrhardt, Höckmann and Schlotzhauer (forthcoming). For cult filiations see Nilsson 1967, 712-3; Fleischer 1973, 132-7; Nick 2002, 278 s.v. Filialheiligtum. Archaic dedications to the Dioskuroi from the Hellenion: Hogarth 1898/9, 30; Bernand 1970.2, 697-700 nos 546, 557, 558, 560, 567, 571, 580 (lamp), all found in spot 34; cf. Jenkins 2000 passim. Hogarth 1898/9, 30 no. 52 found near spot 34; supra n. 17, found in chamber 3. At Naukratis Apollo Milesios and Apollo Didymeus are attested on fragments from the temenos of Apollo, and Apollo Komaios and Apollo Pythios in the context of a so-called prytaneion mentioned by Athenaios 4.149D-E, cf. Nick (forthcoming); Ehrhardt, Höckmann and Schlotzhauer (forthcoming); Herda (forthcoming b). Hogarth 1898/9, 30 nos 65, 85 = Bernand 1970.2-3, 701 no. 588 pl. 26.2 (second row, second sherd left); 703 no. 608; Hogarth 1905, 117 no. 8 = Bernand 1970.2-3, 707 no. 651 pl. 25.1 (first row, second sherd left). Hogarth 1898/9, 62; Bernand 1970.2-3, 701 no. 585 pl. 26.1 (third row, first sherd left). London BM GR 1900.2-14.5; Hogarth 1898/9, 55 no. 54; Bernand 1970.2, 700 no. 577; high foot of an Archaic Ionian vessel, cf. Villing 1999, 191 fig. 1 a-c; cf. infra n. 81. London BM 1900.2-14.6; Hogarth 1898/9, 30, 56 no. 107; Bernand 1970.2, 704 no. 630; Scholtz 2002/3, 232; cf. infra chap. 5 with figs 145, n. 82. Cf. Hogarth 1898/9, 30 pl. 2. London BM GR 1888.6-1.211 and 1888.6-1.212; cf. infra chap. 5 ns 84-5 with Figs 15–19. Cf. Scholtz 2002/3 passim; Koehne 1998, 43 n. 130, cf. pp. 189-95. Scholtz 2002/3, 240. Cf. infra chap. 5. For dedications by hetairai, cf. Williams 1983a, 185 ns 57-9; Steinhart 2003, 220.

The Hellenion at Naukratis: Questions and Observations 64 Möller 2000a, 99-101. 65 Gruppe 1918, 987-8; Laurens 1986, 147-152; Schlotzhauer and Weber 2005, 74-8. 66 Hemelrijk 1984, 50-4, 173-4, 178-9 pls 118-23, figs 39, 41. 67 Herakles treads on the beaten foes like a pharaoh, cf. Wolf 1957, 575 fig. 575 (Abu Simbel); Herakles like a Pharaoh in a chariot, Littauer 1968, 150-2 pl. 62; cf. Shaw and Nicholson 1995, 51 fig. (Beit el-Wali). 68 Cf. the following notes. The inscription of London BM GR 1911.6-6.15 does not relate to Herakles; Hogarth 1898/9, 53 pl. 4 no. 3; Bernand 1970.2, 695 no. 525. 69 London BM 1900.2-14.16; cup type C, concave lip, cf. Sparkes and Talcott 1970, 91-2 pl. 19 no. 409 (500–480 bc), no. 410 (480 bc), or no. 413 (480–450 bc). The inscription reads: )Arte/mwn (Hrak(l)e[i=]; Hogarth 1898/9, 56 pl. 5 no. 84; Lazzarini 1976, no. 390; Bernand 1970.2, 702 no. 607. My thanks to Dyfri Williams who suggested a date in the first half of the 5th century bc. 70 Oxford, Ashmolean Mus. G 141 (1899) 58; the inscription reads: (Hrakle/[oj]; Hogarth 1898/9, 54 pl. 4 no. 33; Price 1931, pl. 1.28; Lazzarini 1976, no. 529b; Bernand 1970.2, 698 no. 555; Möller 2000a, 258 no. 16. My cordial thanks to Udo Schlotzhauer for his information on the date. For more Archaic fragments of vessels from Greece with dedicatory inscriptions to Herakles see Lazzarini 1976, nos 458, 461, 818; nos 526 und 442 are of the 5th century bc. 71 London BM GR 1886.4-1.5; Bernand 1970.2, 743 no. 2; the fragment of a limestone inscription London BM GR 1886.4-1.1364 (Inscription 1092) (Marshall 1916, no. 1092 = Bernand 1970.2, 757 no. 23) may also come from the Hellenion. It was found during the first campaign, Petrie 1886b, 63 no. 7, pl. 31.7. 72 London BM GR 1900.2-14.22; 17 x 10cm; Hogarth 1898/9, 32, pl. 14.9; Prinz 1908, 108. 118; Marshall 1916, 210 no. 1081; Gjerstad 1948, 318; Lippold 1950, 67 n. 15; Lippold 1956; Schmidt 1968, 115; Bernand 1970.2, 746-7 no. 9, pl. 34.2; Davis 1979, 13 n. 7; Floren 1987, 414 n. 3; Kyrieleis 1996, 74 (signed by a Cypriot at Naukratis); Donderer 1996, 87 n. 2; Vollkommer 2004. 73 Rolley 1984, no. 277. 74 Kansteiner 2000, 46-8; cf. also Palagia 1988, 745-6 for the Albertini type. 75 Hogarth 1898/9, 32; the inscription reads: Si/kw[n e)p]oi/hse Ku/p[rio]j I )Aristi/[wn] (Hraklei=; Möller 2000a, 162. The names of Sikon and Aristion are frequently attested (LGPN s.v.). In Naukratis we should assume a connection with Cypriot mercenaries during that time; see infra n.77. 76 Hermary 1989, 299-304; Yon 1992, 156-9; cf. Palagia 1988, 757 no. 566; Karageorghis 1998, 165-9 figs 113-5 (Hellenistic or Roman type Herakles statuettes from Cyprus). 77 The historical background could be provided by the alliance between Euagoras I of Salamis (c. 435–374/3 bc) and the Pharaoh Akoris of the 29th dynasty (393/2–380 bc), see Gjerstad 1948, 501-7; Masson 1983, 356-88; Jansen-Winkeln 1996, 406; Högemann 1998, 201-2. Cf. also the Attic black-glazed cups with Cypro-syllabic inscriptions, supra n. 27; for the presence of Cypriot mercenaries in Abydos and Karnak see Masson l.c. 78 Hogarth 1905, 115, 131-2; cf. the sanctuary of Herakles at the mouth of the Canopic branch, Hdt. 2.113 and Strab. 17.1.18; Gruppe 1918, 987 lines 47-9; preliminary results of underwater excavations in Heraklion see Herold 2002, 22-44; Goddio and Clauss 2006. The Archaic sanctuary of Herakles in Miletos: Rehm 1914, 276-7 no. 132. Representations of Herakles in Egypt in Hellenistic and Roman times are frequent; cf. Cassimatis 1978; Quaegebeur 1987; Palagia 1988; Vorster 1988; Clerc 1994; Felber 2003. For veneration of warlike Herakles cf. Archilochos fr. 324 West (spurious frgt.) (hymn to Herakles Kallinikos); Greek Iambic Poetry, ed. and translated by D.E. Gerber, Cambridge/MA: Harvard UP (Loeb) 1999, 282-7 no. 324; Farnell 1921, 146-8; Graf 1985, 99, 174 n. 103, 181, 296; Graf 1998, 391; for representations of Herakles’ labours on Attic black-figure and red-figure vases from Naukratis cf. Williams 1986, 63-4 (Nereus); Lamb 1936, pl. 20.5 (Triton); Venit 1989, 99-113 (Hydra); Beazley and 2 Payne 1929, no. 23 (tripod), no. 46 (Nessos); ARV 429, 20 (tripod); Piekarski 2001b, 34-5 no. B3 (Middle Corinthian, Hydra); Archaic jaspis from Tell Defenneh and Herakles head vase from Naukratis cf. Boardman 1988, no. 4.11. 79 Graf 1985, 296-316; cf. p. 99 (Chios); for Herakles und his cult on Thasos in Archaic time, see Bergquist 1973. 80 Cf. Graf 1985, 260-1. 81 The inscription on a rather high foot of a late Archaic Ionian vessel from the Hellenion (London BM GR 1900.2-14.5; Bernand 1970.2,

700 no. 577 reads 0Afro]di/thi PU[..., and can therefore not be the same epiclesis; cf. supra n. 57, Figs 8-9. 82 London BM GR 1900.2-14.6; Hogarth 1898/9, 30, 56 no. 107, find spot supra n. 59; rim fragment of red-figured volute-krater, right-to-left meander, in band under rim part of kottabos scene; on upper side of rim incised dedicatory inscription to Aphrodite Pandemos. Fragment of the same side of the krater in Bonn, Akademisches Kunstmuseum Inv. no. 697.90, infra n. 80; first quarter of 5th century bc; cf. Schleiffenbaum 1991, 60-3, V 212; ARV² 228.20 (490 bc); later: V 292, ARV² 287.27 (460 bc, Geras-Painter); Bernand 1970.2, 704 no. 630; Scholtz 2002/3, 232. 83 Bonn, Akademisches Kunstmuseum Inv. no. 697.90; Piekarski 2001b, 40, no. C 13 pl. 14.1. We thank D. Williams for this information. The name of the dedicant could be Psammis, Psammatas, Psammetichos or the like, cf. LGPN s.v.; Jeffery 1990 s.v. 84 London BM GR 1888.6-1.211; Gardner 1888, 66 no. 818; Bernand 1970.2, 688 no. 467; Scholtz 2002/3, 231; cf. for the type Sparkes and Talcott 1970, 92 no. 420 fig. 4 pl. 20. 85 London BM GR 1888.6-1.212; Gardner 1888, 66 no. 821; Bernand 1970.2, 689 no. 470; Scholtz 2002/3, 231; cf. for the type Sparkes and Talcott 1970, 91-2 no. 413 fig. 4 (480–450 bc) or no. 407 pl. 19 (500–480 bc). 86 There is also an interpretation of Aphrodite Pandemos as goddess of light (Usener, Furtwängler), but this does not need to interest us here. 87 Scholtz 2002/3. 88 Cf. Graf 1985, 260-1; Pirenne-Delforge 1994. 89 Scholtz 2002/3, 236. 90 Burkert 1985, 155, combines the distinction between Ourania, who is made the Phoenician queen of heaven, and Pandemos, but gives the latter a political meaning: she literally embraces the entire people as the common bond and fellow feeling necessary for the existence of any state. 91 Xen. Symp. 8.9-10. 92 Paus. 1.22.3. 93 Ath. 13.569D quoting Philemon F3 PCG; Nikandros FGrHist 271-2 F9a; cf. Nikandros FGrHist 271-2 F9b and Apollodoros FGrHist 244 F113 in Harpocration s.v. Pa/ndhmoj 0Afrodi/th. 94 Kruse 1949, 509; cf. Dillon 1999, 68-70; van Bremen 2003, 325-6, who emphasizes Pandemos as being responsible for the collective wellbeing of the civic community. 3 2 95 Paus. 1.22.3; IG I 832 (=IG I 700) of ?480–470 bc (no epiclesis); Hurwit 1999, 41, 212, 276-7; cf. Jacoby 1944, 72-3 = 1956, 254-6. 96 Sokolowski, LSCG 39; cf. Simon 1970. 97 Paus. 1.22.3. 98 Erythrai: SEG XXXVI (1986) 1039 of around 400 bc, decree about building of a temple for Aphrodite Pandemos; Inschriften griechischer Städte aus Kleinasien (Erythrai & Klazomenai II) 201 a 24 of 300–260 bc, selling of priesthoods (cf. Parker and Obbink 2000, 415-49). Sacrificial calendar from Isthmos on Cos: LSCG 169 A 12 (3rd century bc). 99 Scholtz 2002/3, 238. 100 Pind. Ol. 6.63. 101 Scholtz 2002/3, 242. 102 Bowden 1996, 33, cf. Tausend 1992. 103 Thus, A. B. Lloyd in his excellent commentary on Herodotus Book II (vol. 3 [1988], 224) states: ‘We should expect the Hellenium to be a pan-Hellenic religious enclosure and that is exactly what the evidence suggests this area was.’ 104 Morgan 1993. 105 Malkin 2003b, 91-5. 106 Demetriou 2004, a talk which I only know from the abstract. She has informed me that this research is part of her PhD thesis. 107 Hall 2002, 205-20. 108 Hall 2002, 131-2. 109 The only Panhellenion we know of was founded by the emperor Hadrian in 131/2 ad. The idea sprang from an idealized picture of good old Classical Greece inspired by the Second Sophistic. 110 Kron 1988, 135-48; Kreuzer 1998, 32-41, both with earlier literature; Gebauer 2002, 448-70. 111 Villing, this volume. None are, however, recorded as having been found in the Hellenion. 112 Williams 1983a, 186-7; Kreuzer 1998, 32-41; Stissi 2003, 77-9. 113 Hogarth 1905, 112. 114 Leclère 1997 assumes chapels or treasuries of the individual cities and their deities. He kindly provided his manuscript to members of

Naukratis: Greek Diversity in Egypt | 21

Höckmann and Möller the Naukratis research project at Mainz. 115 Von Bissing 1951, 79; Kron (1992, 620-2 n. 58) takes similar chambers at Bitalemi for treasuries or cultic rooms – amphipoleia – i.e. chambers of a priest. 116 Martin 1951, 245-6. 117 Martin 1956, 44-5; for a summary cf. Bernand 1970.2, 858, and Leclère 1997. 118 Williams 1983a, 186-7 n. 66 with earlier literature. 119 Bookidis and Stroud 1997, 393-412. 120 Hogarth 1898/9, 31: in chamber 3 ‘layer of hammered mud’ under the ‘level of the fragment of concrete pavement’, south of it p. 33: ‘a floor of hammered earth overlaid with fine plaster, part plain crimson, part crimson and white stripe, and part blue and white stripe. Only small fragments of the coloured surface were preserved’ in room 20; ‘a similar floor in a similar state, but this time overlaid with yellow and red stripe plaster’ in room 19, where the hoplite relief was found in the ‘the large vessel […] below the level of the plaster floor’. 121 Hogarth 1898/9, 30: ‘concrete pavement’ near 39 in the south; p. 31 ‘concrete paving’ near 2; p. 33 are mentioned remains of ‘brilliant blue stucco’, the pavement here consisted of a ‘thin layer of concrete’ 3.7cm thick, in room 11; ibid. a ‘flooring laid on fragments of coarse plaster’ north of 14 is mentioned, and (p. 34) in spot 30 a ‘thick stratum of concrete laid on chips’, 17.5cm thick, ‘seems to have extended all over the northernmost part of the site’; Hogarth 1905, 115, saw in 58 a ‘patch of pavement of thin concrete’ […] 175cm under ‘the well marked floor level of the Ptolemaic restoration’; in room 10 ‘much fallen wall-plaster of brilliant blue’ on the ‘Ptolemaic floor […] made of a concrete of lime, pounded brick, and pebble’, c. 2cm thick. 122 Hogarth 1898/9, 33, saw in room 14 ‘remains of a conduit made of earthen pipes from 4½ to 6 in. (11.25cm to 15cm) in diameter’, over the terracottas. Cf. Bergquist 1990, 37: ‘common characteristics’ of dining-rooms ‘are a regular square shape, a specific internal walllength, and a frequent multiplication of such rooms in paratactic

22 | Naukratis: Greek Diversity in Egypt

123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142

rows, […] behind a common porch or portico’. The ‘fittings and furnishings’ are ‘a paved or cemented floor with a raised border along the walls and/or couches or supports for couches, an offcentre door, drainage, wall stucco, access to water, etc.’ Supra n. 120. Hogarth 1898/9, 38. Cf. Gazda 1983, 35 fig. 61; Schwartz and Wild 1950, pl. 1. I thank my colleagues from the Egyptological Institute at Mainz for much helpful information. Hogarth 1898/9, 30: ‘ashes mixed with late black glazed ware’ near 40-2, see plan pl. 2; p. 31: ‘patchy layer of ashes’ near spots 2. Hogarth 1898/9, 31: ‘layer of burnt stuff resting on the mud’ near westernmost spot 2. Hogarth 1898/9, pl. 2. Hogarth 1898/9, 30-1. As an example of an altar at which sacrifices to several gods are made, cf. Petrakos 1968, 96-8. Hogarth 1898/9, 32. Hogarth 1898/9, 30 pl. 2, near east-west wall, 39; p. 33; cf. Iozzo 1985, 7-61; Bookidis 1993, 52. Cf. Möller 2000a, 107. Schmitt Pantel 1992, 488-90. For further discussion see Villing, this volume. Jacoby 1912, 731. Tresp 1914, 159, identified this Hermeias with H. of Methymna who wrote a Sicilian history in the 4th century bc. Jacoby remains sceptical. FHG 2: 80 = fr. 112 Tresp ap. Ath. 4.149D. Graf 1985, 187-8. On Apollo Komaios see also Herda (forthcoming b). Möller 2000a, 195-6; Möller 2005, 189. In the 3rd century bc the names of two out of four phylai with 10 bouleutai each are known; cf. the new edition of the inscription by Scholl 1997, 213-28, pl. 18. Ath. 4.149F; cf. Gottlieb 1967, 28-30, who, however, assumed that what Athenaios described goes back to the 6th century bc. Cf. Boardman 1994, 142; Lomas 2004b, 2.

The Delta: From Gamma to Zeta Alan Johnston

Abstract My aim of this short paper is to give a resumé of the range of epigraphic material from Naukratis of the Archaic period, mostly on East Greek pottery, and to focus briefly on just two aspects: nonIonians, and amphora ‘texts’. They comprise a few thoughts ahead of a comprehensive re-study of the material as both texts and pots.* Smaller inscribed vases The ceramic material from the excavations at both Naukratis and Tell Defenneh was published with exemplary speed after each season of excavation. The texts comprise almost entirely either dedications, to a range of deities, or, in the broadest sense, ‘trademarks’, while what may well be owners’ marks are represented by a rather few short texts or ligatures, increasing in number in the classical period, though many of the latter are now seemingly unlocatable. The format of the dedications is varied, and few patterns seem to emerge; but review of the material will yield more specific data in this respect.1 When I first worked through some of the material in the British Museum’s ‘fan room’ in the 1960s, I noted very few corrigenda to be made in the readings in the relevant pages of the publications, save some misleading treatment at the breaks. There were and still are, however, unpublished pieces in the British Museum and probably elsewhere; André Bernand’s massive compilation of 1970 is of use in giving clearer listings of previously published material, but it adds little, since very few comments are added, and is of minimal use to the archaeologically minded. Material not then published include such things as the 25 ‘trademarks’ on East Greek pots which I put in my ‘Trademarks on Greek Vases’, and one particular piece in University College London (Fig. 1), which I noted in BICS 1982, arguing it to be a dedication of Aristophanes (Aristophantos as Dyfri Williams was soon to make him) and Damonidas to Aphrodite (cf. Fig. 7); the pair were active also on Aigina, as demonstrated fully by Williams when publishing the new Chian material from Aigina which so complicates the story of the dedication of bespoke Chian kantharoi: found on the island of Chios, but much more frequently at Naukratis and Aigina – and not at all, we may note, at Gravisca or the Heraion on Samos.2

Figure 1 UCL-742. Chian kantharos rim and handle

The range of ethnics used and a few peculiar alphabetic uses have long been known and discussed by John Boardman, Anne Jeffery, Rudi Wachter, Mario Torelli and Astrid Möller among others.3 On the matter of Aeolic bucchero being marked by Mytilenaians I merely add here some undiagnostic scraps in a long series of some 500 inscribed sherds catalogued in the register of the British Museum’s Greek and Roman Department by Donald Bailey under BM GR 1965.9-30, where we might note two or three alphas with cross-bar ascending right. Nonetheless, the fact that two of these, the first to be looked at in my review, join is either a statistical freak or a sign of substantial surprises in store.4 Teos, Miletos and Knidos are also well-known cases – (Fig. 2) the last reminding us perhaps of the tombstone of a

Figure 2 a) East Greek cup BM GR 1886.4-1.681-2 (N237) b) Stele from Marion. Cyprus Museum

Naukratis: Greek Diversity in Egypt | 23

Johnston a

b

c

Figure 3 BM GR 1888-6-1.173 and UCL-736. Chian lid or krater

Figure 4 ‘Lunate’ gamma in graffiti on Chian chalices: a) N750, b) BM GR 1888.41.420, c) BM GR 1965.9-30.141

a

a

b

b

c

c

Figure 6 ‘Mixed’ gamma: a) BM GR 1886.4-1.650, Laconian? b) N762, Chian? c) BSA 22, Samian?

d

Figure 5 ‘Ionic’ gamma: a) BM GR 1886.4-1.813, Milesian?; b) BM GR 1888.6-1.169, Chian; c) N815,Attic?; d) Cairo 26152N876, Milesian?

a

Figure 7 Dedications by Aristophantos and Damonidas from the temple of Aphaia on Aigina

Figure 9 Chian chalice BM GR 1888.6-1.495

24 | Naukratis: Greek Diversity in Egypt

b

Figure 8 a) Graffito on Corinthian louterion rim from Chios; b) Laconian bronze lion from Samos

Figure 10 East Greek cup BM GR 1886.4-1.260

Figure 11 Large shallow bowl BM GR 1886.4-1.671

The Delta: From Gamma to Zeta Knidian girl from Marion on Cyprus, and thereby the role of Cypriots in the trade with Egypt.5 There is little in the use of script that assists in placing the origin of the more general ‘Ionian’ texts, since the same lettering can be found on pieces of surely Chian or surely Samian origin. I wrote a note on one possibly diagnostic letter, sigma, in BICS 1974, pointing out its diminishing use in the earlier 6th century bc throughout Ionia, as well as demonstrating, I hope, the erroneous interpretation, and occasionally reading, placed on a large percentage of the relevant texts, where nu had become confused with sigma either by the original writer or the modern transcriber and interpreter. Astrid Möller, I add immediately, plausibly suggests that a few of the texts with the letter may be from Aiginetan dedicators.6 I turn to a few notes on individual pieces which I find of interest. First I must add to the note on the unusual gamma that I included in my offering to John Boardman in Periplous; I published one lunate example (Fig. 3) in the bespoke inscription, unusually on a large Chian vase, a lid or krater,7 but I failed to add further Chian examples – graffito on chalices (Fig. 4): N750 (BM GR 1888.6-1.421) and N752 (BM GR 1888.6-1.420), and a kantharos scrap (BM GR 1965.9-30.141), where it is found between an epsilon and perhaps a upsilon. These to add to the one other oft-cited Ionian occurrence from Samos, Jeffery 1990, 341, Samos 7. The letter gamma is not otherwise widely attested at Naukratis – the ‘normal’ form occurs in Aigyptis’ set of bespoke kantharoi8 and a half dozen other texts (Fig. 5): • BM GR 1886.4-1.813 (N309, bowl, Milesian?); • BM GR 1888.6-1.750 (N732, Chian pot); • where? (N815, Attic? This group of graffiti is of various fabrics, as far as can be ascertained. I assume the name to be Megakles); • Cairo 26152 (N876; Gardner 1888, pl. 20 – ?Milesian); • Oxford G141.22 (JHS 33, a later piece, perhaps 5th century bc Attic); • BM GR 1888.6-1.169 (Cook and Woodhead 1952, no. 54, Chian bespoke dipinto; the letter is incomplete, though plausibly forms part of an egrapsen signature); • perhaps also BM GR 1888.6-1.453 (see below); while a form with rising second stroke (Fig. 6) appears more intriguingly on three pieces: • BM GR 1886.4-1.650 (N340, an unusual cup, with many Laconian features, but more likely Knidian); • where? (N762, Chian?); • where? (BSA 22, Samian?). All three were dedicated by the same Hermagathinos who cut the more lunate letter on BM GR 1888.6-1.420 (N752) (Fig. 4b). While no clear picture emerges, and this is not the place fully to demonstrate that this seems a small patch in the protohistory of the letter, the four Chian examples of ‘non-Ionic’ gamma, which represent the ‘hand’ of at least three individuals, should make us pause in pursuing over-relentlessly strict allocations of scripts to poleis.9 Similar thoughts attend, of course, the use of the Doric dialect in the name of Damonidas (Fig. 7), attested in the painted inscriptions on Chian kantharoi from the Aphaia temple, and plausibly once painted on the Naukratis fragment, UCL-742 above (Fig. 1).10 To underline the more inter-polis aspects of our subject, I might mention, inter

alia, the Corinthian dedication on a Corinthian louterion from Chios (Fig. 8a), a Spartan’s dedication on a bronze from Samos (Fig. 8b), and a Corinthian or Megarian dedication on a Chian sherd from Eleusis, albeit, unlike in the case of our lunate Chians, none of these reflect non-epichoric forms used by local dedicators.11 BM GR 1888.6-1.495 (given as 496 in Cook and Woodhead 1952, no. 68, with incomplete reading) (Fig. 9) is a lower wall fragment of a fairly large Chian chalice, perhaps one of the ‘Grand’ pieces, with a bespoke dedication by a ...]midhj. What is intriguing is that the mu is actually incised at a point where the slip had peeled away; that incision is ancient. This suggests, though does not prove, that the pot had suffered in some previous history, but that ...mides or his agent was present at or near the final act of dedication at Naukratis to ‘restore’ the text. I know of no good parallel but regard the piece as an important element in our study of ritual offerings. Full re-study of the material in the British Museum will undoubtedly throw up items of interest. There follows here a few prelimary remarks on pieces worthy of further study: BM GR 1886.4-1.260 (N117, B79) (Fig. 10). ]a?mpuri.[. The reading in N is slightly defective: there is part of a horizontal to the left, almost certainly making the first preserved letter an alpha, and at the end merely a section of a line at the break, but so situated as to suggest it is from a circular letter. The piece is the lip of a low-lipped cup with a little mica. BM GR 1886.4-1.671 (N236, B197, Möller 2000, 174, no. 5b; sample Nauk 73) (Fig. 11). ]ekankley[. The piece is a large flatrimmed bowl, mostly covered with a thin brown ‘glaze’. Hypothetically it may be of Knidian manufacture.12 It is difficult not to construe this as an aorist third plural followed by an aorist participle, a0ne/q]ekan or, as suggested by Möller, ka/qq]ekan kle/y[antej; her Kley[i/aj can perhaps be discounted in view of lack of evidence for any such personal names. A compound adjective perhaps should not be ruled out; but the deus ex machina of a reference to a klepsydra runs against the shallowness of the bowl (even if the full depth is not preserved) and it would be a truly remarkable occurrence. The topos of theft and subsequent action appears in graffiti of the approximate period, though the plural does cause obvious difficulties.13 One final possibility is to read a verb with apocope, retaining perhaps the Aeolic nature of Möller’s suggestion, ka/qq]ek’ a0nkley[... : this would allow a singular dedicator, though introduces an unexpected elision, despite the elegance of the contrast of the two prepositions. While epigraphic probability points in this direction, the prominence of such a statement of theft in a dedication formula would be striking. BM GR 1888.6-1.359 (N754, B405). ]yende[ in the first line, ].oaneq[ below. Cut on the outside of the foot of this small Chian chalice. The first preserved letter is clearly a psi and there is the minimal trace of a vertical at the start of line 2, perhaps an iota. One wonders whether there was room on the small pedestal for a signature egrapsen as well as a dedication, seemingly including a patronymic, below. Figure 12 Chian chalice BM GR 1888.6-1.453

Naukratis: Greek Diversity in Egypt | 25

Johnston

Figure 13 Cup BM GR 1886.4-1.261

Figure 14 Cypriot amphora handle BM GR 1888.6-1.389

BM GR 1888.6-1.453. (N760, B411) (Fig. 12). ]n:o pe.[ | ] aunthj:[. Foot fragment of Chian chalice. Gardner suggested in line 2 Phaid]runtes, but this is epigraphically highly unlikely. He may well have had in mind Olympian polishers of statuary, since no other common or proper nouns in - aunthj or - runthj seem attested elsewhere, save the daughter of the Persian Masistes, Artaunte (Hdt. 9.108). In the first line the final preserved letter has a horizontal at the top – gamma or pi, and in the second the first can hardly be read as rho; at the end there is part of a two-dot punctuation, as after the first letter of line 1 – a0ne/qhke]n : o9 Pe.[ | ]aunthj :[. Regrettably there is not enough of the word commencing Pe- preserved to allow conjecture as to whether it was a patronymic, ethnic or something else; I merely note the lack of any Greek word in Peg- and the extreme rarity of names in Pep- (with short vowel).14 Should one look outside the Greek world but in Greek grammar? BM GR 1886.4-1.261 (N122) (Fig. 13). t]oi zeni t[ Rim of non-micaceous cup. Non-Ionic script, but Ionic dialect, not previously noted as such. The script could be Cycladic. Inscribed amphorae I mentioned Cyprus above, and the immediate epigraphic association that comes to mind is with the graffiti on Cypriot amphora handles from Naukratis, in alphabetic, not syllabic script (Table 1, 33-41), as indeed is also the case with the Cypriot style limestone figurines of more disputed origin.15 I have nothing to add to my 1982 note on the handles, save (Fig. 14) perhaps stressing the possible Knidian origin of the inscriber of 40, if we take the simplest alphabetic explanation of the hourglass sign here as Knidian xi.16 Turning to other inscribed jars, I include in my discussion the more intact pieces from Defenneh and Qurneh; I do not overlook the as yet not fully published material from Eliezer Oren’s excavations at Migdol/ ‘Elephantine’, and the finds from Abusir

Figure 16 Klazomenian amphora from Tell Defenneh, BM EA 22343

26 | Naukratis: Greek Diversity in Egypt

Figure 15 Attic ‘à la brosse’ amphora from Tell Defenneh, BM GR 1888.2-8.60

published by Kvevta Smoláriková, but only two inscriptions are specifically mentioned.17 A number of handles and other fragments were retained from the Naukratis excavations, and, to my knowledge, no uninscribed pieces of plain containers; I cannot as yet allot more than a few to a given provenance, and indeed some may not be from amphorae at all.18 The range of material is scarcely atypical of that found elsewhere in the Mediterranean. I note merely the small number of Corinthian A jars, compared with the record in the central Mediterranean, while the Attic SOS too is thinly represented, despite being a vigorous export until c. 575 bc (though before the Greek arrival at Defenneh, from where we have just one example of the later Athenian ‘à la brosse’ type (1) (Fig. 15), which curiously is, to date, our only assured ‘merchant’s’ mark appearing on amphorae from two widely separated sites.19 Regrettably the set of short dipinti on bowls from Defenneh published by Petrie do not seem to be preserved; it would have been interesting to compare them in all respects, including fabric, with similar dipinto marks on amphorae from Klazomenai; such an apparent system of short painted marks does not appear with regularity in other, probably contemporary, areas of production.20 A Klazomenian jar in the Department of Ancient Egypt and Sudan in the British Museum has a mark of this genre (4) (Fig. 16), though it seems not to be any of the pieces noted by Petrie; my interest in it is its extraordinary repair – it became cracked but was not broken, was repaired with very large drill holes used, and survived in that condition until excavated; very similar is the Milesian piece from Migdol, Oren 1984, 20, fig. 23.5. One large dipinto from Qurneh (59) (Fig. 17) is very different, and recalls published marks from Black Sea sites, while the profile drawing might suggest North Greek production.21 The seal on Chian bobbin jars from Defenneh (2) (Fig. 18) has caused ink to flow, but Amasis’ cartouche on such a

Figure 17 Amphora from Qurneh

Figure 18 Chian amphora from Tell Defenneh, sealed and marked with cartouche of Amasis, BM EA 22356

The Delta: From Gamma to Zeta

Figure 19 Samian amphora from Qurneh, Petrie Museum, UCL, 16391

Figure 21 Fractional Samian amphora, BM GR 1886.4-1.1291

piece need not surprise us, since it was a long-lived amphora type.22 I throw in the observation that there is an as yet unpublished tomb from Rhodes town which had a Chian Late Archaic amphora in it, which I assume was not buried before c. 408 bc. Samian jars seem relatively common;23 this type too has a form of marking unto itself, the pre-firing graffito, which only occurs with any consistency elsewhere at the period on a few Corinthian A jars, employing a different set of signs, though it may well be that both encompass numerical notation. The Samian jar now in the Petrie Museum from Qurneh (60) (Fig. 19) was said by Petrie to have a post-firing mark, but on the original it is clearly not so, and a substantial number of parallels can be cited (Table 2).24 About 40 Samian jars are known to me with pre-firing graffiti (about half of Samian jars with any mark), while I know of about 50 other such marks on Archaic amphorae; 12 are on Corinthian A, the rest very disparate. The Samian marks, found on jars spread throughout the Mediterranean, are more coherent, and in fact a number are duplicated – X is frequent; epsilon, zeta, as on the Petrie jar – with apologies to scientists – lambda, tau and upsilon are repeated, and there are two examples of simple impressed small circle, one hour-glass and one figure-of-eight. Single examples are known to me of beta, ‘horizontal xi’, wavy line, triple vertical and triple horizontal. While it is difficult to see an overall pattern here, some sequences of letters may exist – epsilon, digamma, zeta; tau, upsilon. There are three examples of digamma, not a letter used in Ionian script, while a relatively early, fragmentary jar from Kommos (Fig. 20) has a clear mu digamma; a method of numeration is very clearly indicated, but one that can accommodate 46 and in some way privilege 6, 7

and 30, as well as probably non-numerical signs, is far from easy to discern. Note especially that any straightforward interpretation as capacity is ruled out by the fact that a ‘fractional’ jar in the British Museum from Naukratis (Fig. 21) has the same zeta as the large Petrie piece. One should add immediately that zeta iota (= 17), probably also pre-firing, is found on a lost companion piece to that in the Petrie Museum, Qurneh no. 850 (Fig. 22).25 One may consider the possibility that the marks in some way indicate batches in a kiln, though they presumably do not exist to facilitate the post-firing matching up of pot and lid, as in the later series of Attic pyxides.26

Figure 20 Samian amphora from Kommos, Crete, c. 625-600 BC

Figure 22 Amphorae from Qurneh

Addendum ex-Africa SOS My concern above was for amphorae with inscriptions, but it may be worth adding these few notes on some uninscribed pieces, all of which contribute in some way to the dating of Greek presence at African sites. Rather to the west of our area of prime concern is Cyrenaica, but the dating of an SOS amphora to before 650 bc from the area does reflect on general Greek interests. A piece found in the sea off Ptolemais/Apollonia was so dated by Emmanuela Fabbricotti in 1980. The pot does seem rather to belong roughly to the middle-late period, with a somewhat flat shoulder and roughly painted decoration. At any rate there is no compelling evidence for it to be placed earlier than ‘c. 630 bc’.27 Coming east, in the recent publication of finds from the University of Philadelphia’s excavations on Bates’s Island, Marsa Matruh, one pot of Greek origin is taken to be the earliest Iron Age import on the site, at c. 700 bc, by Don Bailey (Fig. 23). As such it would indeed be a striking fragment, since no other such import into Egypt can be dated to before c. 640 bc.28

Naukratis: Greek Diversity in Egypt | 27

Johnston

Figure 23 SOS amphora neck from Marsa Matruh

Figure 24. SOS amphora neck from Karnak, c .650 BC

However, the piece is most unlikely to be from an Attic SOS amphora, as suggested by Bailey. While its precise typological niche cannot be argued in view of the lack of very close parallels, its characteristics belong generically to what may be termed those of the ‘Laconian’ amphora, and should probably be dated to the period 625–575 bc. Against an attribution to the Attic ‘SOS’ amphora type are the profile with angular, jutting lip, the decoration, with a painted interior to the neck and paint not extending lower than the point of the ridge below the lip, and the clay, described as “hard red clay with white grits”, not typical of Attica.29 Similar material comes from Kommos in south central Crete, an emporium site in the 7th century bc. Such amphorae are, however, not fully comparable with the MM sherd; one of the Kommos variants has a similar lip, but known examples are reserved inside, while the other is painted inside the neck but has a ‘fascia’, not a ridge, below the lip.30 This material has been provisionally assigned to Laconia on adequate geological and rather broad typological grounds; the range of more persuasively Laconian jars, largely of the 6th century bc, again

provides no totally similar parallel.31 However, there can be no doubt that the piece is not Attic of around 700 bc, but of a type current in c. 600 bc, emanating on current evidence from either Laconia or a site yet to be identified. In addition to the piece of ‘à la brosse’ type noted above, a second Attic amphora from the Delta is worth noting, and I thank Helen Jacquet for bringing it to my attention (Fig. 24).32 It is part of an SOS neck from Karnak, from a stratum containing 26th–30th Dynasty pottery. The profile here does conform more closely to the ‘middle’ series, with a relatively low lip and sharp ridge below, as well as having a fairly straight neck, as far as can be judged. One would certainly be tempted to a date ‘c. 650 bc’ for its manufacture, though that of its arrival in Egypt could have been some time later. The question of the contents of the jars is regularly raised, and more recently John Lund has strongly supported a primary use for wine,33 stressing the appearance of Dionysos shouldering an SOS on the François vase as the most important piece of evidence. True, this scene is not to be underestimated, but one must also note that the SOS and the Corinthian A jar are the two earliest types of transport amphora to be made and circulated in large quantities, from the later 8th century bc onwards, and that it is impossible that the Corinthian product could have held wine because of its porosity. We must hope that a carefully conducted programme of organic analysis may some day help resolve this debate.

Table 1 Inscribed amphorae from Egypt34 From Tell Defenneh 1 BM GR 1888.2-8.60 ‘à la brosse’ Johnston 2000b, 236; Fig. 15 I–IET 2 BM EA 22356 Chian Petrie 1888 pl. 36.3; Fig. 18 pentalpha* 3 BM EA 22333 E. Greek? API, retrograde, shoulder 4 BM EA 22343 Klazomenian Fig. 16 glaze squiggle, by handle 5 where? Lesbian Petrie 1888 pl. 33.12b N, probably same as next entry 5 Petrie Museum 19247 Lesbian Petrie 1888 pl. 33.12b? two lines on shoulder 6 where? Lesbian Petrie 1888 pl. 33.12a D 7 where? N. Greek? Petrie 1888 pl. 33.1 E, shoulder 8 BM GR 1977.10-11.2 Samian HI, large, shoulder 9 BM GR unregistered Samian? Johnston 1887, 129, fig. 1 ‘anchor’, large, shoulder 10-12 where? (3 pieces) Samian? Petrie 1888 pl. 33.10b-d various letters part-preserved 13 where? Samian? Petrie 1888 pl. 33.10a ‘arrow’, twice 14 where? Samian? Petrie 1888 pl. 34.39 ‘arrow’, shoulder *The neck is closed with plaster stamped with a seal of Amasis; Petrie (1888, 64) mentions more than one Chian jar so sealed. From Naukratis 15 BM GR 1910-2-22.19 16 BM GR 1910-2-22.27 17 BM GR 1910-2-22.26 18 BM GR 1910-2-22.25 19 BM GR 1910-2-22.24 20 BM GR 1910-2-22.23 21 where? 22 BM GR 1910.2-22.10 23 BM GR 1910.2-22.11 24 BM GR 1910.2-22.12 25 BM GR 1910.2-22.13 26 BM GR 1910-2-22.28 27 BM GR 1910-2-22.29 28 BM GR 1910-2-22.21 29 BM GR 1910-2-22.22 30 BM GR 1910.2-22.35 31 where? 32 BM GR 1910.2-22.1 33 BM GR 1910.2-22.2 34 BM GR 1910.2-22.3

Chian Chian Chian Chian Chian Chian Chian Chian, not amphora? Chian, not amphora? Chian, not amphora? Chian, not amphora? Chian? Chian? Chian? Chian? Corinthian? Cypriot Cypriot Cypriot Cypriot

28 | Naukratis: Greek Diversity in Egypt

N403 N411 N410 N409 N408 N407 N I pl. 16.4 N394 N397 N397a N397b N412 N413 N405 N406 N420 N385 N384, Johnston 1982, 35-7 N386, Johnston 1982, 35-7 N387, Johnston 1982, 35

large O I–I? part of ligature?, shoulder M, shoulder alpha-lambda?, shoulder by handle ]IIL neck large white ‘arrow’, shoulder ‘psi’ handle X, handle lambda-alpha, handle hour-glass, handle mark as Johnston 1979, type 12E, neck NILE, shoulder, handle to right X, part preserved, wall ]IIII, shoulder lambda-epsilon, neck I–I above H, handle ENLH and strokes, handle I–IY, handle deeply cut X, handle

The Delta: From Gamma to Zeta 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56

BM GR 1910.2-22.4 BM GR 1910.2-22.5 BM GR 1910.2-22.6 BM GR 1910.2-22.7 BM GR 1910-2-22.20 BM GR 1888.6-1.389 BM GR 1910.2-22.38 BM GR 1910.2-22.36 BM GR 1910.2-22.242 BM GR 1910.2-22.9 BM GR 1910.2-22.8 BM GR 1910.2-22.34 BM GR 1886.4-1.1261 BM GR 1910.2-22.31 BM GR 1910.2-22.30 BM GR 1910.2-22.14 BM GR 1910.2-22.39 where? BM GR 1910.2-22.37 where? BM GR 1910.2-22.32 BM GR 1910.2-22.33

From Qurneh 57 Qurneh 58 where? 59 where? 60 Petrie Museum 16391 61 where?

Cypriot Cypriot Cypriot Cypriot Cypriot Cypriot Cypriot? E.Greek E Greek Klazomenian N.Greek? N.Greek? Samian amphora? amphora, Samian local ? ? ? ? ? ?

N388, Johnston 1982, 35-7 N389, Johnston 1982, 35 N390, Johnston 1982, 36 N391, Johnston 1982, 36 N404 N764, Johnston 1982, 36-7 N423 N421

N415 N414 N398 N424 N396 N422 N395 N416 N417

heta-epsilon, handle xi, handle XIII, handle three lines crossed by diagonal more complex than sampi, pace N, wall ED and hourglass, handle LI, trace before, shoulder LA, shoulder ‘trident’, pre-firing, shoulder IIIII: , handle ]IXX, handle heta-upsilon zeta, pre-firing, shoulder; dark dipinto, L. alpha-upsilon?, shoulder ‘psi’, shoulder? S, handle phi, on top of mushroom lip; 4th century BC xi, handle ]N ‘psi’ handle mark perhaps as Johnston 1979, type 12E crossed theta, plus, shoulder

Klazomenian Lesbian? N. Greek? Samian Samian

Mys´liwiec 1987, no. 840 Petrie 1909, pl. 55.852 Petrie 1909, pl. 55.855 ; Fig. 17 Petrie 1909, pl. 54.849 Petrie 1909, pl. 54.850

pre-firing? retrograde nu, edge of foot ‘psi’, neck large red(?) AN, body pre-firing zeta; alpha-lambda pre-firing? zeta iota, shoulder

unpublished unpublished unpublished Cat VII 12 Cat VII 11 Cat VII 21 Cat VII 22 Cat VII 1 Johnston 2000c, no. 288 Johnston 2000c, no. 293 unpublished unpublished unpublished unpublished unpublished unpublished unpublished unpublished unpublished unpublished unpublished unpublished Johnston 2005, no. 220; Fig. 20 unpublished unpublished Cat VII 7 Fig. 21 unpublished Qurneh pl. LIV, 850; Fig. 22 Fig. 19 unpublished unpublished Brunnen W2, unpublished Technau 1929, 30, fig. 22 Technau 1929, 30, fig. 22 Technau 1929, 30, fig. 22 Technau 1929, 30, fig. 22 unpublished Cat VII 14 Cat VII 15

L shoulder L under foot retrograde digamma by handle small circle under handle hourglass by handle L by handle. Complex sign on shoulder ‘segno’.‘lettera’ ‘arrow’ by handle L, shoulder T, handle rectangle?, shoulder ‘W’ shoulder B, shoulder T, handle +, neck and Z under handle. AN, shoulder Y, base of handle triple horizontal, neck. + on each handle diagonal, base of handle retrograde digamma. Large horizontal E digamma.TH [ partly over it + (with two horizontals), shoulder. Graffito ‘E’, shoulder mu digamma, shoulder. 7th century BC o on top of handle III on lip above handle X, below handle zeta, shoulder. Dark L, at least, on belly X, shoulder.A on handle. c. 700–650 BC zeta iota, shoulder. alpha-kappa? zeta. alpha-lambda 8 wavy horizontal, neck V, neck L N sidelong xi E?, plus? L L L L to left of handle + on shoulder + on shoulder

N393 N392 N419

Table 2 Pre-firing graffiti on Samian amphorae Athens Athens Capua Cerveteri Cerveteri Cerveteri Cerveteri Cerveteri Gravisca Gravisca Himera Himera Kamarina Kamarina Kamarina Kamarina Kamarina Kamarina Kamarina Kamarina Kamarina Kamarina Kommos Marion (T83, 3) Megara Hyblaea Montalto di Castro Naukratis Pithekoussai Qurneh Qurneh Samos Samos Samos Samos Samos Samos Samos Samos ? ?

Agora P14694 Agora P20809 Capua Villa Giulia, Banditacchia Villa Giulia, Banditacchia Villa Giulia, MA T155 Villa Giulia, MA T546 Villa Giulia, Mengarelli Tarquinia 73/26656 Tarquinia 78/9168 Imera RO3 Imera RO747 Camarina 3511 Camarina 3558 Camarina T1053, 7395 Camarina T1375, 7943 Camarina T1395, 7966 Camarina T1402, 7974 Camarina T1685, 8515 Camarina T1685, 8515 Camarina T870, 7194 Camarina T914, 7226 Pitsidia I112 Nicosia Syracuse Villa Giulia BM GR 1886.4-1.1291 Lacco Ameno 1674-7 where? Petrie Museum 16391 Heraion K3670 Heraion, Brunnen W2 Heraion, lost? lost? lost? lost? Pythagoreion 16/2/1976 Villa Giulia Villa Giulia

Naukratis: Greek Diversity in Egypt | 29

Johnston Illustration credits

Fig. 1 photograph Stuart Laidlaw, author’s drawing; Fig. 2b after Masson 1983, pl. 22.3; Fig. 7 after Williams 1983a, fig. 14; Fig. 8a after SimantoniBournia 1992, pl. 5; Fig. 8b photograph E. Feiler, Neg. D-DAI-ATH1972/269. All rights reserved; Fig. 8c author’s drawing; Fig. 16 detail: photograph Sabine Weber; Fig. 17 after Petrie 1909, pl. 54; Fig. 18 after Petrie 1888, pl. 36.3; Fig. 19 photographs Petrie Museum, University College London; Fig. 20 photograph Kommos Excavation; Fig. 22 after Petrie 1909, pl. 55; Fig. 23 after Bailey 2002, fig. 12.4; Fig. 24 author’s drawing. All other photographs: the British Museum.

Notes *

1

2

3 4 5 6 7 8 9

10

11 12 13 14

15 16

I am grateful to Dyfri Williams and Alexandra Villing for the invitation to participate in the Colloquium and for assistance in restudying some of the material included here. One may anticipate further useful results from the complete re-examination of the material. The primary publications of the material are, for Naukratis, Gardner 1886 and 1888; Edgar 1898/9 and Hogarth, Lorimer and Edgar 1905; and for Tell Defenneh, Petrie 1888. The following abbreviations are used in citing catalogue numbers in these publications: N Gardner 1886 (to N700) and 1888 (from N701) BSA Edgar 1898/9 JHS Hogarth, Lorimer and Edgar 1905 B Bernand 1970 Johnston 1979 and 1982; Williams 1983a. It is clear that the number of unpublished pieces (some simply noted by Bernand) is far greater than I believed until very recently. The apparent lack at the Samian Heraion may not be unexpected; Kyrieleis (1986) noted the meagre evidence for connection between the two islands. Boardman 1999a, 130; Jeffery 1990 index (p. 395); Möller 2000a, 167-81; Torelli 1982, 316-25; Wachter 2001, 214-19. BM GR 1965.9-30.403 and 404, from a larger closed or semi-closed vase; no reading of the two as a text is yet possible, but the boustrophedon system is used. Masson 1983, no. 164, pl. 22.2-3. Möller 2000a, 174-5. Johnston 2000a, 164-6; also noted by Wachter 2001, 216. Cook and Woodhead 1952, 161. Gamma is not so frequently attested in the 7th century bc that we can always see the manner in which each area of the Greek world distinguished its shape from that of the later letters in the alphabet row, lambda and pi. Damonidas – Williams 1983a, 184 and Johnston 1982, 40-1. It would perhaps be a tight fit to get both names, Aristophantos and Damonidas around the lip, but the calculation cannot give a precise figure for the number of letters lost. Corinthian louterion, Simantoni-Bournia 1992, 19, no. 8, pl. 5; Laconian bronze lion, Jeffery 1990, 446, no. 16a; Chian sherd from Eleusis, Johnston 2000a, 166. Clay analysis of the piece (sample Nauk 73) has, however, so far not confirmed this hypothesis and shown it to belong to a group called ITAN of unknown provenance; cf. Mommsen et al., this volume. The best preserved example is from Xanthos, Istanbul 1482, Metzger 1972, 166-70, 200; add two fragmentary ones from Olbia, Dubois 1996, no. 28a-b, with further discussion. A rare example, still unpublished (I owe the reference to Olga Palagia) is Peperia (genitive?) incised on a silver phiale from a rich early 5th century bc tomb from Vergina, noted without mention of the inscription in Andronikos 1988, 2 and Kottaridou 2005, 139. Johnston 1982, 35-7. For the statuettes see Jenkins 2000 and in general Kourou 2002. Johnston 1982, 36, no. 8, giving the abbreviation Erx-. The letter may also be used in the graffito on the neck of an Attic SOS amphora from Porto Cheli, Johnston and Jones 1978, 111, no. 40 and 131-2; the text could be read as Knidian Thox-, i.e. Theox-.

30 | Naukratis: Greek Diversity in Egypt

17 Migdol, Oren 1984, esp. 17-30, noting one ship graffito on a 'torpedo' jar. Abusir, Smoláriková 2001, and 2002, 23-46, with an overview of material from all Egyptian sites and illustration of a post-firing mu on a Samian jar (second figure on p. 117). Add large lettering of Archaic date on the neck of a probably East Greek jar from Marsa Matruh, Bailey 2002, no. 12.45. 18 Some pieces in the Department of Ancient Egypt and Sudan in the British Museum are incised with arabic numerals; one I hesitatingly include as a possibly ancient mark (3, above), though '19A' may well be the reading. Another (BM EA 23776), with '19', is a neck of generically Milesian type though in thick fabric with minimal mica. 19 Johnston 2000b. 20 Petrie 1888, pl. 32; Klazomenai, Doðer 1986, 465, figs 9 and 10. 21 Most accessible in Monachov 1999, 175 and 178-9, with figs 18 and 30, and 2003, 252-3; see also the caption to Tunkina 2003, 340, fig. 23. The material is largely a generation or more later than the Qurneh vase, which seems to be taken as Samian by Smoláriková 2002, 43, n. 276 (with misprinted number). 22 First noted by Petrie 1888, 64. For a summary of recent views see Cook 1989, 165, and for the development of the shape of the type Dupont 1998, 146-8. 23 Unlike Milesian jars, only rarely attested at Abusir and Migdol, Smoláriková 2001, 167 and Oren 1984, 20, fig. 5, and 29, figs 36-8, as cogently suggested by Dupont 1998, 216, n. 200. See also n. 18 above. 24 Petrie 1909, 16 for the generic statement that the jars were marked "after baking". The Petrie Museum amphora is no. 849. 25 Kommos, Johnston 2005, no. 220. Naukratis fractional jar, BM GR 1886.4-1.1291 (Table 1, no. 47), also has a washed out dipinto, lambda or 'arrow' delta, in dark paint, on the belly. It is not simple to disentangle the references in Petrie 1909, pl. 54, bottom right, not least because the present whereabouts of some material is unknown; the two marks drawn above '848' are presumably those on 849, where the zeta is repeated; the ligature of alpha and kappa may perhaps have been on 850. I add two further examples of pre-firing zeta: on a neck and shoulder fragment of an East Greek jar from Kommos (Johnston 2005, no. 227) and on the shoulder of, perhaps, a small, highly micaceous amphora, from Naukratis, BM GR 1886.41.92 (N344, B305); the ductus is unusual in my experience, with thin incised lines which have pushed up low, rounded ridges of clay to either side, as distinct from the gouged grooves of most pre-firing marks. 26 Johnston 1979, 38. Some 25 additional marked lids or boxes have been added in the supplementary volume, which will be published in 2006. 27 Fabbricotti 1980. For the general typology see Johnston and Jones 1978; a further distinction between ‘late 1’ and ‘late 2’ has been made, properly enough, in the Villa Giulia catalogue (see n. 34 below), but is not relevant for my purpose here. 28 Bailey 2002, 126-7, 12.36, with fig. 12.4. 29 Weber 2001, 136 and 142. 30 These details are discussed in Johnston and Jones 1978, as cited by Bailey. 31 Johnston and de Domingo 2003, 32 and 37, with further bibliography on the type. 32 I am grateful to her and to Antigoni Marangou-Lerat for allowing me to include this fragment here. 33 Lund 2004, 213. 34 I have not included dates in the tables since close dating based solely on shape is rarely possible in the period, as learned from my work on 7th century bc material from Kommos. The ‘readings’ in the final column should be taken merely as a rough guide; use of a hyphen indicates a ligature between letters. In the publication column ‘Cat.’ refers to the unpublished catalogue of the exhibition in the Villa Giulia Museum, 1983. Any graffito listed after a full-stop in the final column is post-firing. I am indebted to Federica Cordano for knowledge of some of the Kamarina jars.

‘Drab Bowls’ for Apollo: The Mortaria of Naukratis and Exchange in the Archaic Eastern Mediterranean Alexandra Villing Abstract Shallow grinding bowls made of buff clay were dedicated in considerable numbers in the Archaic sanctuary of Apollo at Naukratis. They belong to a type of mortarium that was widespread across Eastern Mediterranean households but that was also used in the preparation of foods in sanctuaries. The type is known from many sites in the Nile Delta and is also particularly common in Palestine, Cyprus and Ionia; clay analysis suggests that Cyprus was the main production centre and that the type was widely traded. Nevertheless, local imitations of Egyptian clay are also attested. The Cypro-Phoenician type inspired the production of mortaria on the Greek mainland, notably Corinth, which in turn soon became a dominant player in the market in mortaria across the Mediterranean; Corinthian mortaria of the Classical period are also attested at Naukratis.* When Flinders Petrie excavated the temenos of Apollo at Naukratis in 1885, a group of ‘coarse thick drab bowls’ caught his attention; one of them he included among the drawings of pottery in Naukratis I, along with several other fragments of coarse bowls. The majority of them carry votive inscriptions to Apollo and form a quite coherent group.1 Together with some related pieces, these bowls, now in the British Museum, are the focus of the present study. Even if at first glance they hardly seem promising material to study (and indeed they have escaped the attention of most scholars since Petrie), Petrie’s diligence in recording and preserving them was not wasted. We shall see that they provide new and unexpected insights not just

into the cult of Apollo at Naukratis but also into the network of exchange and influence across the ancient Mediterranean. 1. ‘Drab bowls’: mortaria at Naukratis Twenty-six fragments of shallow, open, undecorated bowls are today preserved in the British Museum from among the finds that Petrie made in his excavation of the sanctuary of Apollo in 1884/5. The majority of them (cat. nos 1–20) belong to the same basic type of bowl and are inscribed with graffiti naming Apollo. They are flat-based, of truncated conical shape (cf. the complete profiles of cat. nos 1–2, Figs 1-2), with a thickened, more or less oval rim, slightly wavy/rippled outside and smooth inside, and were probably made on a slow wheel, perhaps sometimes with the help of a mould. Most are made of reasonably hard, dense to slightly porous buff clay that fires pink-orange to yellow-beige, with a yellow-beige slip. Many show signs of having been produced in some considerable haste. They are obviously functional and all show clear traces of abrasion inside as well as on the underside of the base or foot. Two (cat. nos 16 and 26; Figs 3–4, the latter of different type and clay and of uncertain origin) also feature repair-holes.2 Petrie in his Naukratis publication grouped these ‘coarse thick drab bowls’, together with the ‘drab amphorae with loop handles’, in his fabric group P.3 According to his assessment, the bowls are generally early4 and often found in the same levels in the Apollo temenos as the loop-handled amphorae,5 which must have been very common indeed in the excavation:

Figure 1 Mortarium from Naukratis (cat. no. 1), inscribed twpoll[...]i

Figure 2 Mortarium from Naukratis (cat. no. 2; sample Nauk 35: group CYPT); inscribed twpoll[

Naukratis: Greek Diversity in Egypt | 31

Villing

Figure 3 Mortarium from Naukratis (cat. no. 16); repair hole; inscribed ]nun[

Figure 4 Mortarium from Naukratis (cat. no. 26); repair hole

Figure 5 Mortarium from Naukratis (cat. no. 13); inscribed ]wllon[

Figure 6 Mortarium from Naukratis (cat. no. 9; sample Nauk 56: chemical single); inscribed twpol[

Figure 7 Mortarium from Naukratis (cat. no. 18); thread marks and incised E-shaped symbol under foot

Figure 8 Mortarium from Naukratis (cat. no. 19; sample Nauk 55: group EMEA); inscribed nai kr[

Figure 9 Mortarium from Naukratis (cat. no. 20); inscribed panfa

32 | Naukratis: Greek Diversity in Egypt

‘Drab Bowls’ for Apollo: The Mortaria of Naukratis and Exchange in the Archaic Eastern Mediterranean Figure 10 Corinthian ‘tile fabric’ mortarium from Naukratis, with stamped decoration (cat. no. 21)

Figure 11 Corinthian ‘tile fabric’ mortarium from Naukratis, with spool-shaped handle and added grit inside (cat. no. 22)

Figure 12 Corinthian ‘tile fabric’ mortarium from Naukratis, with stamped rosette underneath base (cat. no. 23)

Figure 13 Heavy basin-like mortarium from Naukratis (cat. no. 24)

The great amphorae of thick greenish-drab ware, with massive loop handles, and often made by hand, being scraped down on the outside, are apparently not found above the level of the scarab factory, or 570 bc. They are so common, and at the same time I watched so continually for them in digging in order to settle their age, that this seems probably a real limit; and if so, it is valuable for fixing other dates. The great drab bowls of similar style are evidently early, as the inscriptions on them are very rude, and always retrograde on the inside, while direct on the outside.6

The inscriptions are indeed particularly intriguing. Many are placed on the inside of the rim (e.g. Figs 1–2 [cat. nos 1 and 2], 5 [cat. no. 13], 21 [cat. no. 10]), some on the outside (e.g. Figs 3 [cat. no. 16], 6 [cat. no. 9], 22 [cat. no. 12] and some under the foot (Figs 7 [cat. no. 18] and 8 [cat. no. 19]). All are very large and irregular, and scratched on after firing. Thirteen fragments (cat. nos 1–13), probably belonging to ten vessels, can be restored as dedications to Apollo on the pattern twpollwnov eimi; the others seem to record only parts of personal names, but none is preserved or can be restored completely,7 nor can we tell whether they are male or female. What are these bowls? They can be identified easily as a particular type of mortarium (shallow grinding bowl)8 that was widespread in the Eastern Mediterranean in the Archaic period in Cypro-Phoenician and Ionian contexts; I will refer to this type

in what follows as the ‘Eastern Mediterranean’ or ‘CyproPhoenician’ type.9 Parallels for the general shape are easily found at many sites in the 7th and 6th centuries bc (see below section 4). The type’s longevity, together with much variation that is of little chronological relevance, makes precise dating by shape difficult, yet the character of most of the inscriptions (many of which are retrograde) suggests a date before the middle of the 6th century bc for most pieces; Figure 8 (cat. no. 19) with its low ring base and Figure 9 (cat. no. 20) with its diskshaped base should be somewhat later.10 In addition to this group of inscribed drab bowls, there are also a number of uninscribed, shallow bowls from Apollo’s sanctuary that can equally be identified as mortaria, but as later examples of the shape (cat. nos 21–26). Two of them, with spoolshaped handles, are of uncertain origin (cat. nos 25–26, Fig. 4).11 Others are clearly mainland Greek in type and manufacture, and of 5th–4th century bc date. They comprise three Corinthian mould-made, ‘tile-fabric’ mortaria (cat. nos 21–23, Figs 10–12), two of them with elaborate, impressed designs,12 with added grit in the bottom of the bowl for increased abrasion, and a shallow mortarium of fine, sandy yellow clay (cat. no. 24, Fig. 13), with massively thickened rounded rim and a groove on the outside wall, a thickened flat base, spout and spool handles, which may Naukratis: Greek Diversity in Egypt | 33

Villing

Figure 14 Corinthian terracotta figurine of a monkey playing with a mortarium and pestle, early 4th century BC (BM GR 1903.5-18.3 [Terracotta 957])

Figure 15 Corinthian terracotta figurine of a mule carrying a mortarium, early 4th century BC (BM GR 1873.8-20.576 [Terracotta 969])

be of Corinthian (or Aiginetan?) production.13 From around 500 bc, Corinthian mortaria are found in many parts of the Mediterranean. Elsewhere in Greece, too, shallow pottery bowls had been locally produced for use as mortaria since the 7th century bc, but only the Corinthian production, it seems, gained such a wide distribution and essentially took over the market in mortaria from the Eastern Mediterranean/Cypro-Phoenician type.14 It is only through recent research that we have begun to understand more about the extraordinary spread of both these types and their production centres, and this is what I will concentrate on in the second half of this article. First, however, we should have a closer look at the mortaria from Naukratis and their context at the site, and try to explain why they are such a pronounced feature from the early 6th through to the 5th–4th centuries bc.

What exactly would have been the function of these grinding bowls? With their shallow, round shape and usually thick rim that facilitated being grasped by a human hand, they certainly seem predestined for the grinding, mashing and mixing23 of relatively soft to medium-hard materials; being made of clay, heavy pounding would have been less appropriate. They also strike one as useful for kneading, although we learn that special kneading trays (kardopoi) were also known in Greece.24 Unfortunately, there is very little evidence for exactly what foods were processed in mortaria, and virtually no evidence outside mainland Greece and before 500 bc.25 The few representations that show mortaria being used come from Boiotia and Corinth, while the literary sources largely concentrate on Classical Athens; both have to be used with caution as far as the Archaic Eastern Mediterranean is concerned, and will be discussed in greater detail elsewhere.26 Here I will only briefly summarise what can be gleaned from them. Two Corinthian terracotta figurines are particularly relevant in this context, one a monkey playfully using a mortarium and pestle (Fig. 14),27 the other a mule carrying a mortarium on its back (Fig. 15).28 The mortarium in both cases is clearly of the Classical Corinthian variety, with a long spout and spoolhandles; the pestle is of distinctive angled, ‘finger-like’ shape, as it is known from the 5th century bc onwards, in Corinth and elsewhere, made of wood or stone (cf. also Fig. 18).29 The ingredients in the mortarium carried by the mule are an indication of what might have been processed in the mortarium: a small round cheese next to a cheese-grater,30 and a bunch of garlic or onions. Already some earlier Boiotian figurines show cheese being grated into a mortarium,31 and it seems also from literary sources that in both Greek and Roman cuisine the use of a mortarium for processing cheese was quite common. Mixed with crushed herbs and garlic, it would produce the Roman spicy cheese moretum (its manufacture being described in detail in the poem Moretum), or the Greek spicy cheese sauce called myttotos/myssotos (Ar., Pax 228-88), which went particularly well with tunny, as is pointed out by the mid-6th century bc Ionian poet Hipponax.32 Other ingredients that might be ground in a mortarium are aromatics and spices (such as cumin, vinegar, silphium, and coriander – cf. Ath. 403-4), nuts and fresh green herbs, and seasonings of all kinds, for cakes, sauces, dressings and side dishes, as well as perhaps spices for wine.

2. Mortaria and their uses in the Eastern Mediterranean kitchen The function of mortaria has been the object of considerable discussion among scholars.15 Eliezer Oren,16 with regard to the wide distribution of Cypro-Phoenician mortaria alongside torpedo-shaped and basket-handled amphorae in the Eastern Mediterranean, concluded that these amphorae ‘were probably the standard transport containers for grain and oil throughout the Mediterranean basin, with the mortarium serving as a measuring bowl’. This ‘measuring bowl’ idea was shortly afterwards expounded at great length by Salles, who saw in these bowls measures for soldier’s grain rations.17 Others have put forward the idea that mortaria were used primarily or exclusively as bowls in which milk was left to curdle and turn into cheese.18 Both ideas, however, are easily refuted with reference to the very obvious traces of abrasion in the majority of Archaic and Classical mortaria,19 which are also found in mortaria from Naukratis, most obviously in cat. no. 1 (Fig. 1), where the interior is extremely worn. At Daphnae/Tell Defenneh we even find the centre of one bowl (cat. no. D1, Fig. 19) worn through. Clearly, many if not all mortaria were subject to fairly heavy-duty use,20 such as can result only from grinding with a largish (probably wooden) pestle.21 It is thus with some justification that we use the Latin term for grinding bowl, ‘mortarium’ for these bowls, which formed such an essential part of Archaic and Classical Greek and Eastern Mediterranean kitchen equipment.22 34 | Naukratis: Greek Diversity in Egypt

‘Drab Bowls’ for Apollo: The Mortaria of Naukratis and Exchange in the Archaic Eastern Mediterranean For the staples of everyday cooking mortaria should also have been useful – this would seem to be borne out by the number of grinding bowls found, for example, in domestic contexts in Archaic Miletos. They certainly seem ideal to assist the preparation of mashes and soups made from vegetables or pulses,33 but also, perhaps, porridge and gruel. Of course, heavyduty grinding of flour was usually done with the help of special stone mills – hand-mills, the saddle quern or the hopper rubber – while hulling (of barley – krithe) would have required a large holmos in which the grain was pounded with a large pestle (hyperon),34 at least if large amounts needed to be processed – for smaller quantities a mortarium may well have been useful. Perhaps stone mortaria were also sometimes used.35 A clay mortarium one could imagine, by contrast, might assist in other types of grain processing,36 such as the breaking down of grain into coarse pearl barley, cracked barley (erikis), emmer wheat, coarse meal (alphita), or semolina.37 Barley meal, alphita, certainly was a particularly useful form of cereal. It was the basis for maza (alphita kneaded with water, milk, oil or wine), which was, next to wheat bread, the main grain-based staple of the Greek diet, and for ptisane, barley gruel, a drink or soup with a medicinal function, especially when herbs and spices were added.38 It has been suggested that alphita, in the shape of precooked (moistened and sun-dried) groats,39 may have been ground and mixed with additional ingredients, such as pulses or nuts, in mortaria.40 In fact, one could imagine that all kinds of porridges,41 such as athera,42 a porridge or gruel that could be made from alphita, emmer groats (olyra, a staple in the Near East and Egypt in particular, and also the basis for the Roman porridge, puls),43 semolina (durum wheat flour – semidalis), or wheat (gruel – chondros), might have been mixed in a mortarium. Literary sources, of course, usually mention just cooking pots in connection with porridge (e.g. Ar., Plut. 673), but there is at least one instance in which a bowl is connected with athera: a Hellenistic Cypriot syllabic inscription (pre-firing) on a relatively thick-walled shallow bowl of c. 33cm in diameter, dedicated to the Nymphs, designates this vessel, uniquely, as ‘atharophoron’ (‘porringer’) – whether this means it served in the preparation of athera or was merely a container of athera remains, of course, uncertain.44 The inscription dates the bowl to the year 225/4 bc. Even though typologically the bowl is not a characteristic mortarium, its basic shape and size, as well as its clay (medium-hard creamy-buff clay with creamy-yellow slip), are closely comparable to Cypriot mortaria.45 Spicy cheese sauces for tunny may, thus, not have been all there was to Archaic mortaria. It is likely that mortaria were multi-purpose household implements that were useful in the preparation of daily foods. But what was it then that also made them suitable votive offerings for Apollo?

3. Grinding for the gods: mortaria in sanctuaries The complex of mortaria with votive inscriptions from Naukratis is unique: inscriptions on mortaria are rare,46 and nowhere else, it seems, are they as frequent as in Naukratis. The closest parallel is a single fragment of an Archaic mortarium in the Heraion of Samos (Fig. 16), of the same Cypro-Phoenician type, but with a far more abbreviated inscription: a single ‘H’ (eta) is incised on the outside of the rim.47 The mortarium is thus in all likelihood a dedication to Hera: ‘HR’ or ‘HRH’ (but also, it seems, very rarely just ‘H’) are well known as abbreviations of Hera’s name on the group of sacred pottery with dipinti that has been found in great numbers in the Samian Heraion (as well as on some examples at Naukratis – cf. Schlotzhauer and Villing Figs 14–16),48 and that was presumably used in connection with sacred meals and other rites during the first half of the 6th century bc.49 The inscription makes it likely that this Samian mortarium, too, played a role in the cult, such as in the preparation of sacrificial meals or other sacred foods, rather than being a personal dedication of a domestic implement. The numerous uninscribed mortaria found in the Samian Heraion from at least 630/20 bc onwards50 presumably served similar functions. Cult meals may well have required seasonings, sauces, mashes, meal or porridge, but we could also imagine that the special sacrificial cake made from barley, honey and oil (psaistá, a sweet version of maza) that was offered to Hera on Samos51 might have been prepared in a mortarium. Another alternative, finally, is that the mortaria might have served in the preparation of barley groats for sprinkling on sacrificial animals, but such an idea depends on whether one believes these oulai to have been hulled or ground, rather than whole and unground. Further afield, inscribed mortaria appear on Samothrace: in the Sanctuary of the Great Gods two examples of late 6th or 5th century bc date were found in the fill beneath the floor of the Hall of Votive Gifts.52 Their incised inscriptions, in the local idiom, on the upper surface of the lip read DEL and DE-, like inscriptions on other types of pottery at Samothrace, and have been interpreted as an abbreviation of ‘sacred possession of the gods’.53 Their precise function remains obscure, but they may well have fulfilled a role in the preparation of sacred meals or offerings in the sanctuary. Even more intriguing is the occurrence of inscribed mortaria at the Corinthian Asklepieion in the late 5th century bc. Two typical Corinthian mortaria – spouted, with spool-shaped handles and gritted – carry painted votive inscriptions to Asklepios, ‘Aisklap[io]’, on the outside rim; according to Roebuck they were ‘probably used in sacrificial ritual’.54 However, rather than thinking in terms of ritual dining or food for the god, one might in this particular case instead consider a different function: Aristophanes in his Wealth (710–23) describes how Asklepios in his sanctuary pounds Tenean garlic, fig juice

Figure 16 Mortarium from the Samian Heraion, inscribed ‘H’ (Samos, Heraion inv. K11146)

Naukratis: Greek Diversity in Egypt | 35

Villing

Figure 17 Mortarium from Emecik (inv. ST99K8c-16,78)

and mastix in a stone mortar, mixes it with Sphettian vinegar, and administers the mixture to the sleeping sick by spreading it on their eyelids. Though the effect is less than soothing, the scene may well point to the actual use of mortaria in the preparation of healing salves, as well as food or drinks (such as gruel) for the sick. That mortaria were indeed used for the preparation of medication55 is suggested by a passage in Aristophanes’ Frogs (123–6), where Dionysos and Herakles discuss various ways of committing suicide: hemlock ground56 in a mortarium (thyeia) is among the options rejected by Dionysos. Finally, a bowl with a Cypriot inscription to Apollo Hylates from Chytroi on Cyprus should be noted: the bowl seems to be a mortarium even if its shape is somewhat unusual, and its votive inscription clearly makes it a close parallel for the bowls from Naukratis.57 Less unusual are mortaria without inscriptions. In the Archaic sanctuary of Ephesian Artemis, pottery used in sacrificial meals during the last third of the 7th century bc and subsequently deposited in a bothros included cooking pots and mortaria of Eastern Mediterranean type.58 In the Classical and Hellenistic sanctuary of Demeter and Kore at Corinth, a range of household casseroles, stewpots, grinding-stones and mortaria presumably served for the preparation of ritual dinners as well as votive cakes.59 Many other sanctuaries also feature the odd mortarium as well as other kitchen wares, among them Archaic sanctuaries at Miletos (Aphrodite),60 Assesos in the Milesian chora (probably Athena Assesia),61 Emecik (Apollo; Fig. 17)62 and Vroulia,63 as well as the sanctuaries of Hera Akraia at Perachora64 and of Aphaia on Aigina in the 5th century bc,65 and the sanctuary of Apollo at Klaros in the 4th century bc.66 At Olympia a number of mortaria have been found in wells from the 6th to the 4th centuries bc.67 Finally, mortaria are also known from Etruscan and South Italian sanctuaries, including sanctuaries of Apollo.68 Mortaria (and other kitchen pottery) were thus connected with a wide range of cults in many regions; in most cases, and especially when they were found in bothroi, it seems that, rather than being personal votive offerings, they were used in connection with the preparation of sacrificial meals, but perhaps also of votive foods, healing foods or medication. This also seems to be supported by a much-worn 4th century bc marble Figure 18 Inscribed marble pestle from the sanctuary of Athena at Lindos, 4th century BC

36 | Naukratis: Greek Diversity in Egypt

pestle from Lindos69 (Fig. 18) that carries a votive inscription to Athena: Ka/naio/j m'a0ne/qhken 0Aqhnai/hj e0pi\ bwmo/n . The wear dates from after the inscription was incised,70 suggesting that the pestle was used in the service of the cult, and, given its association with the altar, perhaps of sacred foods or of oulai. That the preparation of foods could be an important activity in a cult is suggested by the role of aletris, (corn) grinder, that was fulfilled by Athenian girls in Athena’s cult (Ar., Lys. 642), and there may have been similar roles in other cults.71 Unfortunately, virtually nothing is known of the cult of Apollo at Naukratis other than that it was a filial cult of Apollo Didymeus Milesios,72 one of the main gods of Ionia with an important sanctuary at Didyma. We can, however, assume that, as for other gods, food played a role in his festivals, notably the Thargelia and Pyanopsia. Our best evidence for these festivals comes from Athens, but they also existed in other Ionian cities, including Miletos.73 The Pyanopsia featured a procession with an olive or laurel branch (eiresione) attached to which were fillets, figs, bread, cakes and small pots with honey, oil and wine.74 And at both the Pyanopsia and the Thargelia, a stew of different vegetables or grains was offered to the god: at the Thargelia in May, the first fruits of not yet ripe grains (thargelia) were boiled in a pot and bread was baked from the grain – seemingly both barley and wheat; at the Pyanopsia in October a sweetened porridge (athare) of beans (pyana) and other pulses as well as wheat flour was cooked in a pot (chytra), offered to Apollo, and consumed in a communal meal. In addition, food was also prepared for the pharmakoi, the two paupers chosen as scape-goats and fed at state cost for some time before the Thargelia. The Ionian poet Hipponax (frr. 5–10 West) describes how the pharmakos is feasted on figs, barley cake (maza) and cheese and is then whipped out of the city with fig branches and sea onions; of these, at least the maza might have required a mortarium. Of course, mortaria might also, more generally, have been used in the preparation of sacred cakes,75 common offerings to all Greek gods. The Milesian socalled Molpoi Inscription, a lex sacra, mentions in an insertion of the 5th or 4th century bc cakes called ‘elatra plakontina’, ‘flat sacrificial cakes’76 made separately for Apollo Delphinios and Hekate.77 Such ‘flat cakes’ were also called ‘plakous’ (after the flat, round seed of the mallow)78 and were served in the prytaneion of Naukratis during the sacrificial meals for Apollo Pythios Komaios and Dionysos (Ath. 4.149–50, quoting Hermeias [of Methymna? – 4th century bc]).79 A Classical version of this cake was close to the Roman ‘placenta’,80 a firm cake shell filled with alternating layers of honey with goat’s milk and dough, both of which, according to Cato (Agr. Orig. 76), required kneading in a mortarium.81 Meals at the prytaneion of Naukratis (Ath. 4.149f) moreover included ‘a small bowl [lekarion] of barley gruel [ptisane] or of some vegetable in its season’, another potential dish for which a mortarium might be required, alongside various types of bread, pig’s meat, eggs, and cheese; on non-festival days communal meals also seem to have been possible at the prytaneion (Ath. 4.150a). Of course, Apollo Pythios Komaios, the main civic, polis god of Naukratis,82 was distinct from Apollo Didymaios Milesios, the apparent recipient of the mortaria, but communal meals were surely also part of his cult – as might also be indicated by the 6th century bc limestone figure of a woman kneading dough that was found in the temple of Apollo.83

‘Drab Bowls’ for Apollo: The Mortaria of Naukratis and Exchange in the Archaic Eastern Mediterranean It seems likely, then, that mortaria were used for some kind of food preparation in Apollo’s sanctuary, either for sacrificial meals and ritual dining or for offerings of food, as in other sanctuaries in East Greece and beyond.84 It remains thus to consider the origin of these bowls: did they come to Naukratis together with the many other imports from the main pottery producing centres of Archaic East Greece? 4. Mortaria in the ancient Mediterranean: types and distribution As has been suggested already, most of the mortaria from Naukratis belong to the common, flat-based variety of the Eastern Mediterranean type that was extremely widespread across the whole Eastern Mediterranean and beyond.85 This wide distribution also included Egypt, and Naukratis is indeed not the only site in the Egyptian Nile Delta where many mortaria have been found. They have been recorded also at Herakleion,86 Tell Defenneh (cat. nos D1, Fig. 19, and D2, Fig. 20),87 Tell elBalamun,88 Migdol (Site T.21),89 Tell el-Herr,90 Tell Tebilla,91 Tell el-Maskhouta,92 Mendes,93 Tanis,94 Heliopolis,95 as well as at Karnak,96 Gourna,97 San el-Hagar98 and Saft el Henneh.99 At Tell Tebilla,100 the maritime port of Mendes, for example, a number of flat-based mortaria were discovered, mostly in mortuary contexts of the Late Period (664–332 bc), often alongside torpedo amphorae. Evidence for abrasion in the base of vessels suggests prolonged use before deposition. At Tell el-Herr,101 both flat-based and ring-footed mortaria of the late 6th–early 5th century bc appear in connection with Cypro-Levantine baskethandled amphorae. Also at Migdol (Site T.21), large numbers of Syro-Palestinian/Phoenician torpedo-shaped amphora as well as basket-handled amphorae were found alongside the mortaria,102 in addition to Greek pottery (especially Chian transport amphorae) and local Egyptian wares. In general, while Greek imports at these sites are usually rare, certainly as far as fine wares are concerned, the pottery finds display a strong connection with the Levant and Cyprus, suggesting that mortaria reached Egyptian trading or redistribution centres in the Delta alongside other imported goods from the Cypro-Phoenician realm. This is also supported by recent research, which has shown that from as early as 700 bc onwards mortaria are exceedingly common, especially in Cyprus and the Levant.103

In the Levant, early examples come from a shipwreck recently found in deep waters off the coast of Israel,104 as well as from Ashdod,105 Horbat Rosh Zayit in Galilee,106 Mersin,107 and possibly Tyre;108 the exceptionally early examples reported from Tarsus/Cilicia remain puzzling.109 In the course of the 7th century bc, the type becomes more widespread, especially in coastal sites, such as En Gedi, Mez.ad H.ashavyahu, Tell Keisan, Batash, Miqne, and Tel Kabri,110 usually alongside other vessels of Cypro-Phoenician origin. It continues until at least the 4th century bc. The situation is similar in Cyprus. Here, the first known occurrence of the type is associated with the first burial in tomb 79 in Salamis/Cellarka (c. 700 bc)111 where several dozen mortaria were found in the dromos. As in the Levant, mortaria are often found together with torpedo and basket-handled amphorae. The type then continues, in tombs at Salamis and elsewhere,112 through Cypro-Classical113 until about 300 bc.114 Variations in the shape of these Cypro-Phoenician mortaria can be observed right from the beginning, but most are only of limited chronological relevance,115 such as the rim shapes (slim and elongated, triangular, rounded, more or less undercut),116 and the degree of waviness of the outside wall. The only truly significant development is the introduction of a variety with a ring foot in the 6th century bc,117 which becomes common in the 5th and 4th century bc in the Levant, Cyprus118 and Egypt.119 As regards the findspots of Cypro-Phoenician mortaria, in Cyprus they were often placed in tombs,120 either unused (i.e. bought especially for deposition in the tomb), or perhaps used just once or twice, e.g. during a ritual meal. It has been suggested that they served as containers for food for the deceased,121 but one could also imagine them more generally as a symbol of wealth and plentiful food supplies. Sometimes they were also used as lids for amphorae in children’s burials.122 That they could also serve regular domestic functions is suggested by the find in Amathus of over 400 fragments of mortaria in settlement contexts, in the area of the harbour and the so-called palace.123 In Palestine they are regularly encountered in residential contexts, and with clear traces of use.124 Cyprus and Palestine, then, are clear centres for the use of Eastern Mediterranean mortaria; however, recent excavations have also firmly placed Ionia on the map.125 Scattered examples had been known for some time from Crete,126 Rhodes,127 Smyrna128 and the Samian Heraion (Fig. 16),129 but recent work

Figure 19 Mortarium from Tell Defenneh (cat. no. D1)

Figure 20 Mortarium from Tell Defenneh (cat. no. D2)

Naukratis: Greek Diversity in Egypt | 37

Villing at sites such as Miletos, Klazomenai, Ephesos, and Knidos/ Emecik (Fig. 17) has dramatically altered the picture. At Miletos in particular, a great number of mortaria has been found in the 7th and 6th century bc settlement. They show obvious traces of abrasion and were clearly much used in the Archaic Milesian kitchen, outnumbering even cooking pots.130 What is interesting, however, is that the ring-footed version of the mortarium, which in Cyprus and the Levant became common from the later 6th century bc onwards, seems to find few parallels in East Greece.131 Furthermore, unlike in Cyprus, mortaria were not placed in tombs, and their association with Cypriot baskethandled amphorae is far less strong than elsewhere, as the latter are attested only relatively infrequently in Ionia.132 Mediation by Ionia probably explains the appearance of mortaria in the Milesian colonies on the Black Sea coast, such as Berezan133 and Histria.134 Related types, finally, also appear in North Africa (Tocra,135 Carthage136), Punic Sardinia (e.g. Nora, Tharros),137 Southern Italy, Sicily and Etruria,138 as well as Spain (e.g. Malaga, Ampurias).139 Notable is the absence of mortaria of the Eastern Mediterranean type on the Greek mainland, with the exception of two examples at Corinth (late 7th–early 6th century bc)140 and one from the Athenian Agora.141 Yet this ‘Great Divide’142 between the Greek mainland and the Eastern Mediterranean world is bridged to some extent later on by the eager adoption of the Eastern Mediterranean mortarium shape on the part of the Corinthians, who at this time maintained close connections143 with East Greece: Corinthian mortaria in their earliest form144 may well owe their existence to inspiration from imported mortaria, as their shape knows no local predecessor and broadly reflects the Eastern Mediterranean type, even if the fabric is from the beginning distinctly Corinthian. 5. Eastern Mediterranean mortaria: production and trade This wide distribution of Eastern Mediterranean mortaria – which in some ways mirrors the 2nd millennium bc spread of stone tripod bowls (mortaria) from Anatolia to the Aegean, a form which also existed in a terracotta version and to which they may in fact be partial successors145 – has given rise to some discussion as to the origin of the type. Twenty-five years ago, Ephraim Stern still wrote that: Since the main distribution of the early bowls [...] is in the Greek colonies on the eastern coast of the Mediterranean […], it seems that we should rather seek its origin in the East-Greek cultural sphere, from which it spread south and east. A Cypriot source for these bowls is also impossible for they first appear on that island at a later date and continue for only a short time.146

Stern further assumed that the (later) variety with ring-foot might have been a local Levantine version.147 Already Petrie had considered the ‘thick drab bowls’ an import into Egypt and Palestine, probably from Greece.148 We know today that in Ionia the type indeed had a significant presence from the early 7th century bc onwards. But at least as early, if not earlier, are the finds from Cyprus and the Levant, where the type was also extremely long-lived. Although in the Levant the bowls appear often in coastal settlements, where also imported Greek finewares may be found,149 in Cyprus there is no particular connection with Greek pottery, and, instead, mortaria are often found with Cypro-Phoenician torpedo and basket-handled amphorae (as, indeed, also in the Levant and Egypt). In fact, new research suggests that the picture is rather 38 | Naukratis: Greek Diversity in Egypt

complex, and involves not just the spread of a type across a large region, but also trade across a wide area over a long period of time. This is indicated especially by clay analysis performed on mortaria in the Levant over the past two decades. Contrary to what one might have expected, this has shown that virtually none of the analysed pieces were produced locally, but almost all must have been imported from one or more centres abroad. One such centre is the coastal region of North Syria, around Ras al-Bassit. Here, a Late Roman workshop supplied mortaria to much of the Eastern Mediterranean and beyond, especially Palestine, Egypt, and Cyprus, but also Athens, Rome, Germania and Britannia.150 It has been suggested that a similar trade might already have taken place in the Archaic period, as some Archaic mortaria have a similar petrographic profile to the later Roman pieces. This includes one of the earliest mortaria from Palestine, the late 8th-century mortarium from the Elissa wreck off Ashkelon.151 However, the petrographic profile of the northeastern coast of the Mediterranean and that of Cyprus can be quite close. Thus, the extensive petrographic and NAA analysis of 5th century bc ring-footed and flat-based mortaria from Tell el-Hesi152 have suggested a North Syrian or a (Southern) Cypriot or possibly South Anatolian origin, i.e. a region with outcrops of limestone and ophiolitic rocks. Similarly, mortaria from Tell Keisan153 are thought to have been imported from Cyprus, North Syria or Eastern Anatolia. Ring-footed mortaria from Tell Anafa show great petrographic similarity to the mortaria of Tell elHesi, but feature additional gastropod shell temper and have been identified as Phoenician ‘White Ware’.154 Petrographic analyses have attributed mortaria from Timnah – Tell Batash155 to a region with metamorphic or volcanic rocks, probably the Eastern Aegean; three ring-footed mortaria of c. 450–350 bc from Apollonia-Arsuf, as well as further mortaria from YavnehYam, Yaoz and Tel Michal, to a (Western) Cypriot or Aegean/ East Greek origin;156 flat-based mortaria from Mez.ad H.ashavyahu to an ophiolitic region, probably Cyprus;157 and a flat-based mortarium of the second half of the 7th century bc from Tel Kabri to a Cypriot coastal centre such as Amathus or Enkomi.158 Furthermore, three ring-footed mortaria from the late 5th century bc shipwreck off Ma‘agan Mikha’el, south of Dor,159 have emerged as Cypriot.160 Cyprus thus presents itself as a highly likely major production centre for mortaria at least from the 7th century bc onwards. Reports of misfired examples of ring-footed, greenish buff mortaria in the harbour area of Amathus, alongside Salamis one of the island’s main cosmopolitan trading communities,161 further support this assumption. Another possible production centre has been proposed at Kition/ Larnaka, for classic ‘Plain White’ pottery; Jean-François Salles claims to have recognized – through visual observation – mortaria of this class of pottery among material in Ashdod, Hazor, Lachisch, Tel Michal and especially Tell Keisan.162 Certainly connections between Cyprus, the Levant and Phoenicia were strong in the Archaic period, continuing the tradition of a ‘West Asiatic Trading Sphere’, possibly involving not just Phoenician but also Cypriot and East Greek traders, with Amathus in particular functioning as a gateway for trade with the Levant and Egypt.163 Other regions, however, should not be ruled out as further production centres for mortaria, notably the Eastern Aegean/ East Greek region. This is suggested by the recent petrographic analysis by Daniel Master of flat-based mortaria from

‘Drab Bowls’ for Apollo: The Mortaria of Naukratis and Exchange in the Archaic Eastern Mediterranean Figure 21 Mortarium from Naukratis (cat. no. 10; sample Nauk 67: group EMEa)

Figure 22 Mortarium from Naukratis (cat. no. 12; sample Nauk 68: group EMEa)

Ashkelon164 – all dating to before 604 bc, the date of the destruction of Ashkelon – which points to the production of two mortaria in North Syria or Cyprus (Master’s Category 9: ‘Ultramafic rocks and micritic sand’),165 and of another in the Southern Aegean/Ionia (Master’s Category 14: ‘Highly micaceous samples with yellow fabric’);166 one further mortarium remained unclear (Category 12: ‘dark brown/black fabric’).167 On the basis of such findings, Master,168 as well as Bennett and Blakely169 assumed specialised workshops in the North Eastern Mediterranean, in Cyprus and/or possibly Northern Syria, which from the 8th and 7th century bc exported mortaria to the Levant, the Aegean and North Africa, thus dominating the Mediterranean market in clay mortaria.170 One is reminded of the spread of basket-handled amphorae (often found in conjunction with mortaria, as we have seen), produced on Cyprus since the late 8th century bc but later, at least from the 5th century bc onwards, also elsewhere.171 However, only further analysis, in particular of mortaria found in Ionia, can ultimately shed more light on the actual range of production centres of Archaic mortaria.172 6. The mortaria of Naukratis: Cypriots, Phoenicians and Greeks in Egypt How does Naukratis fit into this picture? NAA analysis conducted by Hans Mommsen173 suggests a Cypriot origin for four of the mortaria from Naukratis (Mommsen et al. Fig. 1). One (cat. no. 2, Fig. 2, sample Nauk 35) falls into a group (CYPT) that also contains a Cypro-Mycenaean three-handled jar (FS47) of LH IIIA2 date and that is close in composition to Cypriot pottery from Enkomi and Milia.174 Three (Figs 21 [cat. no. 10], 22 [cat. no. 12] and 8 [cat. no. 19], samples Nauk 55, 67 and 68; groups EMEA and EMEa) go together with material from Emecik on the Knidos peninsula that includes Cypriot terracotta figurines175 and one fragment of Cypro-Archaic I painted pottery.176 The mortaria from Naukratis thus mirror what has been established for much of the Levant, and what seems likely also for other mortaria in Egypt. We have already seen earlier that, as

in Cyprus and the Levant, mortaria in Egypt seem to be strongly associated with other types of Phoenician/Cypriot pottery, notably storage jars of torpedo and basket-handle shape.177 We may thus assume that mortaria reached the Nile Delta as part of cargoes of Cypro-Phoenician wares, which might at times have included East Greek pottery,178 and which may well have been carried by Phoenician traders; they might in part have also been targeted at the resident Judaean, Phoenician179 and Cypriot180 (mercenary) population in Egypt. Indeed, as Diodorus (1.68.8) points out, Phoenicians and Greeks were the main traders admitted into Egypt since the time of Psammetichos I.181 That such trade may date back to as early as the late 8th century bc is suggested by the Elissa shipwreck off Ashkelon, which in addition to Phoenician pottery, including torpedo-amphorae, contained not only a mortarium of Eastern Mediterranean type (in all likelihood, though, just for the use of the crew)182 but also a piece of Egyptian pottery.183 The picture at Naukratis, at least at first glance, does not seem to fit particularly well the pattern encountered elsewhere in Egypt and in the Levant. Greek finds dominate by far, dwarfing the very little Phoenician (Schlotzhauer and Villing Fig. 24)184 and not much Cypriot evidence. Cypriot fine ware pottery is hardly traceable;185 there are only two Cypro-syllabic inscriptions, both of Classical date (Höckmann and Möller Fig. 6), some Cypriot-style statuary (some of it perhaps produced at Naukratis, though this is still disputed),186 Cypriot terracotta statuettes,187 and a handful of Greek-inscribed Cypriot amphorae (e.g. Johnston Fig. 14).188 This all mirrors more the situation in East Greece and the Aegean in general, where Cypriot fine wares are also virtually absent, but terracotta figurines and limestone sculptures appear as dedications in sanctuaries. However, we need to remember that Petrie himself had claimed that Cypriot basket-handled amphorae (his class P) were ‘common’ in his early level, and there may have been many more uninscribed or undecorated mortaria, too. The picture at Naukratis, then, may have been somewhat closer after all to what we know from other sites in Egypt, though still undoubtedly of a far stronger Greek character than elsewhere. Both the Cypriot amphorae and mortaria, moreover, bear inscriptions in Greek script, which, as Naukratis: Greek Diversity in Egypt | 39

Villing Figure 23 Mortarium from Naukratis (cat. no. 17, sample Nauk 18: Egyptian Marl)

Alan Johnston189 has pointed out, ‘would not have been used by any Cypriot (Greek or Phoenician) in the archaic or early classical period’. Even in the light of the new evidence from the mortaria, Möller is thus probably right to conclude that there is very little evidence for Cypriots at Naukratis, although this does not exclude the possibility of some of the trade at least having been undertaken by Cypriots, or even of the odd Cypriot trader or mercenary passing through Naukratis. Of course, all this goes decidedly against the common perception that, at least in Archaic and Classical times, coarse, household pottery, was normally not traded, either because such simple pottery could easily be produced locally, or because deeply enshrined local food customs would not allow for the import of cooking and food preparation wares. Yet it seems increasingly that such trade did, in fact, take place, especially, though not exclusively, in a colonial or diaspora environment. Later on, imported items from the home region were eventually supplemented or replaced by locally produced ones or ones imported from elsewhere. This seems to be exactly what happens also with the Cypro-Phoenician mortaria in the West: They first appear in Punic settlements, presumably via Cyprus, in the later 7th and early 6th centuries bc, but from the middle of the 6th century bc onwards, they become part of the Carthaginian cultural koine and are often locally produced and no longer recognizable as Cypro-Phoenician.190 The phenomenon of local production (and import from other sources) can be witnessed in Naukratis, too: a further mortarium that was analysed by NAA (Fig. 23, cat. no. 17, sample Nauk 18) falls into a chemical group (Marl) clearly of Egyptian composition and is also visually recognisable as marl.191 Given the number of examples of locally produced painted Greek-style vessels in a range of shapes that have recently become known from Naukratis,192 as well as the long history of the imitation of other wares in Egypt,193 the find of locally produced mortaria (to which also cat. no. 1, Fig. 1, may belong) may hardly seem surprising. It is, in fact, not a one-off: both mortaria and Cypro-Phoenician storage jars are reported from Tell Tebilla,194 Egyptian mortaria may have been found at Tell El Balamun and Tell Defenneh (Fig. 19, and possibly Fig. 20),195 and there are imitations of Greek trade amphorae from Tell Defenneh and T. 21 (Migdol) as well as a cooking pot ‘made of Nile ware after a Greek shape’.196 At T.21 (Migdol) there is also a smaller bowl of mortarium shape, made of ‘levigated Nile clay, tempered with black grits and mica, fired red-brown’ with a ‘thick matte creamy slip’ – a description which fits quite well with the ‘regular’ Naukratis workshop fabric.197 However, the Naukratis mortarium as well as most of the other mortaria and amphorae are produced of marl, suggesting either that for different kinds of shapes different clays were used or that different workshops were responsible for at least some of these vessels. 40 | Naukratis: Greek Diversity in Egypt

Yet this local production never seems to have flourished all that greatly, probably not least because Egyptian marl clay mortaria proved no match for sturdy Cypriot vessels. Potters and their customers seem to have been quite aware of the properties of certain clays to make pots good for certain functions, particularly if it was a matter of heavy-duty daily use; this seems suggested in particular for cooking pots,198 for which the island of Aigina, termed ‘chytropolis’, appears to have been famous,199 and which also may have been imported from East Greece into the Levant.200 For mortaria, too, such considerations may have played a role201 – presumably already with regard to the widely traded Cypro-Phoenician mortaria, but certainly for Corinthian mortaria from the Late Archaic-Early Classical period onwards. The most common Corinthian types, made from extremely hardwearing Corinthian ‘tile fabric’ with an additional scattering of grit in the lower part of the basin for increased abrasive effect (cf. especially cat. no. 22, Fig. 11),202 were traded to many places across the Mediterranean including Athens203 (where they seem to have replaced local production to a large extent), Miletos,204 Histria,205 Cyrene,206 South Italy and Etruria, as well as, as we now know, Naukratis (cat. nos 21–23, Figs 10–12).207 Also at 4th and 3rd century bc Euhesperides (Cyrenaica),208 Keith Swift has established a very high proportion of imported coarse-wares,209 including Corinthian mortaria and Aeginetan cooking pots: they may have reached North Africa as part of the same trade assemblage, alongside Attic black glaze vessels.210 Quality mattered not just for fine wares, but also for common household pottery. 7. Conclusion We have seen that mortaria form a small but significant group among the pottery from Apollo’s sanctuary at Naukratis from the first half of the 6th to at least the later 5th century bc. Many of them carry votive dedications to Apollo and are likely to have played a role in the preparation of sacred or communal foods in the sanctuary. Several of the Classical examples are of Corinthian manufacture, at the time a prolific exporter of highquality mortaria across the Mediterranean. That such a trade existed already in the Archaic period is, however, indicated by the earlier mortaria from the sanctuary, which could be shown to belong to an Eastern Mediterranean, or Cypro-Phoenician, type that was widespread in a koine encompassing Ionia, Cyprus, the Syro-Palestinian coast, northern Sinai and the Nile Delta,211 and including to some degree the Phoenico-Punic regions of North Africa and the Western Mediterranean and the Ionian settlements on the Black Sea coast – a koine that seems to have extended to culinary habits,212 and in which Cyprus played a large part in the production of goods, and where even coarseware mortaria could be traded long distances. Who the traders were remains open to discussion. In Naukratis certainly the mortaria were used and dedicated by Greeks, similar to the

‘Drab Bowls’ for Apollo: The Mortaria of Naukratis and Exchange in the Archaic Eastern Mediterranean Cypriot basket-handled amphorae inscribed in Greek characters, concerning which Alan Johnston has concluded that ‘Ionians were trading amphorae of a possible Cypriot origin with Naukratis; possibly Knidians and Aeginetans were also involved’. Given the evidence discussed above, however, it seems that Phoenicians, too, as well as perhaps Cypriots, may have been involved,213 and in later periods perhaps mainland Greeks, too. The mortaria of Naukratis thus bear witness to a trade in mortaria, first, from the beginning of the Archaic period, of Cypro-Phoenician origin, and later, throughout the Classical period, of Corinthian manufacture – a situation that certainly contradicts the common assumption that coarse wares were not traded. For later periods the phenomenon of bulk trade in mortaria has of course been known for some time: in Roman times, North Syrian mortaria – popular quite possibly because of their hard and sharp igneous temper214 – were traded, as we have seen, to much of the Eastern Mediterranean and beyond, even as far north as Germany and Britain.215 Here, mortaria for the first time appeared only after the Roman conquest,216 as part of the spread of the Mediterranean cuisine that required new vessels, such as the dolium and the mortarium. Perhaps not quite in the same way, but also as part of a movement of peoples, culinary customs or techniques, the mortarium had arrived in Rome centuries earlier from Greece and the Phoenico-Punic region.217 And some centuries earlier, we now begin to see that it had come to mainland Greece from the Eastern Mediterranean region, at a time when increased contact through trade and settlements led not just to the adoption of important cultural features such as the alphabet, but quite possibly also ‘kick-started’ the development of the Corinthian mortarium, which some decades later was to take over the Mediterranean market in mortaria from its Cypro-Phoenician predecessors. Defying expectations raised by their unprepossessing appearance, mortaria thus shed unexpected light on the network of trade and shared culture that linked the various cultures of the Archaic and Classical Mediterranean. Catalogue of mortaria from Naukratis and Tell Defenneh in the British Museum 1. GR 1886.4-1.790 (Fig. 1) Nearly complete bowl, long slim collar rim, flat base. Clay buff to light orangebrown, fairly hard, some vegetable inclusions (marl clay?). Graffito on inside wall below lip: vac twpoll[…]i vac retrograde. V-shape incised on inside wall opposite inscription. H. of bowl 8.4cm, Diam. 30.7cm. Petrie 1886b, 61 no. 77, pl. 32.77; Bernand 1970, 646 no. 38.

2. GR 1886.4-1.80 (= Nauk 35; CYPT) (Fig. 2) Large fragment of bowl (2 joining fragments), wavy outside wall, flat base. Beige-yellow clay, core light orange, quite dense and hard. H. 15.8cm (original H of bowl 8cm), W. 31cm, Diam. 32cm. Graffito on outside wall below lip: ]twpollw[. Petrie 1886b, 61 no. 78, pl. 4.2 (presumably) and 32.78; Bernand 1970, 646 no. 39, pl. 19.2.

3. GR 1886.4-1.71 3 joining fragments from rim of bowl. Light yellow-beige clay, relatively dense and hard. W 25.5cm, H. 8.2cm, Diam. 36cm. Graffito on inside of lip: ]twpollw.[ retrograde. Petrie 1886b, 61 no. 68, pl. 32.68; Bernand 1970, 645 no. 29, pl. 19.

4. GR 1886.4-1.72 2 joining fragments from rim of bowl. Light yellow-beige clay, relatively dense. W. 11.5cm, H. 6cm, Diam. 29cm. Graffito on inside of lip: ]emi retrograde. Petrie 1886b, 61 no. 69, pl. 32.69; Bernand 1970, 645 no. 30, pl. 19.

5. GR 1910.2-22.18 Probably joining 1886.4-1.79; possibly the same bowl as 1886.4-1.74. Fragment from rim of bowl. Yellow-green clay, core light orange. H. 6cm, W.

10cm, Diam. 41cm. Graffito inside below lip: ]w[. Petrie 1886b, no. 402, pl. 34.402; Bernand 1970, 673 no. 317 (‘il ne s’agit là que de debris’).

6. GR 1886.4-1.73 Three joining fragment from rim of bowl (top lost). Very light, yellowgreenish, friable clay. H. 6cm, W. 10cm. Graffito on inside of rim: ]pol[ retrograde. Left stroke of lambda not quite as in Petrie 1886b. Petrie 1886b, 61 no. 70, pl. 32.70; Bernand 1970, 645 no. 31.

7. GR 1886.4-1.79 Possibly the same bowl as 1910.2-22.18 and 1886.4-1.74. Fragment from rim of bowl. Clay yellow-beige, core light orange. H. 7.5cm, W. 10cm. Graffito on inside of rim: ]oll[ retrograde. Bernand 1970, pl. 19.

8. GR 1886.4-1.74 Possibly the same bowl as 1910.2-22.18 and 1886.4-1.79. Two joining fragments from rim of bowl, Light yellow clay, fairly dense, core pinkish-red. H. 10cm, W. 13.5cm, Diam. 43. Graffito on inside of lip: vac kri[ retrograde (Bernand reads kai). Petrie 1886b, 61 no. 71, pl. 32.71; Bernand 1970, 645 no. 32, pl. 19.

9. GR 1886.4-1.75 (= Nauk 56; single) (Fig. 6) Large fragment from bowl, wavy outside wall, flat base. Yellow clay, light and chalky. H 15cm (H of bowl 8.6cm), W 13.8cm, Diam. 32cm. Graffito on outside wall below rim: twpol[. Last letter has longer vertical than in the drawing in Petrie 1886b. Petrie 1886b, 61 no. 72, pl. 32.72; Bernand 1970, 645 no. 33, pl. 19.

10. GR 1886.4-1.77 (= Nauk 67; EMEa) (Fig. 21) Large fragment from bowl, wavy outside wall, flat base. Light yellow clay, relatively dense. H 13.4cm (original H of bowl 9.4cm), W. 10.1cm, Diam. 29cm. Graffito on inside of rim: ]wpoll[ retrograde. Omega cut lower than other letters. Petrie 1886b, 61 no. 74, pl. 32.74; Bernand 1970, 645 no. 35, pl. 19.

11. GR 1886.4-1.78 Fragment from rim of bowl. Yellow-beige clay, core light orange, fairly dense clay. H 7.5cm, W. 7.2cm, Diam. c. 30cm. Graffito on inside of rim: ]wn[. Messy writing, unclear whether second letter is orthograde or retrograde nu; reading as lambda is unlikely. Petrie 1886b, 61 no. 76, pl. 32.76; Bernand 1970, 646 no. 37.

12. GR 1886.4-1.81 (= Nauk 68; EMEa) (Fig. 22) Fragment from rim of bowl. Yellow-beige clay, fairly dense and hard. H. 12.3cm, H. 8.4cm, Diam. 31cm. Graffito on outside wall below rim: ]lwnoj[. Bernand incorrectly reads ]olwno[. Marked in pencil inside, possibly reads ‘40 ft W out tem Ap’. Petrie 1886b, 61 no. 79, pl. 32.70; Bernand 1970, 646 no. 40, pl. 19.

13. GR 1888.6-1.390 (Fig. 5) Fragment from rim of bowl. Yellow-greenish clay, quite hard. H. 8cm, W. 11.8cm, Diam. 31cm. Graffito on inside wall below rim: ].wllon[ orthograde; first letter difficult, last seems to be nu. Bernand 1970, pl. 19.1 (illustrated but not included in catalogue).

14. GR 1886.4-1.76 Fragment from rim of bowl, very slim profile. Yellow-beige clay, core light orange, fairly dense and hard. H. 6cm, W. 6.5cm, Diam. 35cm. Graffito on inside of rim: ]nj e[ retrograde. Three-bar sigma. Last letter uncertain, might also be iota. Bernand reads, less likely, ]isn[. Petrie 1886b, 61 no. 73, pl. 32.73; Bernand 1970, 645 no. 34; Johnston 1974, 97.

15. GR 1888.6-1.391 Fragment from rim of bowl, rim only barely set off from wall. Yellow-beige clay, core light orange, relatively dense. H. 8cm, W. 11.5cm, Diam. 38cm. Graffito on inside of rim and bowl, in two lines: ]ari[ | ]e[. Inscription is, unusually for an incised inscription on the inside, clearly orthograde. Bernand 1970, pl. 19.

16. GR 1910.2-22.17(= Nauk 52; single) (Fig. 3) Fragment from rim of bowl, slightly thickened rim; repair hole underneath rim. Greenish-yellow clay, quite hard. H. 4.5cm ,W. 11.4cm, Diam. 35cm. Graffito on lip: ]nun[. Petrie 1886b, pl. 34.401; Bernand 1970, 673 no. 317 (‘il ne s’agit là que de debris’).

17. GR 1910.2-22.15 (= Nauk 18; Marl) (Fig. 23) Fragment from rim of bowl, very long and slim collar rim. Clay brown, porous, with vegetable inclusions (Egyptian marl clay). H. 6.5cm, W. 8cm., Diam. 33cm.

Naukratis: Greek Diversity in Egypt | 41

Villing Graffito on outside of rim: ]nel[. Petrie 1886b, pl. 34.399; Bernand 1970, 673 no. 317 (‘il ne s’agit là que de debris’).

18. GR 1886.4-1.792 (Fig. 7) Fragment of flat base of bowl with distinctive thread marks. Graffito underneath foot: E-shaped sign. L. 11.5cm, Diam. base c. 20cm. Petrie 1886b, 61 no. 75, pl. 32.75; Bernand 1970, 646 no. 36.

19. GR 1910.2-22.16 (= Nauk 55; EMEA) (Fig. 8) Fragment of base of bowl with low broad ring foot. Inside of bowl worn, foot slightly worn. Clay light orange, relatively dense and fine, outside surface beige. H 10.5cm, W 8.4cm, Diam. foot 21cm. Graffito underneath base: ]nai Kr[. Petrie 1886b, pl. 34.400; Bernand 1970, 673 no. 317 (‘il ne s’agit là que de debris’).

1 2 3

4

20. GR 1886.4-1.1769 (Fig. 9) Fragment from bottom of bowl, low disk foot. L. 15.2cm, Diam. foot 20cm. Graffito underneath foot: vac panfa vac. Unpublished.

21. GR 1888.6-1.762 (Fig. 10) Fragment from rim of bowl with stamped decoration. Corinthian ‘tile fabric’. H. 8.1cm, W. 12cm, Diam. 38cm. Petrie 1886b, pl. 4.5.

22. GR 1965.9-30.539 (Fig. 11) Fragment from rim of bowl with spool-shaped handle; added grit inside. Corinthian ‘tile fabric’. H. 8cm, W. 10.6cm, Diam. 27cm. Petrie 1886b, pl. 4.7.

23. GR 1965.9-30.540 (Fig. 12) Fragment from base of bowl; part of foot broken off, stamped rosette underneath centre of base. Corinthian ‘tile fabric’. L. 15.1cm, W. 10.6cm. Petrie 1886b, pl. 4.6.

5 6 7

24. GR 1965.9-30.537 (Fig. 13) Large fragment from heavy basin, heavy rounded rim, groove around outside wall, spool-shaped handle. Fine, sandy yellow fabric. H. of bowl 6.9cm, W. 22.2cm, L. 13.7cm, Diam. 29cm. Unpublished.

8

25. GR 1965.9-30.538 Fragment from rim of bowl with ribbed spool shaped handle. L. 7.8cm, W. 8.1cm, Diam. c. 26cm. Petrie 1886b, pl. 4.8.

26. GR 1965.9-30.966 (Fig. 4) Fragment from rim of bowl, with spool-shaped handle; fine sandy red clay; repair hole. H. 10cm, W. 16.5cm, Diam. 27cm. Unpublished.

D1. EA 23685 (1887.1-1.1258) (Fig. 19) Dark yellow-green clay with brown-reddish-grey core, porous with some vegetable matter; inside worn away so much that hole is worn into base; underside of base worn, too. H. of bowl 6cm, Diam. 23.8cm. Unpublished.

D2. EA 23703 (1887.1-1.1215) (Fig. 20) Beige clay, relatively dense but with some vegetable matter; inside worn. H. of bowl 8.8cm, Diam. 27.5cm. Unpublished.

Illustration credits

9 10 11

Fig. 16 author; fig. 17 D. Berges; fig. 18 Copenhagen, National Museum. All other photos the British Museum; all drawings K. Morton/A.Villing.

Notes *

Many colleagues and friends contributed to this study, and I am grateful to all of them. I would like thank in particular Volkmar von Graeve for suggesting to me to study and publish the Archaic bowls from Miletos, and thus kindling my initial interest in ‘drab bowls’; Regina Attula, Ahmet Aydemir, Donald Bailey, Iulian Bîrzescu, Bodil Bundgaard-Rasmussen, Dmitry Chistov, Elizabeth Greene, Alex Fantalkin, Alexander Herda, Bilge Hürmüzlü, Alan Johnston, Michael Kerschner, Gudrun Klebinder-Gauss, Birgit Konnemann, Astrid Lindenlauf, Hans Mommsen, Stavros Paspalas, Elizabeth Pemberton, Marcella Pisani, Udo Schlotzhauer, Jeffrey Spencer, Keith Swift, Jonathan Tubb, and Dyfri Williams for constructive discussions, helpful advice and sharing their own knowledge and material; the audiences at lectures at the Australian Archaeological Institute at Athens in 2005, at the conferences ‘Formes et usages des vases grecs’ in Brussels and ‘Walls of the Ruler’ in Swansea in 2006 for their comments and suggestions; and Helen Clark, Jan Jordan,

42 | Naukratis: Greek Diversity in Egypt

12

13

Guy Sanders, Ioulia Tzonou-Herbst, Wolf-Dietrich Niemeier and Birgit Konnemann for making possible the study of mortaria at the Athenian Agora, ancient Corinth, and Samos. Alan Johnston generously contributed readings of the votive inscriptions. Petrie 1886bb, pl. 4. Repair of mortaria of this type is attested also elsewhere, for example in domestic contexts at Miletos, suggesting that these bowls were valued enough to warrant the effort of repair. Petrie 1886bb, 18, 20, pl. 17.17, 20-1. These amphorae are what we now call (Cypriot) basket-handled amphorae: Johnston 1982, 35-7; Johnston, this volume, Table 1, Johnston Fig 14. See also below, n. 171. ‘A bowl of thickish drab (P), with a short vertical brim’, together with a large ‘P’ amphora, is recorded in the stratum above the burnt layer, but below the scarab factory level, in the southern part of Naukratis town (Petrie 1886bb, 21); ‘P thickish drab bowl with short vertical brim (like that of 320 level)’ in a higher level from the clearing of a road about 200 feet north of the scarab factory (ibid, 22); ‘P, dish, small spout, and a conical bottom of a vase’ at a higher level from road-mending on the east of the town (ibid, 22-3). Basket-handled amphorae of the same type are, in addition to those just mentioned, recorded in Petrie’s ‘deepest strata’ of Naukratis town, at the east of the south wall of the temenos of Apollo (ibid, 21), in the burnt layer of the southern part of the town (ibid), and in the scarab factory layer (ibid p. 22). Petrie 1886bb, 20. Petrie 1886bb, 23. One might be tempted, of course, to associate the inscription ‘Kri-’ on cat. no. 8 with the name of Krithis, attested in a graffito at Abu Simbel as the name of one of the mercenaries in the service of Psammetichos I (cf. Haider 2001, 204, 213 fig. 3). On cat. no. 19 (Fig. 8) a similar name might be identified, but the mortarium is clearly much later in date. The inscription ‘Panpha’ on cat. no. 20 (Fig. 9) must be an abbreviation; the ‘E’-like shape on cat. no. 18 (Fig. 7) may be a symbol (trade mark?). This type of bowl is known under a number of different names in scholarly literature, such as ‘Persian bowl’ (based on the assumption that they are diagnostic for the Persian period in the Levant – see below section 4), ‘cuvettes lourdes’ or ‘plats creux’, ‘open bowls’, as well as ‘Reibschüsseln’ or ‘Reibschalen’ (‘grinding bowls’) and ‘mortaria’; for a discussion of the terminology, see Sapin 1998, 88-90. I have opted here for ‘mortaria’, since the function of these bowls (as discussed below) in essence corresponds to that of the Roman mortarium, which was a grinding bowl much used in the Roman kitchen, especially for the grinding of spicy sauces and the grinding or mixing of other soft creamy and dough-like substances, often together with spices, and distinct from a pounding bowl for grain: Hilgers 1969, 68-70, 225-7, cf. also 252, and Baatz 1977. On the term mortarium and its use in scholarship, see also Blakely and Bennett 1989, 49-50. The term ‘bacino di tipo fenicio-cipriota’ has also been chosen by Bellelli and Botto 2002 in their recent comprehensive survey of the type in the West. Ring bases for mortaria seem to have been introduced in the course of the 6th century bc; see also below, ns 117 and 131. Spool-shaped handles with dense vertical ribbing occur frequently on mortaria of heavy basin-shape like our cat. no. 24 (Fig. 13), from the latter part of the 5th century bc onwards, through to the Hellenistic period (Sparkes and Talcott 1970, 370 no. 1913, pl. 91), but they are also found elsewhere and on other types of mortaria: cf. e.g. Olynthus XIII, 414 no. 1030, pl. 250. A close parallel for cat. no. 21 (Fig. 10) comes from Cyrene (Thorn 2005, 638, 765 fig. 410); it has tongues on the top as well as the outside of the rim. Corinthian mould-made mortaria of ‘tile fabric’ with stamped tongue-decoration on the rim have been found also in the Tile Factory in Corinth, probably dating from the first half of the 5th century bc: Hesperia 23 (1954), 130 fig. 2b-c; cf. also Sparkes and Talcott 1970, 12, 370 no. 1913, pl. 91. For cat. no. 22 (Fig. 11) compare Corinthian mortaria with beaded spool-shaped handles from the mid-5th century onwards: e.g. Corinth 15.3, no. 2158, p. 348; Corinth 7.3, no. 622. Ongoing clay analysis of examples of this type of mortarium from Aigina (Fitch Laboratory, British School at Athens) as well as Miletos and Naukratis (Department of Conservation, Documentation and Science, British Museum, and Hans Mommsen, Bonn) should help to establish its origin. The type is widespread in the Mediterranean from the third quarter of the 5th century bc onwards; examples are

‘Drab Bowls’ for Apollo: The Mortaria of Naukratis and Exchange in the Archaic Eastern Mediterranean

14

15

16 17 18

19

20 21 22 23 24 25

26

27 28 29 30 31

32 33 34

known from Athens (Sparkes and Talcott 1970, 370 no. 1913, pl. 91) and from as far north as Torone and as far south as Miletos (see Pemberton and Villing [forthcoming]). Corinthian mortaria and their typology will be discussed in detail by Pemberton and Villing (forthcoming). A general typology of Greek mortaria had already been drawn up by Matteucci 1986, but is in need of refinement, updating and correction. It seems that Corinth specialised in the production and trade of a number of commodities made in its special ‘tile fabric’, notably roof tiles, perirrhanteria and mortaria: Iozzo 1985, 58-9. For recent summaries and discussions with further references, see Bellelli and Botto 2002, 296-300; Berlin 1997b, 123-4; Matteucci 1986. A new study with particular reference to Corinthian mortaria is in preparation: Pemberton and Villing (forthcoming). Oren 1984, 17. Salles 1985, 1991. Cf. also Defernez 2001, 407-8. Hanfmann 1963, 90, refers to customs in modern Cyprus (‘In Cyprus […] peasants are said to use such mortars as milk bowls and for making cheese’) while Tschumi (1931) suspects that thin-walled Roman Terra Sigillata mortaria might have been used for making mild curdle, citing parallels in latter-day Swiss cheese manufacture; cf. also Amyx and Lawrence 1975, 110. Against: Sparkes and Talcott 1970, 222 n. 5; Baatz 1977, 148. Most of the many mortaria at Miletos, for example, show distinct signs of attrition, as do most of the mortaria in the Levant. In Egypt, attrition is reported for the mortaria from Tell Tebilleh (Mumford 2004) but not for those from Tell el-Herr (Defernez 2001, 407). Perhaps the situation in Cyprus (where many mortaria appear ‘unused’: Salles and Rey 1993, 238-9, with note 23) is different on account of most of the published mortaria having been found in tombs, where they might have been placed while relatively new. On traces of attrition from the point of view of use-wear analysis, see Skibo 1992, 115-7, 132-3. A highly detailed analysis of the traces of use on mortaria from Tell Keisan has been undertaken by Sapin 1998. Their Greek names were thyeia or igdys, the pestle associated with them ‘aletribanos’ or ‘doidyx’; cf. Amyx 1958, Sparkes 1962; Dalby 2003, 101; Buchholz 1963, 67. Berlin 1997b, 123-4; Sparkes 1962, 125-6, esp. 125; Moritz 1958, 22 n. 4. Sparkes 1962, 126; Amyx 1958, 239-41. A passage in the Bible, Proverb 27.22 (‘Crush [‘katash’] a fool in a mortar [‘maktesh’] with a pestle [‘`eliy’] along with crushed grain [‘riyphah’], yet his folly will not depart from him’) – presumably dating back to the Archaic or at least Classical period – may refer to wheat being ground to meal in a mortarium, yet given the uncertainty as to the precise meaning of the crucial words, we cannot be sure whether the pounding of grain in a larger holmos (cf. below, n. 34) might not be referred to instead. Another passage in the Bible (Num. 11.8) contrasts the pounding of manna in a mortar (‘maktesh’) with its being ground in a mill. Pemberton and Villing (forthcoming). The appearance of a spout in many examples from the later 6th century bc onwards, moreover, may well suggest a change in use or a diversification of mortaria for different uses. London, BM GR 1903.5-18.3 (Terracotta 957), Corinthian, early 4th century bc. Higgins 1954, 260 no. 957, pl. 135. London, BM GR 1873.8-20.576 (Terracotta 969), Corinthian, early 4th century bc. Higgins 1954, 263 no. 969, pl. 136. Cf. Corinth 12 (1952), 192 no. 1430 (3971), pl. 86. On pestles in general, see Sparkes 1962, 125; Amyx 1958, 239; Buchholz 1963, 67. On ancient cheese-graters, see also below, n. 71; Weber, this volume, n. 30. On these Boiotian figurines as well as on the phenomenon of ‘daily life’ terracotta figurines in general, see most recently Pisani 2003, esp. 6 fig. 5 (figurine of a man grating cheese into a mortarium/basin), pp. 13-4 nos 49-65 (catalogue of figurines with a mortarium or holmos). Fr. 26 West = Ath. 304b; cf. also Anianos fr. 5 = Ath. 282b. On myttotos, see especially Dalby 1996, 107 with note 48. Cf. Bats 1988, 37-8; Dalby 2003, 307 s.v. soup. Cf. also Garnsey 1998, 218-20. In the Attic stelai, holmoi of stone and of wood (and possibly pottery?) are listed: Amyx 1958, 236-8, 282-4. For representations, cf. Sparkes 1962, 1965, and most recently Neils 2004. Note that Schattner (1995, 81-3) suggested that side B of the Apries amphora (Bailey Fig. 1) showed two women working at a holmos, but this interpretation remains doubtful.

35 Marble mortaria existed in Greece, East Greece and Cyprus at least since Hellenistic times in very similar shapes to the typical Corinthian mortaria, with handles and a (mostly ornamental) spout; cf. most recently Korkut 2002, 236-8. 36 Cf. e.g. Artzy and Lyon 2002, 187 (referring to the possibility that mortaria were used as secondary grinders to achieve a finer grain); Matteucci 1986, 250-2. Cf. Foxhall and Forbes 1982, 77, for practical experiments with hulling barley in mortarium (a stone mortarium with a wooden pestle working best). Note also that Galen (De alimentorum facultatibus 1.9. 502, 566) suggests rubbing soaked barley in a mortarium for removing the spelt, while advising against grinding it raw in a mortarium; cf. Darmstaedter 1933. 37 For Archaic Miletos, for example, it has been shown that barley was by far the most important cereal crop, followed by wheat (including Einkorn, Emmer and spelt) and millet: Stika 1997; cf. also Greaves 2002, 24-5. Wheat was, of course, one of the main export staples from Egypt to Greece, even if such trade is not securely attested before the 5th century bc: Möller 2000, 210-11; Austin 1970, 35 with n. 2. On semolina, see also Ath. 1.24, citing Antiphanes (4th century bc); cf. Dalby 1996, 91; Sapin 1998, 111-2 n. 53; Salles 1991, 220. 38 Dalby 2003, 45-7; Darmstaedter 1933. 39 Dalby 2003, 132; Hill and Bryer 1995; Brumfield 1997, 153-4. 40 As argued by Sapin 1998, 110-17. Cf. also Brumfield 1997, 154-5. 41 Cf. Dalby 2003, 349. 42 Perpillou-Tomas 1992; Sapin 1998, 112; Thompson 1995. 43 Braun 1995, 34-6. 44 Mitford 1980, 100-2, no. 133; Salles 1993, 174. Only rim fragments are preserved and no traces of either wear or burning are recorded in the publication of the bowl. 45 It might be tempting to interpret also the inscriptions ‘kri-’ (cat. no. 8) and ‘-nai kr-’ (Fig. 8, cat. no. 19) as somehow related to krithe, barley, but this would be pure speculation. 46 In addition to these mortaria with votive inscriptions, an unpublished mortarium from Shave Zion, supposedly inscribed with personal name in Greek, is mentioned by Stern 1982, 98. 47 Samos, Heraion K11146. H. 7.2 dm, W. 9.4cm, Diam. 35cm. Yellow, relatively porous clay with a yellow slip. I am indebted to Alexander Fantalkin for pointing out this fragment to me, and to Professor W.-D. Niemeier / German Archaeological Institute for permission to publish it. 48 See Schlotzhauer and Villing, this volume. 49 Kron 1984, 1988. 50 Finds from a bothros dated to about 600 bc: Walter and Vierneisel 1959, 28, Beilage 61. Cf. also Isler 1978, 97 no. 146, 159 nos 597-8, pl. 50, Beilage 3 and 19; Furtwängler 1980, 204 no. I/35 (Phase I before 630/20 bc). Almost all the Archaic mortaria from the Heraion are of the flat-based Eastern Mediterranean type. 51 Cf. Kron 1988, 140. 52 Samothrace 51.922 and 51.923: Lehmann 1960, 49 nos 12-13, pl. 3.1213; Lehmann and Spittle 1964, 209-10 nos A6-A7. 53 Lehmann 1960, 14-16. 54 Roebuck 1951, 131, 135 nos 61 (not inscribed), 65, 66, pl. 50. 55 Roman sources for the use of mortaria in the preparation of medicine are collected by Hilgers 1969, 226; cf. also Matteucci 1986, 249. Apollo, too, was of course at times a healing deity; there is no evidence to suggest such a role specifically for Apollo at Naukratis or Didyma/Miletos, although it is attested in Berezan and Olbia (cf. Herda [forthcoming b], chapter 5). In addition, mortaria could be used for preparing paint or mortar (cf. Hilgers 1969, 226), and there are finds of actual mortaria associating them with workshops. 56 This path to death is described as ‘tetrimmene’, so that it is obviously the word ‘tribo’, to grind, that is associated with the mortar, thyeia. 57 V. Karageorghis and O. Masson, BCH 1960, 260 fig. 29. I am grateful to Ursula Höckmann for pointing out this piece. Note also the rim fragment of a 5th century bc stone bowl from Itanos, presumably from the sanctuary on the acropolis, which seems to carry a dedication by a woman: Sitia inv. Ms 10244; BCH 121 (1997), 822 no. C. I am grateful to Didier Viviers and Athena Tsingarida for this information. 58 Kerschner 1997b, 119 no. 10, p. 122 pl. 4.20, pp. 140-3 no. 63, pl. 9.63, pp. 186, 203. 59 Bookidis 1990, 86-94; Pemberton 1989, 67-8. For the types of cakes that would have been produced in the sanctuary, see Brumfield 1997. 60 Several dozens of 7th–6th century bc mortaria as well as cooking pots. 61 Kalaitzoglou (forthcoming). 62 The opportunity is taken here to illustrate a previously unpublished Naukratis: Greek Diversity in Egypt | 43

Villing

63 64 65 66 67 68 69

70 71

72

73 74 75 76 77

78 79

80

81 82 83 84

85 86 87

example of an Archaic mortarium from the recent excavations at Emecik, from the fill at the southern Temenos wall: Eski Datça, Depot inv. ST99K8c-16,78; H. 4.7cm, W. 8cm; coarse light yellowgreen clay, Munsell 2.5YR 8/4, yellow slip. Photograph D. Berges. I am grateful to D. Berges for permission to publish this piece, and to R. Attula for pointing out its existence. One mortarium in a bothros, together with drinking cups, lamps, amphorae and chytrai: Kinch 1914, 99. One 5th century bc (?) mortarium: BSA 64 (1969), 183 fig. 10.Z35. I am grateful to D. Williams for this information. de la Genière and Jolivet 2003, 127, 134 fig. 44.1-2. Gauer 1975, 157-61. Matteucci 1986, 274. Copenhagen, National Museum inv. 10864; Blinkenberg 1931, 748 no. 3229, pl. 152. I am grateful to B. Bundgaard-Rasmussen for providing photographs of this piece. D. Williams furthermore draws my attention also to an andesite millstone from the sanctuary of Aphaia on Aigina that is inscribed ‘A’. This was observed first by Blinkenberg; I am grateful to B. Bundgaard-Rasmussen for confirming Blinkenberg’s observation. Cf. also Homer, Od. 20.105, where the gyne aletris is a female slave charged with grinding barley and wheat. Note also the term ‘aletribanos’ for a pestle (supra, n. 22). There may also have been other specific female roles associated with the processing of grain: on a Boiotian skyphos one of the women pounding (presumably hulling) grain is called Kodoma, ‘toaster of grain’ – possibly a cult name? Women were traditionally in charge of grain-processing and baking (Lewis 2002, 65-71), but on a Campana dinos of c. 530 bc a woman and a nude man are jointly pounding away at a holmos, possibly hulling grain: Boston, 13.205; Fairbanks 1928, 191 no. 546, pl. 58.546. We may also note the presence of a cheese-grater in a Boiotian 4th-century temple inventory: SEG XXIV.361 line 18. On Apollo at Naukratis, see Ehrhardt, Höckmann and Schlotzhauer (forthcoming); on the cult of Apollo at Miletos and Didyma and in the Milesian colonies, see Herda (forthcoming a, b). Cf. also Bîrzescu, this volume. Simon 1983, 76-8; cf. Deubner 1932, 179-201 (esp. 188-90, 198-9). Simon 1983, 77; cf. Nilsson, 1906, 105-15; Deubner 1932, 181-8. Cf. Dalby 1996, 111, 165; Dalby 2003, 70. Cf. also Kearns 1994, 68; Herda (forthcoming a), 397-8 n. 2820. Line 36; cf. Milet 1.3, no. 133; cf. now also Herda (forthcoming a), 10–14, 396–9. A 3rd-century bc poem mentions a rich plakous filled with cheese among the gifts of a four-year-old boy to Apollo on the occasion of his first haircut: Anth. Pal. 6.155. Cf. Liddell – Scott s.v. plakous II; Herda (forthcoming a), 397-8 n. 2820. Cf. Tresp 1914, 159ff., esp. 160 fr. 112; cf. also Herda (forthcoming a), 143–50, esp. 146 with n. 1027, p. 398 n. 2824. I am grateful to A. Herda for his generous advice on this topic. On Hermeias, the prytaneion and Apollo Komaios, cf. also Höckmann and Möller, this volume, ch. 8. Described by the 4th-century bc poet Antiphanes (cited by Ath. 449b-c); cf. Brumfield 1997, 157. Kandaulos, an Ionian speciality of Lydian origin, seems to be another version: Dalby 2003, 188; Dalby 1996, 111. Cf. also Hill and Bryer 1995, 45-7. Cf. Herda (forthcoming b); Höckmann and Möller, this volume, ch. 8. British Museum, GR 1886.4-1.1288 (Sculpture B 469); Nick (forthcoming), 65, 120, cat. no. 43. I am grateful to U. Höckmann and G. Nick for information on this piece. An ongoing ritual significance of mortaria/holmoi at Naukratis (though probably unrelated) is indicated by the foundation deposits of Ptolemy II Philadelphos (282–246 bc) in the four corners of the entrance building to the Great Temenos, which included, among other ‘ceremonial models’, pairs of sandstone corn-rubbers and limestone mortars/holmoi: Petrie 1886bb, 29, pl. 26.32,33,34. A distribution map, albeit now severely outdated, has been published by Salles 1991, 234 fig. 3. C. Grataloup in Goddio and Clauss 2006, 335 fig. 406. Both mortaria are relatively small, especially cat. no. D1 (Fig. 19), for which one should consider Egyptian manufacture. I am grateful to Jeffrey Spencer for pointing out their existence to me and for allowing me to publish them here. Even though Petrie (1891b, 48) himself stated that ‘thick drab bowls’ were ‘wholly unknown in the Greek town of Defenneh’, the two examples published here show that the type existed there after all, even if manufactured in the Nile

44 | Naukratis: Greek Diversity in Egypt

Delta. 88 Spencer 1996, 89, pls 62.48-59; 79.18-19; 81.4-5; 86.15. The mortaria are described as sometimes made from silt clay but more often a friable, beige to green-grey marl fabric; some might well be CyproPhoenician imports, while others may be local imitations. 89 Oren 1984, 17, fig. 21.10, fig. 31. 90 Defernez 2001, 402-11, pls 91-2, nos 253-7. I owe this reference to Sabine Weber. 91 Mumford 2004. I owe this reference to Sabine Weber. 92 Holladay 1982, 109, pl. 16 (from a well dated to 486 bc). 93 Allen 1982, pl. 14 (Stratum IIB-C, attributed to the Late Period) – Egyptian production?. 94 Khonsu, unpublished; mentioned by Defernez 2001, 403. 95 Petrie 1915, 17, pl. 10.1-2. 96 Mentioned by Defernez 2001, 403. 97 Petrie 1909, pl. 54.821 (used as a lid). 98 Brissaud 1990, pl. X.142; presumably of Egyptian production. 99 Petrie and Duncan 1906, pl. 39F.134. 100 Mumford 2004. 101 Defernez 2001, 402-11, pls 91-2, nos 253-7. 102 Oren 1984, fig. 21.7, figs 27-30 (torpedo amphorae); fig. 21.1,3,5,11, fig. 28 (basket-handled amphorae). 103 On this type of mortarium in the Levant and Cyprus, see most recently and comprehensively Sapin 1998; Gal and Alexandre 2000, 190-2; Mazar and Panitz-Cohen 2001, 51; Stern 1982, 96-8; Salles 1985; Bennett and Blakely 1989, 196-203; Berlin 1997b; Lehmann 1996, types 159 and 173; Waldbaum and Magness 1997, 39-40. 104 Ballard et al. 2002, 162, 160 fig. 9.3; Stager 2005, 242. Cf. also Waldbaum and Magness 1997, 39-40. 105 Destruction levels of Sargon (712 bc) and Nebuchadnezzar: Dothan and Freedman 1967, 157, figs 40.10-11, 41.11; Dothan 1971, 140-1, 100, 104, 110. 106 Gal and Alexandre 2000, 189-92, figs VII.11.19, VII.13. 107 Level IV, 8th century bc: Lehmann 1996, pl. 25.161/1; Barnett 1939/40, 107, pl. 52.1 (cf. also p. 88: Level IV covers the 11th–8th centuries bc but contains mostly 8th-century pottery). 108 It may seem doubtful whether the relatively thin-walled bowls from Tyre cited by Lehmann (1996, pl. 25. 159/1, 160/1) are really mortaria. 109 These mortaria, said to be both local and imported, are cited by Hanfmann (1963, 71, 90-1, 187 nos 297-300, figs 64, 119) as coming from the earliest Iron Age levels and thus dating to as early as 1050 bc; this would suggest that flat-based mortaria were produced in Tarsus (and imported there from another centre) for some 300 years before appearing anywhere else. Later mortaria from Tarsus: Hanfmann 1963, 233-4 nos 921-30, 255 nos 1152-3, 274-6 nos 1310-19, 1328-31, pls 79, 93, 132, 137, 143. 110 Salles 1980, 1985; Lehmann 2003, fig. 10.11-14. 111 Karageorghis 1973, 13, 116, 121, pls 47, 51, 225, 233. 112 Karageorghis 1973, 13, 121-2, pls 41, 136, 281. 113 Karageorghis 1978, 13 nos 7-9, pls 7, 44. 114 Salamis tombs 6, 16, 20, 23, 33A, 52, 58, 104, 106; Tsambres und Aphendrika, tomb 23 (5th century bc): RDAC 1937-9, 89-90, fig. 42.11; Salles and Rey 1993, 237. 115 Salles 1985, 200-1; Sapin 1998, 91. 116 Note e.g. the rim variations found in one tomb alone, Salamis tomb 3, dated to about 600 bc: Karageorghis 1967, pls 41, 125. 117 Cf. Salles 1985, 203; Stern 1982, 98; Berlin 1997b, 124 with note 279; Lehmann 1996, pl. 26-7 no. 168; Roll and Tal 1999, 97-8. It remains difficult, however, to date precisely the appearance of the first ringfooted mortaria; few examples seem to date to the 6th century bc, and there is no firm evidence to date them early in that century or even into the 7th century bc. Unpublished ring-footed mortaria were purportedly (Salles 1985, 203) found in Ashdod Phase 2 (late 8th century bc), but one may doubt this with some confidence. The fragment of a ring foot found in Period V (700–650 bc) in Tell Taanach (Raast 1978, fig. 76.5) has been suspected to be from a mortarium (Salles 1985, 203) but the identification is unconvincing. With regard to an example from Stratum 5 at Al Mina (Lehmann 2005, 78-9, fig. 12.4), usually dated to the late 7th– early 6th century bc, the reliability of the stratigraphic information has been doubted. At Samaria, ring-footed mortaria (Crowfoot et al. 1957,130-2, fig. 12 no. 13) appear in Period VIII which is dated 7th–6th century bc, but the actual end-point of the period appears somewhat uncertain (cf. Crowfoot et al. 1942, 115). Also at Lachish (cf. Salles 1985, 203) the situation regarding the ring-footed mortaria in Level II (c. 700–586 bc) is problematic; the majority of the material there seems to be 5th

‘Drab Bowls’ for Apollo: The Mortaria of Naukratis and Exchange in the Archaic Eastern Mediterranean

118 119 120 121 122 123 124 125 126

127 128 129 130

131

132

133

134 135

136 137 138

139 140

century bc or later: Tuffnell 1953, 279-80, class B.14, pls 80.68, 98.568. Otherwise, ring-footed mortaria appear, for example, in Stratum II (Persian period) at Hazor (Yadin 1958, pls 79.25, 83.7; Yadin 1961/89, pls 190.7, 257.3,6,7), in level 3 (Persian period) at Tell Keisan (Briend and Humbert 1980, 122, 147-8), in Stratum I (Persian Period) at Timnah/Tell Batash (Mazar and Panitz-Cohen 2001, 51), and in Stratum Persian 2 (late 6th–mid-5th century bc) at ApolloniaArsuf (Roll and Tal 1999, 110-1 fig. 4.11). Salles 1985, 201; Salles 1983, 73-4; Salles and Rey 1993, 239. Cf. Defernez 2001, pls 91-2, nos 253-7. Cooking pots, which also may have been used at funerary meals and signified wealth, seem to be rarer: Pilides 2005, 178-9. Karageorghis 1973, 13. Karageorghis 1973, pl. 7, 9. Mentioned by Salles 1993, 237. Sapin 1998, 93 with n. 19. For contexts in which mortaria have been found in Greece and Etruria, see Matteucci 1986, 272-4. Other sites in the region have so far revealed no trace of the type, or at least no finds seem to have been published. Flat-based mortarium in the context of a 7th century bc burial at Arkades: Levi 1927-9, 384 fig. 494, p.493 fig. 592-D, p. 498. Two further late 7th century bc examples from Kommos have just been published by Johnston 2005, 359-61 nos 183-4, fig. 24. Kinch 1914, 99, pl. 23.8. A late-7th century bc mortarium from Old Smyrna in the museum at Izmir is reported by Hanfmann 1963, 60 with n. 221. Above, ns 47 and 50. In the vicinity of Miletos, mortaria have also been found at Assesos (7th century bc; Kalaitzoglu [forthcoming)); Didyma (6th/5th century bc: Schattner in Tuchelt 1996, 171-2 nos 105-14 and 129, 182-7 types 47-50 and 63, figs 114-16; Wintermeyer and Bumke 2004, 106 no. S 10.24, fig. 898), and Teichioussa (Voigtländer 2004, 321 no. 134, 325 no. 164, 326 nos 172-4, pls 167, 171-2). The mortaria from Miletos will be published by the author as part of a study of bowls from the site. At Miletos, with its unbroken and extremely numerous sequence of mortaria in both settlement and sanctuary until the end of the 6th century bc, and a scattering of later, Classical finds, the characteristic form of the Cypro-Phoenician buff ring-footed mortarium is not attested. We see instead the emergence in the 6th century bc of a mortarium very similar in shape, but made in the usual Milesian brown clay and painted with white slip and bands on rim and foot. No more than 20 examples are attested in the Archaic settlement at Miletos: Naso 2005a, 77. One example also comes from the area of the Athena temple: Niemeier 1999, 389-92, fig. 20, 407 fig. 29, 412 no. 21. Pers. comm. D. Chistov; I am indebted to Dr Chistov for showing me his material from the recent excavations at Berezan during a visit to London in 2005. The mortaria published in Solovyov 1999, 94 fig. 91 are of a different, probably Late Archaic or Classical type. Two mortaria from earlier excavations at Berezan, again of different types, have recently been shown through NAA to be imports, from Cyprus and the Troad respectively: Kerschner 2006; I am indebted to M. Kerschner for sharing these results with me pre-publication. We can thus no longer assume that at Berezan ‘locally produced pottery fully satisfied the inhabitants’ cooking and food-preparation needs, regardless of their ethnic roots’ (Solovyov 1999, 52). Alexandrescu 1978, 111 no. 721, fig. 27.111; Alexandrescu 2005, 357 no. C156, 394 fig. 47, pl. 68 (wrongly labelled C 157). Boardman and Hayes 1973, 69-70 no. 2317, fig. 2317, pl. 36.2317 (= Bellelli and Botto 2002, 288 fig. 7d [wrongly captioned]). Versions of this type of bowl (‘collared bowl’), which are not necessarily mortaria, are common in Cyrene also in later periods: Thorn 2005, 643-4 nos 211-2. Vegas 1999, 182-3, Form 50, fig. 89. Bellelli and Botto 2002, 279-85; Bartoloni 1992; Gaudina 1994; Bartoloni 1996, 86-7 with fig. 1.3; The spread of this type in the West has recently been charted in great detail by Bellelli and Botto 2002 (updating Matteucci 1986). They consider it, however, together with another bowl of different shape, which somewhat obscures the picture. Bellelli and Botto 2002, 281-5; cf. e.g. Gran-Aymerich 1991, fig. 39. 4-5 (6th–early 5th century bc). Corinth C-40-312 (Weinberg 1948, 228 no. D79, pl. 84; cf. also Bellelli and Botto 2002, 302) and C-73-163 (unpublished, from lot 73-57). Both examples will be fully published by Pemberton and Villing (forthcoming). It has sometimes been thought that the buff-clay

mortaria of Eastern Mediterranean type are in fact of Corinthian manufacture. This must, however, be ruled out, not merely for the simple reason of the absence of the type at Corinth (bar the two examples jus cited), but also in the light of the recent clay analyses from the Levant, discussed infra. 141 Sparkes and Talcott 1970, 369 no. 1889, pl. 90. Elsewhere in mainland Greece, early pottery mortaria follow completely different traditions: cf. Cavanagh et al. 1996, 73-4 with refs. 142 Cf. Fantalkin, this volume. 143 In the last quarter of the 7th century bc as much as 75-85% of total imports to Corinth may be of East Greek or Island origin, and also some Syro-Phoenician / Palestinian imports are attested: Bentz, 1982, 126-8, 218-41; Siegel 1978, esp. 64-217. 144 Amyx and Lawrence 1975, 91-5; Weinberg 1948, 228 no. D78, pl. 84. 145 Buchholz 1963, 1976/7; Botto 2000. 146 Stern 1982, 96-8. 147 Stern 1982, 98. 148 For a summary of Petrie’s views, see Blakely and Bennett 1989, 50-1. 149 Cf. Niemeier and Niemeier 2002, 240-1, figs 5.96-7, and now also Fantalkin, this volume. 150 Hayes 1967; Blakely, Brinkmann and Vitaliano 1992; cf. also C. Grataloup in Goddio and Clauss 2006, 348 fig. 484. 151 Supra, n. 104. 152 Bennett and Blakely 1989, 198-200, pocket insert 23; Blakely and Bennett 1989, 55-7. 153 Sapin 1998, 97-103. 154 Berlin 1997b, 10-12, 123-32, esp. 124 n. 277; Rautmann 1997, 223-4, sample 10. 155 Mazar and Panitz-Cohen 2001, 19-20 FG 13, 51 BL 20. 156 Goralczany 1999, 186 table 4.10.7-11; and id., 2005. Goralczany’s results from Yaoz and Tel Michal are as yet unpublished. I am indebted to A. Fantalkin for this information. 157 Fantalkin 2001b, 79-82. 158 Analysis by Yuval Goren: Lehmann 2002b, 205 fig. 5.78.14 (Kabri 93/1947/5216-100). 159 Artzy and Lyon 2002, 186-7, fig. 3. 160 Yellin and Artzy 2004, 225-6, fig. 2.6-8. 161 Mentioned by Salles 1993, 237 n. 16. 162 Salles 1985, 202. 163 Cf. most recently Markoe 2000, 53; cf. also Docter 1997, 28-31. 164 Master 2001, 176. 165 Master 2001, 110 no. 50.58 L318 (49), 111 no. 50.58 L318 (50), 134-5, 137-8, 154 fig. 2.8.2, 4, p. 176. 166 Master 2001, 141-2, 155 fig. 2.9.5. In this category falls also an amphora with yellowish clay. The clay is said to be very similar to the clay of Category 13 (‘highly micaceous samples with reddish brown fabric’), which are all Greek imports, probably from the region of Miletos or Samos. 167 Master 2001, 114 no. 50.58 L318 (7). 168 Master 2001, 135; cf. also Master 2003. 169 Bennett and Blakely 1989. 170 See Ballard et al. 2002, 162-3; cf. also Bennett and Blakely 1989, 196203; Berlin 1997b, 123-32. Perhaps the mortaria travelled around the Mediterranean alongside not just Phoenician and Cypriot oil and wine, but also Phoenician semolina and wheat from Cyprus; note the mid-6th century bc Ionian poet Hipponax (fr. 125 Loeb) mentioning ‘those who ate the bread of Cypriot and Amathusian wheat’. 171 Salles 1991, 225-31; Humbert 1991; Gunneweg and Perlmann 1991; Niemeier 1999, 389-92; Bennett and Blakely 1989, 212-3. 172 For speculation on East Greek production and trade of mortaria, see e.g. Bellelli and Botto 2002, 305. A more detailed report on the origins and trade of mortaria is planned for a future study, which will include the petrographic clay analysis of mortaria from Naukratis, Miletos, and Al Mina. 173 Mommsen, this volume. 174 Mommsen, Beier and Aström 2003, 6-8 no. HST 7, 10 table 2. 175 Kleibl 2006, esp. 154-5. 176 Eski Datça, Depot inv. ST00K8c-20,3, sample Emec 122: Attula 2006, 112 cat. no. 202a, pl. 69.3; Mommsen, Schwedt and Attula 2006, 200, 204 table 3. One further mortarium (cat no. 9 [Fig. 6], sample Nauk 56) remains a chemical single, another (cat. no. 17 [Fig. 23], sample Nauk 18) was identified as Egyptian marl (see below). A further mortarium in the group EmeA was found at the northern extremities of the Greek world, at Berezan (sample Bere 025), showing just how widely Cypriot mortaria were traded: Kerschner 2006, fig. 22. 177 On Levantine pottery in Iron Age Egypt, see e.g. Maeir 2002. He lists mortaria in his catalogue of Palestinian pottery found in Egypt (pp. Naukratis: Greek Diversity in Egypt | 45

Villing 238-9), thus disregarding the abundance of analyses indicating their status as imports also to Palestine. 178 The Ma‘agan Mikha’el wreck, for example, carried Cypriot, Phoenician and East Greek pottery; cf. Artzy and Lyon 2002; Yellin and Artzy 2004. 179 Cf. Maeir 2002. For Jews and Phoenicians serving as mercenaries in Egypt, see also Kaplan 2003, 7-9. Cf. also Vittmann 2003, 44-83 on Phoenicians in Egypt, and 84-119 on Aramaic documents from Egypt, and Holladay 2004 on the Judaean diaspora in Egypt. On SyroPalestinian pottery, see also Gjerstad 1948, 241, 242; Petrie 1888, pl. 3. 180 On Cypriot mercenaries in the service of Egypt in the Archaic period, see Kaplan 2003, 10; cf. also Fantalkin, this volume, n. 75. 181 As argued by Carrez-Maratray 2005, they may at this period have docked at a number of ports in the Nile Delta. 182 Small numbers of mortaria have been found in a number of ship wrecks. In addition to the Elissa and the above-mentioned Ma‘agan Mikha’el wreck one might list the late 6th century bc Pabuç Burnu wreck (Greene 2003; the ship carried Samian, Milesian and other East Greek transport amphorae), the third-quarter-5th-century bc Tektaþ Burnu wreck (on display in Bodrum Museum; cf. Carlson 2002; the ship carried mainly East Greek transport amphorae and appears to have traded along the Ionian coast), the Giglio wreck (Bellelli and Botto 2002, 296, 288 fig. 7e; Bound 1991, 224, fig. 59; Cristofani 1996, 43-5, fig. 14.2), the mid-6th century bc Antibes wreck (Bellelli and Botto 2002, 296-7, fig. 10a; Bouloumié 1982, 34 nos 300-301, fig. 9; the Etruscan (?) ship carried mostly Etruscan but also East Greek pottery presumably destined for Marseille), the wreck Pointe Lequin 1A (Long, Miro and Volpe 1992, 219, 221 fig. 37.1, 4), the early-5th-century bc Gela wreck (Panvini 2001, 30 fig. 32, 44-5 cat. nos 36274 and 36275, 54-5 cat. nos 38/85 and 36345, pls IV.23-4, XL.65-6), and the Late Classical Porticello wreck (Eiseman and Ridgway 1987, 31 no. G10, 32 fig. 3.9-10). Again, in all cases, mortaria seem to have been part of the galley kitchen equipment rather than cargo – raising the (albeit remote) possibility of mortaria found in sanctuaries having been dedicated after a successful journey. 183 Ballard et al. 2002; Stager 2005. Stager believes that the destination of the Phoenician Elissa with its cargo of wine may have been either Carthage or Egypt. Egyptian pottery is found occasionally in the Levant through the 7th and 6th centuries bc; in the late 7th century bc, of course, the Phoenician coast even found itself under Egyptian rule. Cypriot trade amphorae, and occasionally fine wares too, also reached the Levant. Cf. Fantalkin 2001b, 97-8; Fantalkin, this volume; Markoe 2000, 44-7; Vittmann 2003, 44-83. 184 On Phoenician evidence see Villing and Schlotzhauer, this volume. 185 For a summary of Cypriot material at Naukratis, see Möller 2000a, 161-3. Gjerstad (1948, 241) cites two pieces from Naukratis in the Cairo Museum: ‘a Black-on-Red III (V) globular miniature hydria with base-ring, neck widening upwards, swollen rim, horizontal handles on the belly, a vertical handle from below rim to shoulder, and decorated with encircling lines around shoulder and belly’ (Cairo Museum C 3132), and ‘a Black-on-Red II (IV) handle-ridge juglet with funnel-shaped mouth and flat rim; encircling lines around lower part of neck; upper part of neck and rim covered with mat, red paint’ (Cairo Museum C 3133). A Cypro-Archaic II small feeder-jug with eye decoration in Oxford, Ashmolean Museum 1987.62, is also supposedly from Naukratis (gift Peter Fraser; I am grateful to Helen Whitehouse for information on this piece). The unpublished imitation of a Cypriot Black-on-Red I (III) juglet in the collection of UCL that is mentioned by Davis (1979, 16-7, and 1980, 7) is in fact of faience (Gjerstad 1948, 411); its findspot, however, seems to be uncertain. As regards other types of material, a bronze bowl (Petrie 1886bb, pl. 12) that may be of late 7th / early 6th century bc Cypriot production was kindly pointed out to me by U. Höckmann. 186 Cf. Höckmann and Möller, this volume; Schlotzhauer and Villing, this volume. 187 Such as, e.g. BM GR 1888.6-1.100 (Terracotta C586). 188 See supra n. 3. 189 Johnston 1982, 36. 190 Cf. e.g. Peserico 1999; Peserico 2002; Bartoloni 1992; 1996; Gaudina, 1994.

46 | Naukratis: Greek Diversity in Egypt

191 Mommsen et al., this volume. I am grateful to Jeffrey Spencer for confirming the visual identification as marl clay. Marl clays must have been available on the desert fringes of the Nile Delta, and thus at least in the wider area of Naukratis. On Egyptian Nile and marl fabrics, see Arnold and Bourriau 1993; Bourriau, Smith and Nicholson 2000. 192 Mommsen et al., this volume; Schlotzhauer and Villing, this volume. 193 Cf. e.g. Holladay 2004. 194 Mumford 2004. Very different, local, carinated mortaria shapes seem to be found among the materal from Tell El-Herr from the Persian period onwards: Defernez 2001, 288-93, pl. 65. 195 Cf. supra, n. 85 196 Cf. Oren 1984, 27; Weber 2001, 142 with n. 115 (not all the pieces listed by Oren are actually local). One example of an Egyptian imitation of a Samian trade amphora is British Museum EA 22333; I owe this information to S. Weber. There are no certain Greek cooking pots at Tell Defenneh: a kind of chytra (Petrie 1888, 64, pl. 35.43; present whereabouts unkown) was identified by Petrie as of likely Greek origin, but has been described as reminiscent of late Iron Age Judean cooking jugs by Maeir (2002, 239, 236 fig. 1.5); although no certain identification can be made based on the drawing, the fact that Petrie mentions the existence of similar double-handled vessels might indeed point to a Judean rather than a Greek type. For the related phenomenon of both imported and locally-produced (marl) ‘Judean jugs’, presumably manufactured by Judean potters for a Judean diaspora community, see Holladay 2004. 197 A small fragment of similar shape and similar-looking clay and slip has recently also been found at Miletos; its clay analysis, however, is inconclusive (Sample Milet 41; Mommsen et al., this volume). 198 Cf. e.g. Tite and Kilikoglou 2002. 199 Cf. Hiller 2000, 467, with pp. 497-9 (discussion by P. Alexandrescu, S. Hiller and P. Dupont). For the modern Siphniot parallel, see Jones 1986, 861-4. 202 Waldbaum 2006, 65; Master 2001, 160-71, esp. 167; Master 2003; contra Fantalkin, this volume. 201 Master 2001, 220-1; for the possibility of the development of specialised production centres for mortaria see also Bellelli and Botto 2002, 299. 202 On Corinthian clays and the manufacturing of mortaria, cf. Sparkes and Talcott 1970, 222. 203 Sparkes and Talcott 1970, 370 nos 1899-1911, fig. 20, pl. 90-1. 204 E.g. Voigtländer 1982, 79 fig. 37.230, 144 no. 230, pl. 27.1. Further Corinthian mortaria have also been found in the recent excavations in the Archaic settlement at Miletos, to be published by the author. Cf. also Voigtländer 2004, 325 no. 164, pl. 171 (Teichioussa). 205 Alexandrescu 1978, 111-2 no. 729, fig. 27; Alexandrescu 2005, 357 no. C157, 394 fig. 47. 206 Thorn 2005, 638 nos 184-5, figs 377, 410. 207 The type of the ‘collar rim’ mortarium is attested, for example, at Velia: Gassner 2003, 97, fig. 41. 208 The mortaria of Euhesperides have been studied by Keith Swift, to whom I owe this information. For a preliminary report, see Swift 2003, 219-20. 209 Swift 2003, 215. Compare also the evidence from Late Archaic– Early Classical Velia, where c. one fifth of coarse-ware is imported (mortaria from Calabria and Corinth): Gassner 2003, 220-4. I owe this reference to Michael Kerschner. 210 Swift 2003, 219-20, and pers. comm. 211 Oren 1984, 8-9, has stressed the position of Northern Sinai and the Eastern Delta as a densely populated region of commercial, industrial and military importance that linked Egypt with Canaan. On connections between Egypt and the Levant, see also Maeir 2002. 212 An observation made already by Lehmann 2005, 74. 213 The prominent role played by Phoenician traders in the Iron Age is underlined by Fantalkin, this volume; see also Docter 1997. 214 Blakely, Brinkmann and Vitaliano 1992, 208. 215 Blakely, Brinkmann and Vitaliano 1992; Hayes 1967. 216 Cf. Baatz 1977, 154-5. 217 As already suggested by Baatz 1977, 155.

Carian Mercenaries at Naukratis? Dyfri Williams and Alexandra Villing

Abstract Two fragments of pottery from Naukratis in the British Museum are identified here as probable products of Caria. As such, they are the first examples of pottery from Egypt to be associated with Caria, and they raise the possibility of the presence of Carians and in particular Carian mercenaries at Naukratis. As research progresses on the wealth of pottery excavated by Flinders Petrie, E.A. Gardner and D.G. Hogarth at Naukratis, there are bound to be surprises. This article publishes two pottery fragments that certainly fit the bill, for they raise the unexpected question of Carians at Naukratis. The first fragment (Fig. 1) comes from the wall of a krater.1 The clay contains much golden mica, white and black grits, and has a pinky red core. The exterior is covered with a thin, whitish wash; the interior is covered with a dull black slip. The decoration is done with a yellowish brown paint and is organised in panels. On the right, there is a panel of vertical chevrons, framed by a pair of vertical lines on each side (on the extreme right another line at an angle to the vertical suggests an unidentified part of another motif). On the left a cock with three drooping tail feathers (rather than a horse) is depicted in silhouette to the left; on the extreme left there is a blob. The style of this fragment seems to combine Geometric, or rather Subgeometric traits with Orientalising ones. The date might be somewhere in the second half of the 7th century bc. Neutron Activation Analysis carried out on the fragment by Hans Mommsen has linked it with a Wild Goat oinochoe in Bochum attributed by Mommsen and Schlotzhauer to a Carian workshop.2 Furthermore, the fabric of the fragment seems to compare quite well by eye with the Carian Geometric pieces in the British Museum, excavated in tombs at Asarlik by W.R. Paton in 1887, but it is to be hoped that more Carian material will be analysed in due course and the attribution to Caria further supported.3 Carian Orientalising pottery in a local Wild Goat style has been recognised and studied by a number of scholars since the discovery in the early 1970s of a necropolis near the village of Damlýboðaz near Mylasa, when the bed of the River Sarýçay was

diverted in connection with the construction of a dam.4 The Naukratis fragment has points of contact with some of these pieces, including the panel of vertical chevrons, which was also frequently used as a horizontal band in the Carian style,5 but we have not been able to parallel the cock among the published material. Further research and the appearance of additional publications may shed more light on this fragment and its likely Carian origin. The second fragment is from a closed vessel, probably an amphora, made from a coarse, brown clay with white grits (Fig. 2).6 The decoration is divided into three zones. There is a finely drawn set of six concentric black circles. A zone of elaborate banding takes the form of two reddish-brown bands between lines, with a wavy line between them. Finally, there is a panel of feathered, or dotted, multiple strokes. This fragment has not been analysed by Neutron Activation but the colour and consistency of the clay and the decorative motifs and scheme can only be paralleled on fragments of a fabric found at Kaunos and Damlýboðaz which has been studied by Bernhard Schmaltz (Fig. 3).7 Sadly, there are no clues from the stratigraphy at Kaunos as to the chronology of this fabric, which is known in the form of fragments of dinoi or kraters, amphorae, cups, plates and dishes. Schmaltz, however, has posited the idea that it is essentially 7th century bc in date.

Figure 1 Fragment of a krater, BM GR 1888.6-1.653

Figure 3 Fragment of an amphora from Kaunos, Kaunos PT – Ö – 9/8/98

Figure 2 Fragment of an amphora (?), BM GR 1924.12-1.37

Naukratis: Greek Diversity in Egypt | 47

Williams and Villing Both the above fragments are the first examples of their fabrics to be found outside Caria. It seems unlikely, therefore, that we are dealing with normal examples of the export of pottery for commercial reasons, as must be the case with so much of the East Greek pottery found at Naukratis. The Carians were, however, well known as mercenaries in Egypt and the Near East.8 Greek mercenaries were a significant presence in the Orient. Herodotus (2.152–4) and Diodoros (1.66.12–67.2) report that Psammetichos I (664–610 bc), the last of the Saite pharaohs, employed Carian warriors. He is said to have settled them later in ‘Stratopeda’ in the eastern Delta, on the Pelusiac branch of the Nile. From Polyainos (Strategica 7.3) we also learn that when Psammetichos overcame Tementhes in a battle at Memphis the Carian Pigres was his advisor and that he had many Carian mercenaries. Indeed, it has been suggested that a grave-stele of Pigres found at Memphis is the same Pigres as mentioned by Polyainos.9 Carian mercenaries are also attested in Lydia, especially under Gyges, who may well have been responsible for sending Carian mercenaries to the aid of Psammetichos. Others may also have served in Assyria, Judah and Tyre.10 The archaeological evidence for such mercenaries is, of course, slight and particularly open to question: it is also regularly unspecific as to the home city of the mercenary. Thus, a Greek bronze greave and shield found at Carchemish have been connected with a Greek (or Carian) mercenary in the pay of Necho at the time of the city’s conquest by Nebuchadnezzar I of Babylon in 605 bc.11 Similarly, a fragmentary silver bowl of Cypro-Phoenician workmanship, found at Amathus on Cyprus, probably shows a city-siege with Greek (or Carian) mercenaries in both armies.12 Two recent studies have looked at the remarkable series of grave stelai of Carians from Saqqara, which may be associated with the mercenary presence.13 Furthermore a pair of bronze objects found at Sais bear Carian and hieroglyphic inscriptions.14 Less formal Carian inscriptions include graffiti on the statue of Ramases II at Abu Simbel, in the tomb of Mentuemhet, the governor of Thebes in the time of Psammetichos I, and in the temple of Seti I at Abydos.15 In addition, attention has recently been drawn to a brief graffito on a small oinochoe with ribs round the neck that is made of local Egyptian clay (Schlotzhauer and Villing Fig. 36).16 This oinochoe, now in Berlin, has no precise find-spot, only the Nile Delta, but it may well have come from a tomb at Saqqara. The graffito has been read by Masson and Yoyotte as naming one Mik(k)ylos, but Kammerzell’s transcription system would perhaps produce something like Pyhra.17 Other possible Carian graffiti were reported by Petrie on transport amphorae found at Tel Defenneh, but there is no way to be sure that any of them are Carian rather than Greek.18 The identification of one of Herodotos’ Stratopeda with Daphnae/Tell Defenneh has been questioned by many scholars. Jeffrey Spencer, however, has argued recently and convincingly that the balance of the evidence points to the tower-like brick

48 | Naukratis: Greek Diversity in Egypt

structures at Tell el-Balamun, Naukratis and Daphnae all being fortified camps.19 They were not elaborate forts, like the site of Migdol in Sinai, on the edge of the Delta Plain, which perhaps accommodated a variety of mercenaries, including Greeks, Phoenicians and Jews.20 They were rather, as Spencer writes, ‘fortified barracks, capable of serving the dual purposes of watchtower and redoubt.’ It is clear that East Greeks were involved in Egypt both by way of trade and as mercenaries. The Carians, however, seem to have been mercenaries par excellence. The discovery of two fragments of Carian pottery at Naukratis could suggest a number of scenarios. Perhaps the fragments indicate the presence of Carian mercenaries who showed their gratitude to the gods for their safe arrival at the nascent port of Naukratis. They may even have been stationed there by the Egyptians as some sort of trusted police force, tasked with keeping order in such a bustling environment full of foreigners. Or perhaps some Carian mercenaries retired to Naukratis after service elsewhere in Egypt and dedicated their last remaining pots from home in a Greek sanctuary, as they took wives and settled down. Notes 1

2

3 4 5 6 7

8

9 10 11 12 13 14 15 16 17 18 19 20

British Museum GR 1888.6-1.653 (sample no. Nauk 66): greatest width 7.3cm; greatest height. 5.0cm; thickness 0.6cm. The fragment is Mommsen, Nauk 66, see Mommsen et al. in this volume. The Bochum oinochoe (inv. S987) is Mommsen, Kari 2. For the Bochum oinochoe see further: Cook 1993, 112-3, fig. 7; Cook and Dupont 1998, 64 fig. 8.26; Cook 1999, 80, list B no. 8. Paton 1887, 68-71. On early Carian pottery see also Özgünel 1979. Carian pottery is currently the subject of a research project by U. Schlotzhauer, Ý. Fazlýoðlu, and H. Mommsen. Gercke 1981; Hemelrijk 1987, 33-55; Lenz 1997, 29-61; Cook 1998, 636; Fazlýoðlu (forthcoming). Vertical chevrons: Gercke 1981, nos 13 and 27; Lenz 1997, no. 3. As a horizontal motif: Lenz 1997, nos 6 and 7; Gercke 1981, nos 1, 8, 11, 12, 27, 35; Hemelrijk 1987, 43, figs 19-21 (Amsterdam APM 10189). British Museum GR 1924.12-1.37: greatest dimension 10.0cm. Schmaltz 2003, 37-42. The fragment illustrated in Fig. 3 is no. 16 on p. 40, fig. 5, 7. We are very grateful to Bernhard Schmaltz for permission to publish it here and for generously sharing his findings in Kaunos with us. For mercenaries in Egypt see Haider 1988, 153-211; Laronde 1995; Möller 2000a, 33-8; Haider 2001; Kaplan 2003, 1-31; Höckmann and Vittmann 2005. For Egypt and the Near East see, for example, Haider 1996, 95-115; Niemeier 2001, 16-19; and Alexander Fantalkin in this volume. Brussels E2483: Kammerzell 2001, 240-1, and 246 fig. 2-1. Niemeier 2001, 19-21. Boardman 1999a, 51 and 115; Niemeier 2001, 19-20. Boardman 1999a, 50, fig. 19; Niemeier 2001, 21. Höckmann 2001b, 217-32; Kammerzell 2001, 233-55. See also Vittmann 2003, 161-79. Vittmann 2003, 160-2, pl. 19 c and fig. 75. Kaplan 2003, 6-7. Schlotzhauer 2006, 308-11. Masson and Yoyotte 1956, 12-3; Cf. the chart, Kammerzell 2001, 245. Petrie 1888, 64 with pl. 33. We are grateful to Sabine Weber and Alan Johnston for their comments on these pieces. Spencer 1996, 56-8. On Tell Defenneh see also Sabine Weber in this volume. For the Naukratis structure, which was clearly renovated in the Ptolemaic period, see Möller 2000a, 109-13. For Migdol see Oren 1984 and 1993.

The Study of East Greek Pottery John Boardman

It is singularly appropriate that a conference devoted largely to East Greek pottery should be held in the British Museum, rather even than somewhere in the Mediterranean world. The study of Greek vases, especially of the Archaic period, is becoming more and more one of excavation and modern techniques in the evaluation of finds, but many of the basic techniques of analysis are those which have been practised and refined in museum environments. The pottery resources of this museum, not least those from Naukratis, bring us away from the pottery sheds and laboratories, back into a more scholastic environment. Here, the pots are objects requiring study and understanding individually, per se, as well as in relationship with the whole range of relevant material in other museums and from excavation. My paper breaks no new ground, but reflects a little on the study of East Greek pottery, where it has gone and where it is going. It is a lecture which should have been given by Robert Cook, who died in the last year of the last century. To his sharpness, in every sense of the word, Greek pottery studies owe a great deal. His scepticism was usually more positive than negative, and we might do well to be equally and more often critical of our own ideas and sources. The Aegean Sea has been a great comfort to students of the history and archaeology of ancient Greece. It seemed to effect both a geographical and a cultural divide between the arts of the Greek so-called homeland and of the coast and islands of Asia Minor, and for a while one might conveniently forget that the heart of ancient Greece was the Aegean Sea itself, even more than the Balkan mainland. For the historian, the east was an area where Phrygians, Lydians and Persians dictated events, where merchants and tyrants flourished, where there were very special forms of lyric and epic poetry, and where Herodotus, an East-Greek-cum-Carian was the authority. It was all just a little foreign to scholars brought up on Athens and Corinth and Sparta and Olympian ideals. For the archaeologist, what was, generically, East Greek, seemed to be readily definable and distinguishable from the mainland products in almost all areas: it was more broadly Ionian than Dorian, it was Wild Goat not black-figure. This was a comfortable situation but fraught with danger, and the overall uniformity of much East Greek production in pottery, exemplified by the Wild Goat style, probably led to a more casual attitude in definition of its history within the East Greek world itself. It has taken a long time for the situation to be rectified, and we are still some distance from the sort of solution which, in the homeland, enables us to be quite confident in identification of most local wares, their development and dating. In fact, this meeting is exceptional in concentrating on this analysis and its products, since many pottery studies have moved into areas so far removed from consideration of the pottery itself as to seem unreal, if not selfindulgent. We may take comfort in our more traditional approach and what it may offer of archaeological, art-historical

and socio-historical importance. Robert Cook’s account of the history of the study of Greek vases in his Greek Painted Pottery (its first edition – Methuen, London – was as long ago as 1960) is nowhere more exciting than where he describes the way scholars came to assess and understand the pottery of the East Greek world. He reveals a catalogue of disastrously wrong assumptions over more than a century of study. These were sometimes based on superficial observation of similarities to mainland Greek wares, whose dating was only slightly better understood, sometimes on almost wilful disregard of historical evidence for date. The way that much was allowed to depend on the accident of discovery is understandable, though we are most of us more cautious nowadays. There was also the handicap that the coastal sites were not in Greek territory but Turkish, and little had been done there since the 19th-century expeditions, which were to the major architectural sites, rather than to Archaic ones which seemed to promise less substantial returns than things like the Mausoleum, Assos and Pergamon Altar sculptures. There was also one ideological argument which proved attractive to many and that seems to have carried much weight, and that was the expectation that Ionia, which could be deemed to be the mainspring of Greek thought, literature and the higher arts, especially if one placed Homer on Chios, must also have been supreme in pottery decoration, which was demonstrably of importance in Athens and Corinth. This was not a position that was easy to sustain except by assuming very high dates. The ideology was Panionism, which seemed almost a cult and was particularly flourishing a hundred years ago. Cook was very scornful of its effects on the study of East Greek pottery, forcing impossibly high dates and ignoring influence from the mainland, indeed preferring to judge that even in pottery mainland Greece learned from East Greek models. Cook was right to be scathing, but perhaps now we need to revisit the question, even à propos of the pottery. Here I would certainly have reservations, since in it we can, here and there, detect an echo of greater things in other media in the East Greek world, notably wall painting, which has only become known from western Asia Minor in the last 50 years and may have been a legacy of Anatolia’s past. There is also the strong possibility that what happened to Athenian pottery in the second half of the 6th century bc was due in no small degree to influence from the East Greek world, even if not always from its potters. Jackson wrote well about such matters nearly 30 years ago, and probably did not go far enough.1 We may well still wonder where Lydos, the Andokides Painter or Amasis came from, and observe the influence of patterns and shapes. Francis Croissant’s work2 on the finer terracotta heads of Ionia has given good grounds for believing in direct Ionian influence on mainland sculptural and drawing styles, beyond what seemed obvious from the marble sculpture. Indeed, the debt in sculpture, not least in the korai, is Naukratis: Greek Diversity in Egypt | 49

Boardman clear enough and well acknowledged, as it is in architecture, although the mainland was slow to adopt the florid Ionic and was in no position to compete with the colossal Ionic temples of the east, backed by eastern gold. We have learned much in the last 50 years. Looking back at the earlier history from an archaeological viewpoint, we see now on the Asia Minor coast and islands that there is evidence for substantial occupation in the Bronze Age by Minoan non-Greeks and Mycenaean Greeks. The so-called Ionian Migration has to be viewed very differently now, less perhaps as a migration and more as something like an East Greek parallel to, as well as reflection of, that regeneration of Greek culture in the early Iron Age which occurred also on the mainland, followed by some movement of peoples. The main difference is that in the east there was also a local Anatolian population which seemed mainly indifferent to what happened on its western coast. Can we even be sure that the cultural break with the Bronze Age was as complete as it seems to have been in mainland Greece? À propos of the Anatolian peoples, in many ways, and especially in pottery studies, we may need to begin to view the relationships of East Greeks with Carians, Lydians or Phrygians, as not unlike those of Greeks with Etruscans and Phoenicians in the west. Should we not look for give and take – think of the bucchero in the north and Rhodes – and not assume automatically that non-Greeks can only be the learners in these matters, or that the East Greek world was as isolated culturally and geographically as was mainland Greece in the early Iron Age? In the Geometric and Archaic periods it is clearly a mistake to judge from pottery quality alone. This may take us far in mainland studies, but gets us nowhere farther east. East Greek pottery always looked poor stuff, because it lacked the many figure and narrative scenes of Corinth or Athens, but we know now that the strength and originality of East Greek narrative art was expressed in higher crafts than pottery. It was expressed on relief metalwork, such as appears in Samos and is identifiable by style in Olympia, and almost as certainly in major painting, but we have only come to know about the paintings at Gordion and Elmalý in the last 50 years, or found an East Greek painting on a wooden panel in Egypt. When it comes to direct dealings with the foreigner, other than colonial, East Greeks were on the heels of mainlanders in the east, at Al Mina from the start of the 7th century bc, and before them in Egypt later in the century. Samos shows clearly enough that it was as closely in touch with Near Eastern, Black Sea, Syrian and Cypriot arts and techniques as Olympia or Corinth, even Crete. Perhaps there still remains a case for a measure of Panionism, given the achievements also in sculpture, architecture, literature and thought, as well as the special relationship with the east, which the East Greeks seem to have taken up where the mainlanders left off, with the Black Sea, and with Egypt. It still will not make our pottery any more intrinsically important as a source for cultural history, except for the basic archaeological problems of defining places, people and trade, while for trade it is the plain carriage wares that may be the more important, although they are also the pots that travelled as much on account of their contents as for the folk who made or used them. There were other problems introduced by early pottery studies in this area. The analysis of East Greek pottery in terms of Early, Middle and Late follows a pattern set by the study of 50 | Naukratis: Greek Diversity in Egypt

prehistory. I am not sure that it has been particularly helpful, and further, closer definition on the same lines, even by Cook, leaves one with, for instance, a type of Middle Wild Goat that overlaps Late Wild Goat in different places. The scheme seems to imply that each phase has a certain entity and implies a cycle of growth and decline, while what happens historically may indeed involve a genesis, but thereafter simply involves change, for different reasons in different places. Thus, elsewhere in Greek pottery studies, it is easier to define ‘Late’ Corinthian pottery than to distinguish between ‘Early’ and ‘Middle’ Corinthian. We have come to understand ‘Late’ Protocorinthian more in terms of its painters not its style. ‘Late’ Protoattic is virtually the same as ‘Early’ black-figure. Dr Schlotzhauer’s paper here shows us, I think, that it is time to forget ‘Middle Wild Goat’ subdivisions and use group designations that can be justified in terms of date, style, geography and analysis. He is proposing a new classification, with Michael Kerschner;3 the only danger, I think, may be over-complication. The old system tended to distract attention from other questions. How and why can a style be said to decline? Is it an internal matter to be explained by some sort of psychological study of the painters and potters; how can a style be described or explained as ‘exhausted’; how and why does ‘taste’ change; what part, and why, is played by the market, and is it the home or export market? We have been spared any general use of an Early, Middle and Late classification for Athenian black-figure, largely thanks to the analytical work which started over a hundred years ago in presenting the history of the ware in terms of workshops and individual potters and painters, their associations with each other and with their market, at home and abroad. This is historically far more useful. Does it need to be applied even more rigorously by now to the East Greek world? British pottery studies of the 1930s were inspired by Beazley’s work on Athenian vases and the application of Morellian analysis to determine hands, combined with traditional analysis in terms of shape and general decoration, which had been developed mainly by prehistorians, who had often nothing more sophisticated to work from, and from observation of stratigraphy and historical dating points where there were any. It started with Humfry Payne, from Oxford, and his book Necrocorinthia (Clarendon Press, Oxford, 1931). In Athens, under Payne’s wing, Robert Cook, John Cook and Arthur Lane set to the task, with Fikellura, Protoattic, and Lakonian black-figure respectively. In fact Robert had been intended to deal with Lakonian, but something changed his mind, and he was warned off Wild Goat by Payne who was reserving it for someone else to study. With characteristic caution, not to say scepticism, Robert was the only one who tended to shrink from assigning artists and stuck to groups, whereas his brother John, Arthur Lane, and Humfry Payne, were happy with the individual. I make no special claim for British achievement here with the non-Athenian, since in the 1920s Andreas Rumpf had already sorted the so-called Chalcidian black-figure vases into their painters with the greatest skill. But this was to be the way forward. This approach has not been much used in the study of East Greek pottery except for the later manifestations, in Chian, Fikellura and black-figure, and eventually, by Robert Cook himself, on Klazomenian vases and the sarcophagi, which he called ‘these deplorable monuments’ or ‘these ungainly

The Study of East Greek Pottery creations’, soon after writing a whole book about them. These are areas where hand and workshop can be readily distinguished because the painting is detailed. Surprisingly, though giving full credit to Beazley’s skills in his book, Cook nowhere describes the Morellian process of analysis which is what he practised, and which is generally agreed to be as near scientific as one can hope for in such studies,4 and applicable, with care, even to non-figurative vase painting and anything as repetitive as Wild Goat. It could be objected that too much East Greek pottery, mainly the Wild Goat, looks alike, that one wild goat is much like any other, while subsidiary decoration is repetitive. But this is defeatist and all that is called for is closer perception of drawing styles and conventions, as well as principles of composition. It is surprising that more of this had not been done long ago, but encouraging to discover that, hand-in-hand with clay analysis, the role of style in East Greek studies is reviving strongly. The broad styles distinguished by Wolfgang Schiering long ago, and by Chrysoula Kardara and others since, do not attempt such detail, and too much still, especially of old finds, is or has to be judged from drawings, not photographs. Kardara came close to doing what was needed, but not close enough, while so far as I know little serious study has been devoted to potterwork. Potters/painters generally do not move about much, though there are exceptions to which I shall revert. If you cannot define the man at least you may define his home. Regionality is of course a major issue, by which I mean the definition of regional styles in a manner which can be applied even when they are found far from home. Two sources for the solution exist – observation of a predominance in local finds, and science. For many years the main problem was inadequacy of excavation and publication. It was natural to assume that Rhodes was a major centre as long as Italian excavations on Rhodes were so prolific and well published. It was probably only the thought that Miletos ought to be responsible too, given its historical importance in the early period, that created the term RhodoMilesian. Excavation has led to better understanding of regional styles, but too much remains unpublished and the major resource for East Greek pottery still lies, I suspect, locked away in Greek and Turkish museum storerooms. At least we can hope for more now from the latest work at major sites like Miletos, Ephesos and Klazomenai. We need much more raw material to ensure progress. Think of the improvement in studies of Corinthian pottery once the Corinth cemeteries and Perachora were published. Historically, the different fortunes of the main East Greek cities are fairly well defined for the Archaic period, even alliances and dislikes, for instance, that between the neighbours Chios and Samos, which seems well reflected in the way they ignore each other’s pottery and pottery styles, while both have little to do with the major centres on the coast opposite. I do, however, detect a possible source of confusion comparable to that with Early, Middle or Late, in defining styles as North or South Ionian. Where is the divide? For Cook, ‘North’ was Klazomenai and farther north, but not Chios. He also properly distinguished Aiolis at the north and the Dorian south. The convenience is clear, but it may be outweighed by the possibility of being led to think that there are real differences determined by being north or south, by being Aeolian or Ionian or Dorian,

rather than simply by being Lesbian or Samian or Chian or Milesian. One cannot, however, discount the usefulness of the terms when it comes to, for instance, indeterminate black-figure styles of almost anywhere from Klazomenai to Phokaia, for which ‘North Ionian’ is a very useful description. Black-figure seems to me a serious problem outside the obvious major groups. There are far too many isolated pieces, for instance, from Smyrna and Naukratis, which are not obviously from any of the major workshops and yet testify to accomplished potters and painters working somewhere. Perhaps more chemical analysis should be devoted to these isolated pieces and not just to major groups. Science can provide better definition of what clay beds were in use locally, and this has rendered great service in both defining local wares and identifying them far from home, despite whatever was done to the clay after it left its bed. But the same might have been done if the opportunity had been given, and taken, of closer stylistic analysis of local finds, since we are far from the time when clay analysis will become cheap and as readily applicable as, for instance, taking our own temperatures. We shall still have to rely on judgement by eye most of the time – another reason for more work on stylistic analysis. And where there are possibilities of various different clay sources for a single centre it becomes absolutely important that what is analysed is intelligible archaeologically and not merely by provenance. By which I mean that it must be stylistically and with certainty attributable to a group or workshop whose other members are already known. Observation that there is more than one clay type attributable to a single city, as Miletos, is very important. In the past, time and money have been spent on analysis without a proper strategy which can lead to useful results that go beyond local record; those days, I think and hope, are past. The by-products of a proper analysis and attribution of East Greek pottery centres are important, and they do not apply only to the homelands. The nature of their influence on local nonGreek peoples is easier to observe than explain and may not have depended always on Greek hands – I think of Caria, where what might be called provincial East Greek styles are easily defined, and Lydia, where we have more to learn about elaborate patterned styles, with more of their Anatolian past in them than appears elsewhere, and the possibility of their influence on Greek centres. Thasos seems to offer evidence for wandering potters/painters from Chios, but still in Greek lands. Then, identity of wares may reveal identity of those who carried them for use. This has yet to prove of great importance for East Greek pottery in the western Mediterranean beyond Italy, but there is hope. The so-called Ionian cup is particularly elusive without clay analysis, and commonly not Ionian. Naukratis plays a major role in the study. The pottery found there matches quite closely the identity of the states involved, as recorded by Herodotus; at least, I do not think we are being over-ingenious in such an observation. Perhaps the absence there of the Rhodian situlae should tell us something too. Where there is a prolific site, what is missing may be as informative as what is present. The rather Samian aspect of some of the earliest East Greek pottery at Al Mina may be historically important – later it is perhaps more Milesian. Farther afield, it seems to have been characteristic of Greeks that wherever they settled in some numbers and with the Naukratis: Greek Diversity in Egypt | 51

Boardman intention of staying they sought to make their own pottery. This is most obvious in the western colonies. The recently fashionable idea that you cannot identify people from their pottery has fortunately given place to common sense and judgement of individual cases. We can see that Greeks in particular were fussy about pottery for their tables, especially their drinking ware. In the Black Sea the local production of Milesian wares has been demonstrated by clay analysis, and there may have been other centres on the coast of the inland sea. In the west there is a distinguished succession of immigrant styles in the colonies, but also working for non-Greeks, it seems; from the Swallow Painter of wild goats, through the Northampton and Campana groups of black-figure, to the great Caeretan series. Egypt is an interesting playground for East Greek pottery studies. The studios at home were certainly aware of the possibility of, not an Egyptian market, except for their plain storage vases which were so much better than the Egyptian, but of a Greek market for fine-wares in Egypt. Yet when Egyptianinspired motifs appear, cartouches or Greek scenes mirroring Egyptian practices, one wonders whether they were meant only for Greek eyes. And are they not also sometimes an indication of Greek production in Egypt? Thus, the only Wild Goat vase I know on which the bulls are shown with their horns drawn across the top of their heads and not just pointing forward was found at Saqqara, and this is of course the normal Egyptian way of drawing horns, not Greek.5 Surely this was made in Egypt. And at Saqqara there is evidence for the presence of an East Greek painter, close kin to those who went on to paint the Caeretan vases in Italy, creating a processional scene with bulls on a wooden panel, and drawing horns in the Greek way.6 Cook thought a piece of a Wild Goat vase from Naukratis could be of Nile mud, not decent potter’s clay, and there is more evidence

52 | Naukratis: Greek Diversity in Egypt

for such pottery presented in this volume. The question of whether potters clay was ever carried for use elsewhere is a thorny one, but most acute for East Greeks in Egypt, where there is no good clay. It is easy to say there is no evidence for it, since the evidence is a matter of archaeological judgement, unless kilns are ever found. The Greeks were well used to carrying raw materials around the Mediterranean in the Archaic period – think of the hundreds of tons of Greek island marble – so the movement of clay cannot be ruled out when there are other arguments in favour. We must not be timid about assessing what a Greek might or might not do. Greek enterprise overseas has been a constant theme in the history of the east Mediterranean lands, nowhere more than at Naukratis, and I end with an anecdote which brings us away from ancient Greek pottery in Egypt and into Greeks in Egypt in the last century or so. Dick Nicholls and I travelled to Naxos in 1949. There we met an elderly Greek who said he had been with General Kitchener at Khartoum in the Sudan, when the British were fighting the forces of the Mahdi at the Battle of Omdurman in 1898. It was perfectly plausible; he was then a boy, perhaps already in his early teens, and he said he had helped his father to sell lemonade to the British troops – and, moreover, at a lower price than that in the local market. The entrepreneurial skills of the Greeks must never be underestimated. Notes

1 2 3 4 5 6

Jackson 1976. Croissant 1983. Kerschner and Schlotzhauer 2005. Boardman 2001b, ch. 2. Boardman 1998b, fig. 305. Ibid., fig. 500.

East Greek Pottery from Naukratis: The Current State of Research* Udo Schlotzhauer and Alexandra Villing Pottery, most of it from East Greece, is by far the largest category of material discovered at Naukratis. Over 7,000 pieces, mostly small fragments, are currently held in over 40 collections and museums, yet this number must be but a small proportion of what was originally found at the site: we hear of up to 5,000 sherds of pottery being excavated per day, so that looking at the current sherd count one realises just how much was not kept.1 The finds, speedily published at least in selection by Petrie, Smith, Gardner, Edgar and Lorimer (from Hogarth’s excavations)2 fell on fertile ground in the scholarly community, and soon an intensive discourse developed on East Greek pottery, led by scholars such as Boehlau, Prinz and Price.3 For the first time, bird bowls and eye bowls were studied and discussed, Chian pottery was recognised as a distinct fabric – even if christened ‘Naukratian white-faced fabric’ and attributed to local production – and speculation flourished about Egyptian influence on Greek beliefs and iconography.4 Even today the material from Naukratis still plays a crucial role in scholarship on East Greek pottery – though now of course new excavations in many of the production centres of East Greek pottery in the East Aegean, and in sites all over Greece, Western Greece, the Black Sea, the Levant, and North Africa, have completely revolutionised our understanding of East Greek pottery production and trade.5 In addition, scientific analysis of clay has greatly helped in further distinguishing and understanding East Greek pottery. The articles in this section present some of the latest developments in these fields, encompassing new interpretations of old finds and new results from recent fieldwork, art historical and scientific analyses, and in particular the results of a recent programme of clay analyses on the East Greek pottery of Naukratis (see Table 1, and the detailed report given by Mommsen et al. in the present volume). In the following paragraphs we try to summarize some of the new research, and to highlight some of the insights that have emerged recently but that are not addressed by other scholars in this volume, in particular the production and trade of Ionian cups and the identification of long-suspected but hitherto elusive local pottery production at Naukratis. Archaeology and science: the study of the pottery from Naukratis Over the past few decades clay analysis has become an integral part of pottery studies, and the pottery from Naukratis, and East Greek pottery in general, have particularly profited from this development.6 In the 1980s, Pierre Dupont’s analysis of pottery from Istros (Histria) and comparative material from other sites7 already pointed to the Milesian/South Ionian origin of the clay of much of Middle Wild Goat II and Fikellura pottery and Ionian cups, the North Ionian origin of Late Wild Goat style pottery, and the Chian origin of the clay of Chian pottery (including the

so-called ‘Grand Style’), with a possible branch workshop at Erythrai. Dupont also pointed to the existence of a local workshop of East Greek style pottery at Histria.8 The results of his analyses of over 70 samples from Naukratis, covering a great range of fabrics, are here fully published for the first time.9 Among other things they include fragments of East Greek style pottery attributed to a local workshop at Naukratis (now confirmed also by Hans Mommsen’s analyses; see below). At about the same time as Dupont produced his work, Richard Jones, together with John Boardman, analysed over 50 samples of pottery from Naukratis, including Wild Goat, East Greek black-figure, Klazomenian, Fikellura and Chian pottery, bird bowls, eye bowls, rosette bowls, Hera cups and locally produced terracotta figurines, as well as situlae from Tell Defenneh.10 Chian pottery once more emerged as Chian and clearly distinct from local Naukratis clays, in accordance with Dupont’s results,11 while ‘situlae’,12 again in accordance with Dupont’s results,13 fell into a ‘Rhodian’ group together with Vroulian cups.14 On a smaller scale, some fragments from Naukratis had also been analysed with NAA by Mike Hughes at the British Museum;15 this included a Hera mug (Fig. 15) now also analysed by Hans Mommsen (sample Nauk 2), four lamps and a Samian amphora from Naukratis (Johnston Fig. 21), all of which fell into the same Samian Group ‘L’. More recently, pottery from Naukratis and Tell Defenneh has been the subject of an extensive programme of analysis by Hans Mommsen, initiated by Udo Schlotzhauer of the Mainz Naukratis project as part of the database of Mediterranean pottery established by Hans Mommsen and Michael Kerschner and now including the British Museum among its many contributors. A list of analysed British Museum pieces, as well as some pieces in other collections, is given here in Table 1, correlating them with the Mommsen’s sample numbers; elsewhere in the volume, sample numbers are usually given in brackets after inventory numbers of analysed pieces. The main results of this programme of analysis are presented in this volume by Hans Mommsen et al., while the method is set out by Mommsen and Kerschner, taking a particular group, G, and its subgroup, g, as a case study. Most importantly, however, the results have contributed to many of the other articles throughout this section, and significantly added to our understanding of the production centres of the pottery from Naukratis. One problem which remains, however, in spite of all the increased efforts in research on the pottery of Naukratis, is the fact that our material basis is but a small portion of the actual pottery profile of the site. Some categories of material seem particularly badly affected by this; note, for example, Petrie’s mention of an ‘abundance of Roman pottery’ in the area of the temenos of Apollo,16 although there is clearly no abundance of Roman pottery among the known preserved material. Likewise, Naukratis: Greek Diversity in Egypt | 53

Northern Ionia (Teos?) Northern Ionia (Teos?) Northern Ionia (Teos?)

Northern Ionia (Teos?)

Northern Ionia (Teos?)

Northern Ionia (Teos?) Nauk 21 Northern Ionia (Klazomenai?) Nauk 20

Northern Ionia (Klazomenai?) Northern Ionia (Klazomenai?) Northern Ionia (Klazomenai?) Northern Ionia (Klazomenai?)

Northern Ionia (Klazomenai?) Defe 15

Northern Ionia (Klazomenai?) Nauk 23

Aiolis (Kyme/Larisa)

Aiolis (Kyme/Larisa) Aiolis

Aiolis Troad

Troad

Troad

Samos Samos

Samos Samos Miletos

Miletos

Miletos

Miletos

Miletos

Miletos

B B B

B

B

BE

E E E E

E

E-

G

G g

g B-Troy

B-Troy

B-Troy

J J

J J A

A

A

A

D

D

Nauk 39

Nauk 4

Nauk 7

Nauk 42

Nauk 32

Nauk 3 Nauk 72 Nauk 26

Nauk 1 Nauk 2

Nauk 65

Nauk 63

Nauk 64 Nauk 62

Nauk 13 Nauk 77

Nauk 12

Nauk 58 Defe 7 Defe 9 Defe 13

TbEgy 1

Nauk 87

Nauk 37 Nauk 54 Nauk 76

Nauk 10 Nauk 22 Nauk 24

Northern Ionia (Teos?) Northern Ionia (Teos?) Northern Ionia (Teos?)

Naukratis Naukratis Naukratis

Findspot

BM GR 1886.4-1.1040

BM GR 1888.6-1.544F (= 1924.12-1.1114) BM GR 1886.4-1.570

Cambridge, Fitzwilliam Museum GR 22.1894 BM GR 1886.4-1.830a

BM GR 1888.6-1.405 BM GR 1965.9-30.498 Boston, MFA 88.972

BM GR 1888.6-1.401 BM GR 1888.6-1.403

BM GR 1965.9-30.508

BM GR 1888.6-1.613a

BM GR 1888.6-1.637 BM GR 1888.6-1.634

BM GR 1886.4-1.1294 BM GR 1888.6-1.658

BM GR 1888.6-1.573b,c

BM GR 1888.2-8.117 (Vase B.128.1) Boston, MFA 88.840

BM GR 1888.6-1.544d BM GR 1888.2-8.139b BM GR 1888.2-8.86 BM GR 1888.2-8.171

Black-figured amphora with band of cartouches of Pharaoh Apries Dinos stand (?) Painted plate with Wild Goat style decoration Painted plate Wild Goat style jug or amphora Wild Goat style lidded amphora Klazomenian black-figure slim amphora; Petrie Painter Klazomenian black-figure amphora; Urla Group Amphora or oinochoe with black-polychrome decoration Oinochoe with black-polychrome decoration Dinos of the London Dinos group Male terracotta head, painted

Large North-Ionian black-figure plate

Lid with scale decoration Wild Goat style dinos (?) Jug or amphora with black-polychrome decoration Painted plate with patterned decoration Painted plate with metope decoration North-Ionian black-figure plate

Description

Naukratis

Naukratis

Naukratis

Naukratis

Naukratis

Ionian cup type 11 with votive graffito to Apollo Fikellura amphora

Plate with metope decoration

Krater with votive graffito by Polemarchos

Lid with metope decoration

Grey ware fenestrated stand Grey ware dinos rim with votive graffito to Aphrodite by a Mytilenean Naukratis Grey ware stand with votive graffito ‘M]aloeisio[s’ Naukratis Grey ware vessel with incised wavy lines (or Tell Defenneh?) Naukratis Samian Hera mug with dipinto Naukratis Samian Hera mug with dipinto (bottom right) Naukratis Samian Hera mug with dipinto Naukratis East Greek painted plate Naukratis Krater

Naukratis Naukratis

Naukratis Naukratis

Naukratis

Naukratis

Tell Defenneh

Naukratis Tell Defenneh Tell Defenneh Tell Defenneh

BM GR 1924.12-1.1107 Naukratis BM GR 1924.12-1.1124 Naukratis BM GR 1888.6-1.551 Naukratis + 1965.4-28.1 BM GR 1924.12-1.1127 Naukratis (belongs to GR 1965.9-30.704) Petrie Museum UC30035a-b Thebes, Egypt (joining Basel, Cahn, HC 1175) Boston, MFA 88.815 Naukratis Boston, MFA P4631 Naukratis

BM GR 1886.4-1.1267f Boston, MFA 88.851 Boston, MFA 86.544

Sample no. Museum no.

Place of production

NAA Group B B B

Table 1 List of pottery analysed by NAA

Möller 2000a, 91, 241, 297, pl. Ib; Walter-Karydi 1973, 59, pl. 77.645

Fairbanks 1928, 117 no. 330.2, pl. 36; Schlotzhauer 2001a, 122, pl. 17.2

Schlotzhauer 2006, 311-13 no. 4.A, 318 fig. 11

Schlotzhauer 2006, 311-13 no. 4.B, 318 fig. 12 Schlotzhauer and Weber 2005, 81, 106 fig 10

Möller 2000a, 173 no. 3

Möller 2000a, 172-3 no. 1a

Gardner 1888, pl. 14.5

Fairbanks 1928, 116 no. 328.1, pl. 36.

Cook 1954, pl. G.B. 587.16

Petrie 1888, pl. 31.6; Cook 1954, pl. G.B. 605.7 Cook 1954, pl. G.B. 587.7

Fairbanks 1928, 116 no. 336, pl. 37.

Fairbanks 1928, 110 no. 321.8, pl. 34. Fairbanks 1928, 116 no. 328.2, pl. 36.

Publication

2nd–3rd quarter 6th cent. Walter-Karydi 1973, 3, pl. 2.26

1st half 6th cent.

Late 7th–early 6th cent.

Late 7th–early 6th cent.

1st third 6th cent. 2nd quarter 6th cent. ? 2nd–3nd quarter 6th cent. Late 7th–early 6th cent.

1st third 6th cent. 1st third 6th cent.

7th-6th cent.

1st half 6th cent.

Last third 7th–1st third 6th cent. Last third 7th–1st third 6th cent. 1st third 6th cent. Mid–3rd quarter 6th cent. 6th cent. 1st half 6th cent.

c. 530/20

1st half 6th cent. 1st third 6th cent. 1st third 6th cent. c. 540/30

1st half 6th cent. 1st half 6th cent.

2nd–3rd quarter 6th cent.

Mid-6th cent.

1st third 6th cent. 1st third 6th cent. Mid-6th cent.

1st half 6th cent. Late 7th–early 6th cent. 1st half 6th cent.

Date BC

Fig. 22

Fig. 21

Fig. 20

Fig. 19

Fig. 16 Fig. 17 Fig. 18

Fig. 14 Fig. 15

Fig. 13

Fig. 12

Kerschner Fig. 10 Fig. 11

Kerschner Fig. 8 Kerschner Fig. 11

Kerschner Fig. 9

Fig. 10

Fig. 6 Fig. 7 Fig. 8 Fig. 9

Bailey Figs 1-5

Fig. 5

Fig. 2 Fig. 3 Fig. 4

Fig. 1

Fig. no.

Nauk 82 Nauk 83

Nauk 34

n.l. (Eastern Doris?)

n.l. (Eastern Doris?)

n.l. (Eastern Doris?)

n.l. (Eastern Doris?) n.l. (Eastern Doris?) n.l. (Eastern Doris?)

n.l. n.l.

n.l.

n.l. n.l. Attica (?)

Attica (?) Attica? Cyprus Cyprus Cyprus Cyprus Egyptian marl Egypt (Naukratis?)

Egypt (Naukratis?)

Egypt (Naukratis?) Egypt (Naukratis?) Egypt (Naukratis?) Egypt (Naukratis?) Egypt (Naukratis?)

Egypt (Naukratis?) Egypt (Naukratis?) Egypt (Naukratis?)

Egypt (Naukratis?) Egypt (Naukratis?)

Egypt (Naukratis?) Egypt (Naukratis?)

TD

TD

TD

TD TD TD

RHc1 ITAN

ITAN

ITAN ITAN KROP

KROP perb CYPT EMEA EMEa EMEa MarlQANN

QANN

QANN QANN QANN QANN QANN

QANN QANN QANN

QANN QANN

QANN QANN

QANN- Egypt (Naukratis?)

Nauk 79 Nauk 81

Nauk 25 Nauk 27 Nauk 33

Nauk 14 Nauk 15 Nauk 16 Nauk 17 Nauk 19

Nauk 9

Nauk 57 Nauk 88 Nauk 35 Nauk 55 Nauk 67 Nauk 68 Nauk 18 Defe 10

Nauk 73 Nauk 74 Nauk 43

DlEgy 1

Nauk 53 Abus 1

Defe 4 Defe 5 Defe 8

Defe 3

Defe 2

Defe 1

Knid 1

Knidian peninsula

EMEb

Defe 11 Nauk 6 Nauk 51

Miletos n.l. Knidian peninsula

D DD EMEB

Naukratis Naukratis Naukratis Naukratis Naukratis Naukratis Naukratis Tell Defenneh

Naukratis Naukratis Naukratis

Egypt, Delta

Naukratis Abusir

Tell Defenneh Tell Defenneh Tell Defenneh

Tell Defenneh

Tell Defenneh

Tell Defenneh

Near Datcha

Tell Defenneh Naukratis Naukratis

Naukratis Naukratis

Naukratis Naukratis

Cambridge, Museum of Naukratis Classical Archaeology NA 256

BM GR 1965.9-30.739 BM GR 1886.4-1.83

BM GR 1965.9-30.501 BM GR 1965.9-30.536

BM GR 1924.12-1.43 Naukratis (same vessel as1924.12-1.42) BM GR 1910.2-22.232b Naukratis BM GR 1910.2-22.233 Naukratis BM GR 1910.2-22.243 Naukratis BM GR 1886.10-5.12 Naukratis Berlin, Ägyptisches Museum Egypt und Papyrussammlung 7206 Boston, MFA 86.533 Naukratis Boston, MFA P 4864 Naukratis Cambridge, Museum of Naukratis Classical Archaeology NA 48

BM GR 1886.4-1.1324 BM GR 1965.9-30.494 BM GR 1886.4-1.80 BM GR 1910.2-22.16 BM GR 1886.4-1.77 BM GR 1886.4-1.81 BM GR 1910.2-22.15 BM GR 1888.2-8.57

BM GR 1886.4-1.1271 Bonn,Akademisches Kunstmuseum 2002.5 Bonn,Akademisches Kunstmuseum 1524 BM GR 1886.4-1.671 BM GR 1888.6-1.608.a BM GR 1886.4-1.678

BM GR 1888.2-8.65 (Vase B106.19) BM GR 1888.2-8.16b & 17 (Vase B 106.12-13) BM GR 1888.2-8.20 (Vase B106.11) BM GR 1888.2-8.42a BM GR 1888.2-8.44a BM GR 1888.2-8.25

BM GR 1893.11-13.4

BM GR 1888.2-8.46a BM GR 1886.4-1.1025 BM GR 1886.4-1.96

Shallow burnished plate, Egyptian, 26th dynasty Trefoil-mouthed oinochoe with wavy-line decoration Burnished small dinos with graffito Amphora or hydria with painted bands and votive graffito to Apollo Lamp

Plate with East Greek style decoration Plate with pierced lug handle Plate with East Greek style decoration

Undecorated plate with graffito Undecorated plate Black-glazed dinos with votive graffito by Phanes, son of Glaukos Head kantharos Hydria or stamnos, white-on-black decoration Mortarium (Villing cat. no. 2) Mortarium (Villing cat. no. 19) Mortarium (Villing cat. no. 10) Mortarium (Villing cat. no. 12) Mortarium with graffito (Villing cat.no. 17) Amphora with East Greek style patterned decoration; NaukA II Plate with East Greek style decoration and pierced lug-handle Jug with ribbed neck and graffito ‘deka’ Pot stand/kiln furniture with pre-firing graffito Pot stand/kiln furniture with pre-firing graffito Closed vessel with painted floral decoration Jug with ribbed neck and Carian graffito

Black-figure amphora

Stamnos related to East Greek ‘situlae’ Stamnos related to East Greek ‘situlae’ Amphora with with representation of men beside tripod Wild Goat style plate Small closed vessel with painted bands

East Greek ‘situla’ with representation of bull

East Greek ‘situla’

Fikellura amphora Cup with everted rim (Ionian cup); type 10,2.B East Dorian cup with Phoenician inscription East Dorian plate with representation of ship (Attula cat. no. 4) East Greek ‘situla’ with representation of owl

Cook 1954, pl. G.B. 605.5 Cook 1954, pl. G.B. 605.4 Cook 1954, pl. G.B. 598.10; 605.6

Petrie 1888, pl. 26, 12-12a; Cook 1954, pls G.B. 598.3-4, 602.1-2 Petrie 1888, pl. 26.15; Cook 1954, pl. G.B. 603.2

Cook 1954, pl. G.B. 599.3

6th cent.

2nd–3rd quarter 6th cent. 1st half 6th cent.

6th cent. 1st half 6th cent.

2nd–3rd quarter 6th cent. 2nd–3rd quarter 6th cent. 2nd–3rd quarter 6th cent.

6th cent. Ptolemaic (2nd cent.?) Ptolemaic (2nd cent.?) Ptolemaic 6th cent.

2nd–3rd quarter 6th cent.

1st half 6th cent. Late 6th cent. 1st half 6th cent. 1st half 6th cent. 1st half 6th cent. 3rd quarter 6th cent.

2nd half 6th cent.

6th cent. 6th cent. 2nd half 6th cent.

Schlotzhauer 2006, 308-11 no. 3.C, 317 fig. 10

Schlotzhauer 2006, 308-11 no. 3.B, 317 fig. 9

Petrie 1888, pl. 32.4; Cook 1954, pl. G.B. 606.3; Weber (forthcoming)

Walter-Karydi 1973, 30, pl. 57.485

Schlotzhauer 2006, 294-301 no. 1, 315 figs 1-3

Late 7th–early 6th cent. Last quarter 7th cent. – 1st half 6th cent. 2nd–3rd quarter 6th cent. Geominy 1992, 48-50 figs 10-11; Mielsch 2003, 58-9

2nd third 6th cent. 2nd third 6th cent. 2nd third 6th cent.

2nd third 6th cent.

2nd third 6th cent.

2nd third 6th cent.

2nd third 6th cent. Walter-Karydi 1973, 66, pl. 89.683 2nd–3rd quarter 6th cent. Late 7th–first third Schlotzhauer 2006, 301-7 no. 2, 316 figs 4-6 6th cent. (epigr. com.W. Röllig) Late 7th–first third 6th cent.

Fig. 34 Fig. 38

Fig. 31 Fig. 33 Fig. 30 and Dupont & Thomas Fig. 1, Nau 9 Fig.41 Fig. 37

Fig. 35 Fig. 43 Fig. 44 Fig. 42 Fig. 36

Fig. 32

Villing Fig. 2 Villing Fig. 8 Villing Fig. 21 Villing Fig. 22 Villing Fig. 23 Fig. 39

Johnston Fig. 11 Fig. 26

Fig. 25

Weber Fig. 21 Weber Fig. 20 Weber Fig. 23

Weber Fig. 20

Weber Fig. 19

Weber Fig. 18

Attula Figs 5-6

Schlotzhauer Fig. 3 Fig. 23 Fig. 24

Schlotzhauer and Villing Table 1 cont. List of pottery analysed by NAA Pairs Defe 16: BM GR 1888.2-8.139g (1) = Pair 1 Defe 17: BM GR 1888.2-8.139g (2) = Pair 1 Nauk 69: BM GR 1888.6-1.574 (Vase B103.11) = Pair 2 Nauk 70: BM GR 1886.4-1.1113 (Vase B 102.13) = Pair 2 Nauk 85: BM GR 1965.9-30.972 a = Pair 3 Nauk 86: BM GR 1965.9-30.972 d = Pair 3 Nauk 8: BM GR 1886.4-1.1263 + 1924.11-1.1113 = Pair 4 Emec 31: Emecik excavation, inv. no. ST -1-I 8b-8c,74 = Pair 4 Kari 2: Bochum,Antiken- und Kunstsammlung der Ruhr-Universität S 987 = Pair 5 Nauk 66: BM GR 1888.6-1.653 (Williams and Villing Fig. 1) = Pair 5 Singles Nauk 5: BM GR 1886.4-1.1031 Nauk 11: BM GR 1888.6-1.573a Nauk 28: Boston, MFA P.4714 Nauk 29: Boston, MFA 88.1091 Nauk 30: Bonn,Akademisches Kunstmuseum 697.25 Nauk 36: BM GR 1886.4-1.1219 Nauk 41: BM GR 1888.6-1.561 + 1924.12.1.1058+1102 Nauk 44: BM GR 1888.6-1.640b Nauk 47: BM GR 1888.6-1.644c Nauk 52: BM GR 1910.2-22.17 Nauk 56: BM GR 1886.4-1.75 Nauk 59: BM GR 1888.6-1.569; Weber Fig. 25 Nauk 78: BM GR 1886.4-1.1311; Weber Figs 14-15 Nauk 80: BM GR 1965.9-30.504 Nauk 84: BM GR 1886.4-1.82.b Defe 6: BM GR 1888.2-8.139a Defe 12: BM GR 1952.5-5.11 Defe 14: BM GR 1888.2-8.77a (Vase B 109) Kami 2: BM GR 1860.4-4.44 Kari 1: Bochum,Antiken- und Kunstsammlung der Ruhr-Universität S 985 Milet 41: Miletos excavation inv. no. K86.97.6 Rhod 20: BM GR 1868.4-5.78; Weber Fig. 24

of the numerous pieces of basket-handled amphorae reported by Petrie very few are known today (e.g. Johnston Fig 14). 76 In general, one of the main criteria for keeping pottery seems to have been the presence of inscriptions; even the tiniest scraps of sherds, if they preserved part of a letter, were kept, be they decorated or not. Painted pottery was often kept even if there was no dedication, although one gets the impression that some must also have been thrown away. Undecorated coarse-ware was by and large only kept if it was a complete pot (Edgar even devoted a short article to the publication of a group of unpainted pots found in a well in 1899)18 or if a fragment carried an inscription or a sealing; large numbers of stamped handles of trade amphorae, for example, are known to have survived.19 In addition, a few samples may have been kept as specimens of particular types of fabrics, which Petrie in particular had been very keen to distinguish. The surviving pottery from Naukratis is thus but a skewed sample of the actual pottery profile of the site, heavily tilted towards inscribed and decorated votive pottery. This always needs to be kept in mind when drawing any conclusions based on the surviving evidence. It is particularly irritating to think that most of the kitchen and household pottery that must have existed at Naukratis and that might have added important additional information to our understanding of the site is lost to us. No cooking pots, for example, have yet been identified among the surviving material.20 However, as the contribution by Villing on the mortaria from Naukratis shows, the study of coarse-ware sherds can provide new information on cults at the site as well as on trade links, in this case with Archaic Cyprus and Classical Corinth. An additional distortion in the perception of the pottery from Naukratis has been caused by selective publication: some fabrics and styles have been studied and published extensively, with nearly every single known sherd 56 | Naukratis: Greek Diversity in Egypt

being catalogued, while others have been severely neglected, with only a small selection or nothing at all being included. As a consequence, even among the preserved material the importance of some types has erroneously been over- or underestimated, a problem which will only be solved once the ongoing project of gathering and collating all the known material is completed. The painted pottery of Naukratis: the current state of research As has been observed many times before, the painted pottery found at Naukratis is mostly of East Greek origin, and in particular stems from poleis said to have been involved in the foundation of Naukratis. In addition, much material was also found that originated in Athens, Corinth, and Laconia. Such wares were widely traded around the Mediterranean at the time so that their appearance at Naukratis is hardly surprising.21 Attic pottery in particular is of a varied character and covers the full time-span of Athenian pottery trade, including many early imports,22 but also much late black-glazed material.23 Following on from earlier studies,24 it is currently being examined by Valerie Smallwood and Susan Woodford. The Corinthian pottery, too, encompasses some early material but is not particularly numerous; Laconian pieces are fewer and later. These wares, however, are not discussed in the present volume; instead, emphasis is placed on East Greek pottery, where great advances have been made recently in our understanding of the various production centres (for a map of East Greece, see Villing and Schlotzhauer Fig. 1b). North Ionia North Ionian pottery, rich and varied and only slowly becoming better understood,25 furnishes a large proportion of the painted pottery from Naukratis – not surprising, perhaps, given that according to Herodotus Klazomenai, Teos and Phokaia were among the founding cities of Naukratis. At Klazomenai and Smyrna in particular, new excavations (and the restudy of old finds) have revealed much relevant material, providing new insights into the complex variety of styles produced at those sites. Other sites, such as Phokaia and Teos, however, remain archaeologically obscure. It has been raised as a possibility, however, that the hitherto only little known, but seemingly high quality, production of Teos can now be recognised in Dupont’s archaeometric Group D (Dupont and Thomas Figs 1.15,11,14,16; 4.51-52) and particularly Mommsen and Kerschner’s Group B (Figs 1–5), though this remains to be confirmed by further analyses. Group B, otherwise also known as the ‘Bird Bowl Workshops’, certainly encompasses a wide variety of styles and shapes, from Wild Goat to black-figure, including also some high-quality black-figured plates (Figs 4-5) as well as the wellknown Apries amphora (Bailey Figs 1-5), decorated with a band of Egyptian cartouches, and now thanks to Donald Bailey known to have been found in Egyptian Thebes. Although the group can hardly be said to form a stylistically coherent whole, there are similarities among some of the pieces,26 and certainly several instances of painters creating a very interesting and unusual iconography. Some of the shapes are also broadly paralleled in material from North Ionian sites, such as at Klazomenai, where, for example, amphorae related in shape to the Apries amphora have been found.27 Both stylistic and scientific observations,

East Greek Pottery from Naukratis: The Current State of Research

Figure 1 Lid, sample Nauk 10, Group B (North Ionia – Teos?)

Figure 2 Painted plate, sample Nauk 37, Group B (North Ionia – Teos?)

Figure 4 Black-figure plate, sample Nauk 76, Group B (North Ionia – Teos?)

Figure 6 Painted plate, sample Nauk 58, Group E (North Ionia – Klazomenai?)

Figure 5 Fragment (left) from blackfigure plate (right), sample Nauk 87, Group B (North Ionia – Teos?)

Figure 7Wild Goat style jug or amphora, sample Defe 7, Group E (North Ionia – Klazomenai?)

Figure 3 Painted plate, sample Nauk 54, Group B (North Ionia – Teos?)

Figure 8 Wild Goat style lidded amphora, sample Defe 9, Group E (North Ionia – Klazomenai?)

then, support a North Ionian origin of this and other pieces in the group. Visual observations alongside clay analysis have also helped to change our assessment of the highly exceptional large figured plate from Naukratis (sample Nauk 87, Fig. 5). Once attributed to a Chian workshop by Walter-Karydi,28 clay analysis has now confirmed a North Ionian (Group B) origin, which might also have been suspected purely on stylistic grounds. Another highly unusual figured plate, though different in style (sample Nauk 76, Fig. 4), belongs to this same group, as (on visual evidence) do several other examples of plates and other shapes from Naukratis.29 One may also consider the sherd with a representation of a black African (book cover) in this context; its attribution by Walter-Karydi30 to a Chian workshop seems doubtful, and close examination places it near the Apries amphora.31 As expected, there is also a fair number of bird bowls, from around 620/10 bc, and other hemispherical bowls

Figure 9 Klazomenian black-figure Figure 10 Klazomenian black-figure amphora, Petrie Painter, sample Defe amphora, Urla Group, sample Defe 15, 13, Group E (North Ionia – Klazomenai?) Group E (North Ionia – Klazomenai?)

(‘Kalottenschalen’) at Naukratis: rosette bowls, lotus bowls and eye bowls, banded bowls and maeander bowls are all to be found (even if eclipsed in number by South Ionian cups with everted rim).32 They have a wide distribution but their origins lie predominantly in North Ionian centres; Group B (possibly Teos) has been detected by clay analysis of pieces from various sites, but also E (possibly Klazomenai), F (presumably Smyrna), G (presumably Kyme; e.g. Kerschner Fig 26) and other groups.33 Klazomenian pottery is present at Naukratis in considerable amounts, and includes examples of the well-known Klazomenian black-figure. However, this type of pottery has so far been quite difficult to isolate scientifically: two sherds form a chemical pair (samples Nauk 69-7034 = pair 2), another sherd remains a chemical loner (sample Defe 14),35 yet two pieces from Tell Defenneh – one sherd by the Petrie Painter and another from the Urla Group (samples Defe 13 and 15, Figs 9-10) – fall into the chemical provenance pattern group E, which has tentatively been associated with Klazomenai, not least since Naukratis: Greek Diversity in Egypt | 57

Schlotzhauer and Villing

Figure 13 Grey ware vessel, sample Nauk 65, Group B-Troy (Troad)

Figure 11 Grey ware dinos, sample Nauk 62, Group B-Troy (Troad)

Figure 12 Grey ware stand, sample Nauk 63, Group B-Troy (Troad)

another fragment of an oinochoe of the Urla Group (from Smyrna) belonged to this group in an earlier set of analyses.36 Further pieces that fall into Group E show that it encompassed a wide range of shapes and decorative schemes, including amphorae from Tell Defenneh somewhat reminiscent of ‘Borysthenes’ amphorae (Figs 7-8).37 The material from Tell Defenneh will shortly be presented in detail by S. Weber.38 There can be little surprise, therefore, that the decoration of a plate from Naukratis, Nauk 58 (Fig. 6), closely resembles a plate found at Klazomenai.39 This latter plate, moreover, features a sofa rim with a characteristic volute pattern that is also found on another fragment from Naukratis (Paspalas Fig. 2) as well as, slightly modified, on a plate from Smyrna (Paspalas Figs 1). The plate is, of course, a common shape across North Ionia, and could be decorated with a variety of motifs. The most characteristic type, perhaps, is that with a meander-rim, apparently produced in several North Ionian centres. One of these seems to be Group B (Teos?), but there is also the possibility of a local production in the Aiolian pottery centre G (presumably Kyme; e.g. Kerschner Fig. 27) and at Smyrna. As Stavros Paspalas in his survey on plates from the old excavations at Smyrna points out, Smyrna seems to have had its share of imported Klazomenian, Tean(?) and other North Ionian pottery, as did Naukratis. The pottery production of Smyrna itself is, however, still somewhat difficult to distinguish. A chemical provenance group, Group F, might be of local origin.40 Some of the other plates from old excavations in Smyrna may also be of local production, notably those with simple floral ornaments (Paspalas Figs 4, 18, 22). They are found less widely spread and, unlike many of the other plates from Smyrna, do not seem to occur at Naukratis. As a phenomenon, the distinct floral patterns of these Smyrnaean dishes, however, seem to be mirrored in Aiolian ‘drop-style’ pottery (Group G/g) (Kerschner Figs 24–5), one of several instances of interconnections between Aiolian (G) and Smyrnaen (F) pottery.

actual place of production of Group G/g cannot yet be determined with certainty, Kerschner argues for Kyme and perhaps, on a smaller scale, Larisa. The wide distribution of the pieces might in part at least be explained by Phokaian trade. Grey wares were found in considerable numbers in Naukratis, but have to date remained largely unpublished.42 Given that Aiolis is usually assumed to have been a major production centre for this type of pottery,43 it is not surprising to find that one such fragment (Nauk 64, Kerschner Fig. 10) falls into Aiolian subgroup g (presumably Kyme). In addition, analysis has established a second Aiolian group of Grey wares from Naukratis, which falls into the chemical Group B-Troy. This group includes two fragments found in the temenos of Aphrodite which, through their inscriptions, seem connected with Mytilene on Lesbos, one of the founding cities of Naukratis. One (Nauk 62, Fig. 11) is the rim of a dinos inscribed on the shoulder with a dedication to Aphrodite by a Mytilenean.44 Traces of red pigment can be discerned on this piece, a feature also witnessed on an Archaic Grey-ware dinos from Troy.45 Added decoration – in red, white and black – is also found on a few sherds from the Aiolian region, Smyrna, South Ionia, Rhodes, Delos and Crete.46 Bayne47 had speculatively suggested production of painted Grey wares in Naukratis, but both our analyses as well as other studies now show that several places must have produced such painted wares, and so far there is no evidence for Naukratite production.48 The second fragment in Group B-Troy (Nauk 63, Fig. 12) belongs to an open-work stand preserving a name (?) seemingly related to a Lesbian cult title of Apollo, Maloeisios;49 its shape, too, can be paralleled on Lesbos.50 Rather than being located on Lesbos, however, Group B-Troy is, in fact, a pattern which is found in pottery from the region of Troy from the Bronze Age to the Archaic period, including Archaic Grey wares.51 It is believed to be local to the area52 and, as Mommsen argues in the present volume, may be conntected with what Posamentir has christened the ‘Hellespont workshops’. Indeed, the third analysed fragment from this group (Nauk 65, Fig. 13) with its incised wavy lines finds parallels in pottery from Troy, though also in pieces from Lesbos.53 The question arises, therefore, how the NAA result can be reconciled with the Mytilenean/Lesbian credentials of at least two of the analysed pieces. Certainly Grey wares had been in common use both on Lesbos and in the Troad for centuries, and at least from the end of the 8th century bc at least some of the Grey wares in the Troad may have been produced (and inscribed?) by Lesbian colonists.54 Perhaps Lesbos and the Lesbian peraia also exploited related clay beds. It remains to be seen how future archaeological studies and chemical analyses

Aiolia and Grey wares Aiolian pottery is one group for which a good chemical profile has now been established in Group G/g, as Michael Kerscher reports in his contribution in the present volume. Interestingly, a very wide variety of styles of decoration makes up this group (Kerschner Figs 8–11, 13–30), such as Wild Goat, black-figure and black-polychrome styles, including the so-called ‘London Dinos group’, banded wares, maeander-rim dishes and Grey ware pottery, to mention but a few. The phenomenon of extremely widely ranging production has also been witnessed elsewhere in East Greece, such as at Klazomenai.41 Although the 58 | Naukratis: Greek Diversity in Egypt

East Greek Pottery from Naukratis: The Current State of Research

Figure 14 Hera mug, sample Nauk 1, Group J (Samos)

Figure 15 Hera mug, sample Nauk 2, Group J (Samos)

Figure 19 Painted krater, sample Nauk 42, Group A (Miletos)

will define the role played by Lesbos in the wider Aiolian region, both with regard to decorated and undecorated pottery. Chios Far better known is the pottery of the island of Chios, which forms a particularly large and conspicuous group among the pottery from Naukratis – so much so that it was at first widely believed to a be a locally produced fabric. Well researched and published, there still remains much to be studied, as can be seen from the latest research on the workshops that produced Chian pottery by Dyfri Williams in the present this volume. We find, for instance, a Laconian artist at work in a Chian workshop, another example of the well-attested phenomenon of the migration of artists in the Archaic period and surely one of the reasons for the transfer of stylistic features from one region to another. Remarkable, too, is the existence at Naukratis of phallus-shaped drinking cups, uniquely appropriate perhaps to the cult of

Figure 22 Fikellura amphora, sample Nauk 39, Group D (Miletos)

Figure 16 Hera mug, sample Nauk 3, Group J (Samos)

Figure 17 Painted plate, sample Nauk 72, Group J (Samos)

Figure 18 Painted krater, sample Nauk 26, Group A (Miletos)

Figure 20 Painted plate, sample Nauk 7, Group A (Miletos)

Figure 21 Cup with everted rim (Type 11), sample Nauk 4, Group D (Miletos)

Aphrodite in a harbour town famous for its attractive hetairai. Even if it seems increasingly unlikely that any Chian pottery was ever produced at Naukratis, more research needs to be devoted to the question of workshops on Chios and perhaps its vicinity. There are indications that Chian workshops were operating on the neighbouring mainland, perhaps at Erythrai, and it is hoped that more chemical analyses (planned by the British Museum, H. Mommsen and M. Kerschner) may shed further light on the question.55 South Ionia and Caria As regards South Ionia, pottery from Samos and especially Miletos (both founding cities of Naukratis) is well attested at Naukratis. From Samos, in addition to trade amphorae (e.g. Johnston Fig. 21), there are ritual vessels with dipinti (Hera cups and mugs) used in the Samian filial cult of Hera at Naukratis (Figs 14–16, 29). That these were imported and not locally made has been confirmed by analyses and demonstrates the close cultic link that remained between the Samians at Naukratis and their home island.56 In addition, a decorated plate falls into the same chemical provenance group, J (Nauk 77; Fig. 17). Of the hundreds of so-called ‘Ionian cups’, or cups with everted rim (Knickrandschalen), many have come from Miletos (Nauk 4, Fig. 21), although some may also have been produced elsewhere, as will be discussed in more detail below. Wild Goat style and Fikellura pottery, too, is well attested at Naukratis, and most has been attributed to two Milesian workshops, Group A (the ‘Kalabaktepe workshops’: Nauk 7, 26, 32, 42; Figs 18–20) and Group D (Nauk 39, Fig. 22; cf. also Defe 11, Schlotzhauer Fig. 3). That Miletus was indeed the major producer and the driving force behind the development of the Fikellura style has Naukratis: Greek Diversity in Egypt | 59

Schlotzhauer and Villing

Figure 23 Cup with everted rim (Type 10, 2B), sample Nauk 6, Group DD (not located)

Figure 24 East Dorian cup with everted rim, sample Nauk 51, Group EMEB (Knidian peninsula)

now become clear through the finds from recent excavations at Miletos. As Udo Schlotzhauer points out in the present volume, Milesian pottery not only encompasses shapes that combine the Fikellura and Wild Goat styles (Schlotzhauer Fig. 1),57 but also features a previously unparalleled wealth of figured images, such as the ‘Potnia Theron’ painted on the inside of a cup (Schlotzhauer Fig. 2). Fikellura pottery thus presents itself increasingly as on a par with the other figured pottery of the Archaic II period in East Greece, notably the rich tradition of North Ionian black-figure painting. Finally, pottery in a South Ionian style also seems to have been produced at a few Milesian colonies in the 6th century bc. Production at Histria had already been suggested by Dupont, but recent archaeometric analyses by Mommsen, Posamentir and Kerschner have now established workshops also in the Hellespont region,58 suspected by Posamentir to be identical with Dupont’s Ionie du Sud 3, which exported their products preferentially to the north, towards other Greek colonies such as Berezan (see Posamentir Figs 17–20). Stylistically related in many ways, but of a highly distinctive nature, is the pottery of the region of Caria, around and to the south of Miletos. None of the characteristic examples of Carian pottery as we know them for the region of Mylasa seems to have been found at Naukratis. There is, however, one Subgeometric fragment which through NAA has been associated with a typical Carian vessel (Pair 5), and another fragment which seems to belong to local Carian pottery production at Kaunos on the Carian/Lycian border; both fragments in all likelihood date to the earliest period of Naukratis (Williams and Villing Figs 1–2). One is tempted to think of Carian mercenaries passing through or settling at Naukratis, or perhaps of traders from Halikarnassos, another founding city of Naukratis.

plates fall into the chemical provenance Group EmeB, which must be attributed to East Dorian, probably local Knidian, workshops. The group also includes a cup found in Naukratis of a characteristic East Dorian type, paralleled especially on Rhodes, and bearing a Phoenician inscription (Fig. 24)60 – a further instance of the cosmopolitan nature of trade around the Mediterranean. Rhodes itself, of course, also had a stake in the port of Naukratis. While most East Greek pottery used to be thought of as Rhodian, this idea has now been dispelled by recent research;61 indeed, there is little left that can with any certainty still be attributed to the island. The Rhodian origin of the ‘situlae’, discussed here by Sabine Weber, which are so common in Tell Defenneh, remains a strong possibility. Analysis has shown several of them (Weber Figs 18–23) to fall into a chemical provenance Group TD, which cannot yet be localized but (particurly considering the results of earlier analyses that placed them in a Rhodian context) which may well eventually find its home on Rhodes or in its vicinity, as may perhaps another fragment of a plate, Nauk 53 (Fig. 25; Group RHc1) with an unusual representation of a deer. Finally, one may speculate about whether the enigmatic Archaic plate from Naukratis with the ‘kleps-‘ inscription (Johnston Fig. 11) may also stem from somewhere in this region. Both in shape and in clay composition it goes together with a larger fragment (Fig. 26; Group ITAN), and even though the origin of these pieces cannot be localised yet, one may point to very similar fragments found on the acropolis of Lindos.62

East Doris For another founding city, Knidos, new excavations at the Apollo sanctuary at Emecik have now established a large body of Archaic pottery,59 which includes a number of decorated plates. As Regina Attula demonstrates in the present volume, they are notable particularly for their figured decoration, which includes a ‘Potnia Theron’ (Attula Fig. 4 – compare the Fikellura ‘Potnia Theron’ from Miletos mentioned above, Schlotzhauer Fig. 2) as well as several representations of ships, an appropriate theme for a sanctuary close to an important marine port. Many of the

Cups with everted rim (‘Ionian cups’): a much-neglected class of its own Cups with everted rim have so far not been systematically studied as a group of material at Naukratis,63 as opposed to other classes of material such as painted Attic,64 Chian,65 Fikellura (MileA II),66 or certain North Ionian groups of pottery.67 Even though over a hundred years have passed, not all the preserved finds from the four seasons of work in the late 19th and early 20th century have been taken into account by modern scholarship. This selective treatment of finds from Naukratis68 has repeatedly led to a distortion in their assessment.69 As has been mentioned above, one of the cups with everted rim from Naukratis (Fig. 24)70 has recently been shown to have probably been produced in a workshop in the region of

Figure 25Wild Goat style plate, sample Nauk 53, Group RHc1 (not located)

Figure 26 Undecorated plate/bowl, sample Nauk 74, Group ITAN (not located)

60 | Naukratis: Greek Diversity in Egypt

East Greek Pottery from Naukratis: The Current State of Research

Figure 27 Cup with everted rim (Type 5.C), BM GR 1886.4-1.1036

Figure 28 Cup with everted rim (Type 9.A), BM GR 1886.4-1.1035

Knidos/Emecik (NAA Group EmeB). This piece, indeed, belongs to the East Dorian variant of cups with everted rim, which in Naukratis is a far rarer import than its prolific South Ionian sibling, an observation which fits in well with the known pattern of distribution of both these types in the Levant as well as in North Africa.71 The South Ionian variant is attested by several hundred fragments in the store-room of the British Museum alone. Four of these are illustrated here; they are broadly representative of the variations that are found of the South Ionian type among the imports to Naukratis (Figs 21, 23, 27–28). South Ionian cups with everted rim, in their various variations, are found in Naukratis from the last quarter of the 7th until the end of the 6th century bc. The earliest type of cup with everted rim to be found at Naukratis is Type 5.72 This type still belongs to the late 7th century bc and in its latest Variant 5.C, to which our cup belongs (Fig. 27), comes to an end already in the first quarter of the 6th century bc. The span of production of Type 5 thus overlaps with the earliest phase of Greek presence at Naukratis. Cups with everted rim of Type 5 are, accordingly, rare among finds at the site. Type 9, by contrast, belongs exclusively to the 6th century bc .73 Its production begins around 590–80 bc at the earliest. At Naukratis both early variants of this type, such as a cup of Variant 9.A (Fig. 28) from the second quarter of the 6th century bc, and late variants from the second half of the 6th century bc are found. Type 9 is of higher quality and is indeed quite frequent at Naukratis, which is surely connected with its dedication in sanctuaries. Cups of this shape could also be adorned with figured relief-appliqués74 or with painted figured decoration in MileA II-style (Fikellura).75 The best-known examples are of course the Ionian Little-MasterCups.76 A few of these have also been found in Naukratis and continue to be counted among the examples of highest quality in their class.77 Two further types of South Ionian cups with everted rim that are found at Naukratis are Type 10 (Fig 23) and 11 (Fig. 21).78 The emergence of these two types in the last quarter of the 7th century bc marks a significant change in the production of this class in Miletos.79 While the 7th century bc had been characterised by a multitude of different types being produced in relatively small numbers, the 6th century bc was dominated by mass production of types 10 and 11 and their variants.80 What is interesting in this respect is that at Naukratis Type 11 (Fig. 21; Group DD) is only very rarely attested, whereas Type 10 (Fig 23; Group D) is present in numerous variants. This might suggest that only products of a specialised workshop were exported to Naukratis, and the producers of Type 11 only rarely traded with Naukratis. The current state of research, however, does not yet allow us to either support or contradict this thesis. Further chemical analyses to determine the provenance of cups with everted rim are necessary to gain an overview of the clay

composition of Types 10 and 11. Perhaps these will reveal fundamentally different clay paste recipes used by different workshops, as is hinted at by the very limited NAA results obtained so far. Or it might emerge that both types were made from similar clay pastes, which might mean that they come from the same workshop. If so, the evidence from Naukratis could be interpreted differently, and a greater role be attributed, for example, to choices made by traders and customers in Naukratis. The re-evaluation of the cups with everted rims may finally help us to close a gap that had long been recognised between the historical tradition and the archaeological evidence.81 If Samos and especially the large Ionian metropolis Miletos are the main production centres of cups with everted rim, then these, together with MileA II painted pottery, would finally provide archaeological proof of a significant number of Milesian wares in Naukratis.82 The archaeological evidence would thus reflect the historically attested significance of the two sites which is also manifest in the two filial cults of the main deities of Samians and Milesians at Naukratis. Yet the attribution of this class of pottery is not quite as clearcut as is suggested in many handbooks. The scientific analyses aimed at the determination of the provenience of cups with everted rims caution against the wholesale attribution of classes of pottery to only few or even a single centre of production. According to the current state of research the production of cups with everted rims in Miletos can be considered as certain,83 not least through the analysis of a cup of Type 11 from Naukratis (sample Nauk 4, Fig. 21) which falls into the Milesian Group D. For the island of Samos, the results of analyses conducted by Hans Mommsen on the Hera mugs of Naukratis (confirming earlier results to the same effect achieved by Mike Hughes) have shown provenance Group J to be Samian.84 The cup with everted rim with a Hera-dipinto (Fig. 29) that was a part of the ritual dining pottery in the cult of Samian Hera in Naukratis can thus indirectly be attributed to Samian production,85 as can a cup with everted rim from Ephesos that has the same Samian clay composition J.86 But in addition to Samos and Miletos, production in Ephesos – a site still within the wider radius of South Ionia – has also been suspected.87 Furthermore, P. Dupont has raised the possibility of production in Klazomenai88 and Aiolis,89 and attributed examples of cups with everted rim from Naukratis to these sites (Dupont and Thomas Figs 5.60 [Klazomenai], 62–3, 100–1 [Aiolis]). And as D. Williams points out in the present volume,90 Chian workshops, too, produced their own version of these cups, decorating the rim with myrtle or laurel wreaths.91 Further examples of this class that have been analysed by Mommsen provide other interesting information. A cup from Naukratis (Figs 23) that forms a chemical pair together with a cup of exactly the same type (though lacking red and white bands) from Tel Kabri (sample TeKa 3) cannot yet be attributed to a production place with any certainty. Its chemical

Figure 29 Samian cup with everted rim with Hera-dipinto, BM GR 1911.66.23

Naukratis: Greek Diversity in Egypt | 61

Schlotzhauer and Villing provenance Group DD so far consists of only three pieces (the third is a jug found in Berezan on the northern Black Sea coast),92 with no kiln waster nor any find concentration to hint at any particular location.93 Nor do the three pieces show any chronological spread: all date to the first half to third quarter of the 6th century bc. What is clear, however, is that their place of production must have been a trading centre in the Eastern Aegean which exported its wares during this period to Naukratis, the Levant and the Black Sea. Such a distribution could plausibly be reconciled with production at Miletos. Yet in none of the places attested by our ancient sources to have been Milesian spheres of interest have only Milesian wares been found. For this reason, other East Aegean regions cannot be excluded as possible places of origin for Group DD. Still, the shapes of all three pieces are usually connected with South Ionia, and the decoration of the jug (of phase MileA II) in particular is closely associated with Miletos. The place of production of DD could thus be Miletos itself – perhaps as a third local group in addition to A and probably D – or in its immediate vicinity, or even a colony of Miletos, such as the ‘Hellespont workshops’ proposed by Posamentir.94 This production centre, not yet precisely located in the Troad or on the shores of the Hellespont, is known to have created, among other things, pottery similar to that of Milesian workshops.95 At any rate, the findspots of the two cups fit in with an assumed trade route from the Eastern Aegean to Naukratis/North Africa, which has also been suggested by Schaus, D’Angelo, Villing and Schlotzhauer in the present volume, with Tel Kabri being a stop-over en route,96 an assumption supported also by the fact that the typological range of cups with everted rim from Naukratis is a striking match for 7th and 6th century bc imports in the Levant. But as Fantalkin in the present volume argues from a more historical point of view, such direct imports in the Southern Levant seem to cease following the Babylonian disaster. It is thus possible that the cup reached Tel Kabri with Greek mercenaries, perhaps via Naukratis. From the evidence presented so far it seems likely that further production centres for cups with everted rim will be added. Indeed, if all the results referred to above are correct, it would seem that at times all the various East Greek regions produced cups with everted rims! Only for the two South Ionian centres of Samos and Miletos, however, are there also archaeological reasons, in addition to the scientific analyses, to assume a major production of this class of cup. Only here do we find a broad spectrum of types as well as a continuous development from the Geometric to the Archaic period. In no other region, neither Aiolis nor North Ionia, nor even East Doris (where a close variant was obviously developed) is a comparably complete line of development attested. Imitations, however, are quite conceivable in other regions, too, not least since other imported pottery was imitated frequently.97 Finally, it is also possible that the imitations attributed to Aiolis by Dupont may in fact have been produced in a Milesian colony, such as Posamentir’s ‘Hellespont workshops’.98 Local pottery production at Archaic Naukratis One of the most valuable results to have emerged from the archaeometric work on the pottery from Naukratis is the confirmation of local pottery production at the site itself or in its immediate vicinity.99 Much of this work is based on material in 62 | Naukratis: Greek Diversity in Egypt

the British Museum.100 It was P. Dupont who in 1983 was the first to demonstrate local, Egyptian production of Greek pottery,101 as a parallel for local Greek production in the Milesian colony of Istros (Histria) on the Black Sea coast.102 A mere few lines in a footnote refer to his important analyses of two pieces from Naukratis in the Museum of Archaeology and Anthropology of the University of Cambridge. One of them he considered to be a Late Wild Goat plate of Aiolian type (Dupont and Thomas Fig. 1, NAU 9), the other an ‘Ionian’ oinochoe with banded decoration (NAU 71). For both pieces he assumed local production with Nile clay: ‘présenté des compositions typiques du delta du Nil’.103 The present volume finally contains Dupont’s complete and detailed report on his investigations on the pottery from Naukratis and includes, among the 78 analysed pieces, also the two abovementioned sherds attributed to local production in Egypt (Dupont’s Group G).104 A further 14 products of the Naukratis workshops, located in five museums (Berlin, Bonn, Boston, Cambridge, and the British Museum, London; cf. Table 1), have now been added as a result of the recent investigations on the part of the British Museum’s Department of Greek and Roman Antiquities and the Naukratis Project of SFB 295 at Mainz University together with the laboratory of Hans Mommsen in Bonn.105 In addition, the same plate in Cambridge as analysed by Dupont was analysed once more by Mommsen and local production in Egypt confirmed (Fig. 30, sample Nauk 33 = Dupont and Thomas Fig. 1, sample NAU 9). The majority of pieces in the provenance group thus gained (QANN - see the detailed discussion in Mommsen et al, this volume) are Archaic (12 examples, see Table 1). This provides a more complete picture of the range of production of the local pottery workshop of early Naukratis. It is now clear that both painted and undecorated vessels were produced at Naukratis. To the shapes already recognised by Dupont – plate and oinochoe with banded decoration (Fig. 30, Dupont and Thomas Fig. 1, NAU 9; NAU 71) – we can now add with certainty further plates (Figs 31–33), but also dinoi (Fig. 34), small oinochoai with a ribbed neck (Fig. 35), large oinochoai with trefoil mouth (Fig. 37), amphorae (Figs 38 and 39 – from Tell Defenneh106) and lamps. In addition, there are mugs107 and cups with everted rim, which have not (yet) been analysed.108 Surely further shapes were produced which have not been preserved or not yet been discovered among the preserved material. It is particularly interesting that non-Greek shapes, too, are found as part of provenance Group QANN, such as a plate identified by J. Spencer as a 26th dynasty Egyptian shape (Fig. 41). The pottery of the Naukratis workshop has a character of its own. This is obvious already in the idiosyncrasy of the shapes as well as in their eclectic decoration. A group of three small oinochoai with a ribbed neck are a case in point (Figs 35, 36). Although jugs with a trefoil mouth and jugs with a ribbed neck are known among East Aegean pottery, the combination of both these features is not. Nor can one agree with the assessment of P. Dupont, who detected merely an Aiolian character in the plate fragment Fig. 30 (sample NAU 9 = Mommsen Nauk 33). Rather, we find here an angle-filled cross as it is know from Aiolian and North Ionian pottery, next to a flower of a form known both in Klazomenai and in Miletos before the middle of the 6th century bc.

East Greek Pottery from Naukratis: The Current State of Research

Figure 30 Painted plate, sample Nauk 33, Group QANN (Egypt – Naukratis?)

Figure 31 Painted plate, sample Nauk 25, Group QANN (Egypt – Naukratis?)

Figure 33 Undecorated plate, sample Nauk 27, Group QANN (Egypt – Naukratis?)

Figure 32 Painted plate, sample Nauk 9, Group QANN (Egypt – Naukratis?), with fragment from same vessel

Figure 34 Undecorated dinos, sample Nauk 82, Group QANN (Egypt – Naukratis?)

Figure 35 Undecorated jug, sample Nauk 14, Group QANN (Egypt – Naukratis?)

Figure 36 Undecorated jug, sample Nauk 19, Group QANN (Egypt – Naukratis?)

Figure 37Wavy-line oinochoe, sample Nauk 81, Group QANN (Egypt – Naukratis?)

Figure 38 Amphora or hydria, sample Nauk 83,Group QANN(Egypt – Naukratis?)

Figure 39 Painted amphora, sample Defe 10, Group QANN (Egypt – Naukratis?)

Naukratis: Greek Diversity in Egypt | 63

Schlotzhauer and Villing

Figure 40 Bowl with graffito, BM GR 1888.6-1.739

Such eclecticism is, of course, not without its parallels even within the workshops of East Greece itself, or indeed Archaic Greece as a whole. Motifs that are characteristic for production centres such as Corinth, Laconia, Attica etc. are on occasion adopted also in the production of other centres. In his contribution on Chian pottery from Naukratis in the present volume, for example, D. Williams points to several instances of such adoptions and influences among Chian pottery of the 6th century bc, which testify to the vivid exchange of motifs and ideas among the workshops and pottery centres of the Greek mainland and the Greek centres of the East Aegean. They include the possible migration of a Laconian painter closely associated with the Boreads Painter (575–65 bc) to Chios, and there christened Sirens Painter by Williams (Williams Figs 14–21), and the adoption of Laconian elements, such as flying winged creatures, into South Ionian pottery. Or does the adoption happen the other way round? Certainly, in this period, as already in the previous Orientialising period, much influence also moves from East to West, and Eastern traits remain influential.109 In the third quarter of the 6th century bc adoptions continue to be attested in Chian workshops, as Williams points out with reference to atticising kantharoi and chalices.110 Looking at the whole of the output so far attributed to the local Naukratis workshops, it must be noted, in addition to the interestingly broad spectrum of shapes, that there are more decorated than undecorated pieces. This may be a somewhat skewed picture, as also the locally produced pottery – which had not been recognised as such by the excavators, at least not explicitly in print111 – was subject to the usual criteria for collection and preservation applied to the all the pottery from the site by Petrie, Gardner und Hogarth.112 Consequently, only a small percentage of the local pottery of Naukratis is likely to have been preserved, and that the undecorated percentage may well once have been far higher. As for the decorated pottery, its study shows that the typical decorative schemes current in 6th century bc East Greece are present here, too: wavy lines, bands, various types of ornamental decoration, and half or completely slipped vessel. The decorative designs mostly belong to the phase A II – filling ornaments as they are found in the late Wild Goat and early Fikellura styles, and floral ornaments as we know them from Miletos. Most of the painted examples, moreover, feature a whitish slip or a distinctive thick, pinkish slip clearly intended to mimic the whitish-beige slips of pottery in the East Greek homeland. A further aspect of the local pottery from Naukratis is the addition of graffiti. Williams and Villing in their contribution on Carians at Naukratis in the present volume have already mentioned the Carian inscription on a locally produced oinochoe in Berlin (Fig. 36) that may have been an offering in a tomb. It has two nearly identical siblings from the temenos of 64 | Naukratis: Greek Diversity in Egypt

Figure 41 Egyptian plate, sample Nauk 79, Group QANN (Egypt – Naukratis?)

Apollo at Naukratis itself (Fig. 35), one of which through clay analysis has been shown to be a product of the Naukratis workshop, too. They carry the Greek inscription ‘DEKA’, which may be an abbreviation of deka/th, signifying the dedication of a tenth to the god.113 More explicit is the retrograde votive graffito to Apollo on an amphora or hydria with banded decoration:…]wpollwso emi[… (Fig. 38).114 A small bowl (not analysed but of typical Nile clay and covered with the distinctive pinkish slip of the Naukratis workshop) bears the graffito AFRO[... (Fig. 40) and may have been a dedication to Aphrodite,115 and the small undecorated dinos (Fig. 34) with the graffiti TH[… was probably a votive offering, too. There finally remains the question of figured decoration on the local pottery of Naukratis. Not a single example with figured decoration was identified among the chemically analysed pieces from Naukratis and Tell Defenneh. Nor is there – as far as we are aware to date – any piece from Naukratis in any collection in Europe or America that one might suspect to be local figured pottery.116 Nevertheless, for a long time there have been discussions regarding figured pottery from other find places in Egypt which were suspected to be local products of Greek character.117 Not all speculation in this regard has been proven right. The so-called Apries amphora (Bailey Figs 1–5, Group B),118 for example, is surely no Egyptian product in spite of the cartouches of the Pharaoh Apries. It now needs to be seen alongside the situlae from Tell Defenneh,119 which equally feature Egyptian motifs in their decoration (e.g. Weber Figs 16–17) but are also East Greek products.120 The case may be different, however, with the amphora from Saqqara.121 Its iconographic traits (John Boardman in his contribution to the present volume points, for example, to the characteristically Egyptian way of representing the bull’s horns) and its stylistic closeness to the painted pottery from the Naukratis workshop lead one to suspect that it was produced in Egypt, most likely in the same workshop, although this has to remain pure speculation until the clay of this piece, too, is examined. That products of the workshop were exported within Egypt is also suggested by the amphora from Tell Defenneh (Fig. 39), which falls into the same provenance Group QANN. This aspects will be discussed at length by S. Weber.122 This brings us to the question of the origin and localisation of the provenance Group QANN. So far we have talked of ‘the Naukratis workshop’; what is the justification for this association with Naukratis, as opposed to a localisation at, for example, Tell Defenneh? When did local pottery begin to be produced and by whom? Even beyond their decoration, most of the pieces of provenance Group QANN are distinctive already at first glance. Visually, their clay stands out as typically micaceous red-brown Nile silt with a grey core, very different from other Greek clays, but resembling locally produced Egyptian pottery. Yet what is

East Greek Pottery from Naukratis: The Current State of Research

Figure 42 Hellenistic painted vessel, sample Nauk 17, Group QANN (Egypt – Naukratis?)

Figure 43 Hellenistic kiln furniture, sample Nauk 15, Group QANN (Egypt – Naukratis?)

obvious, too, is that the Nile silt in the Greek examples is processed very differently to regular Egyptian clay (as exemplified, e.g., in the plate Nauk 79, Fig. 41)– its much denser, finer texture clearly points to Greek potters at work who brought with them processing techniques different to those used by their Egyptian counterparts, in addition to East Greek shapes and East Greek-style decoration. The enterprise thus appears to have been an essentially East Greek one, the result of one or more potters and painters settling down in Egypt to cater to the needs of other Greeks. Egyptian involvement is not certain, at least not after, perhaps, some guidance in locating the local clay beds. That Egyptian potters may have worked in the same workshop – and produced typical Egyptian plates, according to traditional Egyptian methods of clay preparation and finishing, such as that shown in Figure 41 – cannot be excluded, though one may equally consider the joint use of the same clay beds. The origin of the potters and painters, on the other hand, is hard to pin down; as has been observed above, both the shapes and the decoration display an eclectic mix of styles that covers several East Greek regions. Egyptian influence, by contrast, is elusive, unless one counts the distinctively Egyptian way of drawing the horns on the above-mentioned Saqqara amphora. That the workshop that produced this pottery was located in Naukratis or its immediate vicinity, rather than elsewhere in the Nile Delta, is made likely by several factors. One is the chronological span that is covered by the pottery in production Group QANN, between at least the Archaic (second third of the 6th century bc) and Ptolemaic (2nd century bc) periods – the date of an attractive decorated fragment from Naukratis (Fig. 42) – since it would seem unlikely that the inhabitants of Naukratis imported their pottery over a long period of time from another, far removed site in Egypt. Another is the fact that local pottery production is securely attested in Naukratis in later periods. Writing in the early 3rd century bc, Athenaios (9.480) referred to a flourishing pottery production at Naukratis, with a gate in the potters’ quarter being called ‘Keramike’.123 A potters’ quarter had in fact also been noted by Petrie, and potters’ rubbish and kilns are marked on his map to the east of the temenos of Hera.124 Little is know about what exactly Petrie’s evidence was, but even though his description of the potters’ rubbish may be read to include Archaic material, it is more likely to be mostly Hellenistic or later.125 Kilns, kiln furniture and local pottery made from Nile clay are certainly attested through the American fieldwork at Naukratis, from at least the 3rd century bc onwards. The local potters appear to have produced imitations (particularly in a characteristic grey fabric) of Greek

Figure 44 Hellenistic kiln furniture, sample Nauk 16, Group QANN (Egypt – Naukratis?)

(Attic black-glazed) shapes, as well as some local shapes.126 A fair number of such easily identifiable local wares from the early seasons of excavations at Naukratis are also preserved in the store-rooms of the British Museum. Most importantly, however, there are strangely shaped objects with pre-firing inscriptions, which in all likelihood are some kind of pot stands or other kiln furniture inscribed, one may speculate, with a potter’s name so as to separate different lots in kiln.127 Parallels have been found, for example, in a 2nd century bc pottery workshop at Athribis.128 Two of these objects from Naukratis have now been analysed and fall into the local provenance Group QANN (samples Nauk 15 and 16; Figs 43, 44).129 They not only provide direct evidence for local pottery workshops at Naukratis in the Ptolemaic period, but also – given the extremely high likelihood of their having been made locally – securely anchor the provenance Group QANN at Naukratis. As regards the likely beginnings of this local Naukratite production, these seem to fall approximately into the second third of the 6th century bc. Perhaps we may, in fact, connect them with the re-organisation of Naukratis under Amasis as it is reported by Herodotus, or with the decades immediately following that event. Finally, it should not be forgotten that the fine wares of the Archaic ‘Naukratis workshop’ are just part of a wider picture, as is pointed out by Villing in this volume. Local imitations in marl clay of Archaic Greek (Samian and Lesbian) transport amphorae are attested at Tell Defenneh and T21.130 Mortaria made from marl clay at Naukratis and Tell Defenneh (Villing Figs 1, 19, 20, and esp. 23) and many other sites may go back to Greek or Cypriot models, and may have been produced by potters of the ‘Naukratis workshop’ working with different clay, or perhaps by different workshops at Naukratis or elsewhere, perhaps even ones run by Cypriots, Phoenicians or Egyptians. Other foreign potters at work in Egypt are certainly attested in the case of the ‘Judean juglets’, which appear to have been manufactured by Judean potters for a Judean diaspora community.131 The Naukratis workshop and related production of Greek pottery shapes in Egypt surely falls into the same category of local pottery produced by immigrants for an immigrant community. Archaic Naukratis thus emerges less and less as place where traders merely passed through going about their trade and depositing offerings in sanctuaries, but as centre of production in its own right, with a pottery workshop now having joined the already attested workshops for scarabs and faience, the likely terracotta and sculpture workshops, and the possible alabaster alabastra and perhaps even flower garland workshops.132

Naukratis: Greek Diversity in Egypt | 65

Schlotzhauer and Villing Illustration credits

Figs 17b, 26b, 37b, 38b, 40a, 41b K. Morton; Figs 23b, 24b, 27, 28, 30, 32b, 34b, 35b, 36, 44b U. Schlotzhauer; Figs. 18, 31, 33 ©2004 Museum of Fine Arts, Boston; all remaining photographs the British Museum.

Notes * 1

2 3 4 5

6

7 8 9 10

11 12 13 14 15

16 17 18 19 20 21

We are grateful to Dyfri Williams, Michael Kerschner, Richard Posamentir and Sabine Weber for helpful comments on this essay. See also Gardner 1888, 15; who refers to large numbers of painted and undecorated fragments coming to light every day: ‘I accordingly had to content myself with glancing through the contents of each basket, to make sure that it really consisted of painted fragments of vases, and had no rubbish put in to fill.’ For a provisional list of collections with material from Naukratis, see Villing and Schlotzhauer, this volume, n. 12. On the significance of the finds from Naukratis for scholarship of the late 19th and early 20th century, see e.g. Cook 1997, 296-9, 305. E.g. Boehlau 1898; Prinz 1908; Price 1924, 1928. For a discussion of scholarship on the pottery of Naukratis, see also in particular Kerschner 2001; Schlotzhauer 2001a. See e.g. Stevenson 1890/1. Two recent publications in particular summarise the results of this development: the late R.M. Cook's magisterial survey on East Greek pottery (a joint publication with Pierre Dupont), and Sir John Boardman's handbook on Early Greek vase-painting: Cook and Dupont 1998; Boardman 1998b. Akurgal et al. 2002; Coldstream and Liddy 1996, 480-1; Dupont 1983, 1986, 2000; Hertel et al. 2001; Harbottle et al. 2005; Hughes et al. 1988; Jones 1986; Kerschner et al. 1993, 2002; Kerschner and Mommsen 2005; Kerschner 2006; Mommsen, Kerschner and Posamentir 2006; Mommsen, Schwedt und Attula 2006; Posamentir and Solovyov 2006; Schlotzhauer 2006; Seifert 1998, 2004; Seifert and Yalçin 1996. Dupont 1983. Dupont 1983, 36. Dupont 1983, 36 n. 38, where he refers to this series of analyses for the first time and also mentions two fragments that might have been produced locally at Naukratis. Jones 1986, 698-702, 663 pls 8.11-13. Mostly samples in Oxford were analysed, but some pieces in the British Museum were also included; see ibid., 701 table 8.18. The three situla fragments in Oxford are wrongly indicated as being from Naukratis; in fact they are from Tell Defenneh. Jones 1986, 662-3. Oxford, Ashmolean Museum, Inv. 1925.608 a-c; Beazley et al. 1931, pl. G.B. 401.25-7; Jones 1986, 669-70. Dupont and Thomas, this volume; cf. Weber, this volume. The new results by Mommsen, Weber, Schlotzhauer and the British Museum, by contrast, show the situlae in a chemical group of their own, to which none of the analysed Vroulian vessels belong. Hughes et al. 1988, 475; Mommsen et al., this volume. A later batch taken by Hughes, analysed independently, included 14 black-glaze and black-figure pottery samples from Naukratis in Oxford and the British Museum; the results of this analysis have recently been published, establishing that all analysed samples are of Attic production: Harbottle, Hughes and Seleem 2005. Unfortunately, the descriptions of the analysed pieces given in the article (table 1 on p. 513) are incomplete or flawed; as regards the ten pieces in the British Museum (all from Naukratis), no. 16 is from a 6th century bc bf dinos stand (Venit 1982, 481 no. B839); no. 17 is from a bf dinos, circle of Sophilos ; no. 18 is from a rf column-krater, ca. 500-490 bc (Venit 1982, 486-7 no. C8); no. 20 is from a bf dinos stand, circle of Sophilos (Venit 1982, 480-1 no. B838; cf. Williams 1983b, 15-16 fig. 10); nos 19 and 21-25 are all from black-glaze vessels. Of the Ashmolean Museum pieces, two (nos 7 and 9) are catalogued among the Attic material in Venit 1988, 339 no. B407, 435 no. B702, pls 171 and 214. Petrie 1886b, 13. Petrie 1886b, 23. See also Villing, this volume, and Johnston 1982, 357, and this volume. Edgar 1905, 123-6. Petrie 1886b, 42, reports the find of some 1200 such pieces, which were all brought back to Britain. Villing, this volume, n. 196; cf. also Kerschner 2001, 75-6. For a summary of Attic, Laconian and Corinthian pottery at Naukratis, see esp. Möller 2000a, 119-27.

66 | Naukratis: Greek Diversity in Egypt

22 Venit 1984; Schlotzhauer 2006, 294-301, esp. 301 with n. 61. 23 For clay analysis of black-glaze pottery from Naukratis, see supra n. 15. 24 Notably Beazley and Payne 1929; Venit 1982, 1988. 25 For a summary of the current state of knowledge on North Ionian pottery see especially the recent work by Kerschner and Mommsen: Kerschner 2001, 85-7; Kerschner in Akurgal et al. 2002, 63-92; Kerschner 2006; Posamentir and Solovyov 2006; see also for Klazomenai: Ersoy 1993, 2000, 2003, 2004; Özer 2004; Hürmüzlü 2004a. 26 The simpler plate Nauk 37 is clearly related. Note also the prominent use of added red and white dots in many of the examples in this group. 27 Özer 2004, 215 n. 15; cf. Bailey, this volume. 28 Walter-Karydi 1973, 68-9, cat. no. 732, pl. 99.732. 29 Note, e.g., a similar plate which may have been produced in the same workshop:Walter-Karydi 1973, 68, cat. no. 731, pl. 97.731. 30 Walter-Karydi 1973, 70, cat no. 819, pl. 100.819. 31 Doubts had already been expressed by Lemos 1991, 179. Further examination of this topic is envisaged for the future. 32 Kerschner 2001, 79. 33 Kerschner in Akurgal et al. 2002, 63-72, 97-105. 34 Cook 1954, pls G.B. 587.18 and 590.15. 35 Cook 1954, pl. G.B. 586. 36 Kerschner in Akurgal et al. 2002, cat no. 88; cf. also an amphoriskos with scale pattern: ibid. cat no. 86; N-Ionian bf column krater: ibid. cat no. 87. 37 Cf. Kerschner 2006. 38 Weber in Schlotzhauer and Weber (forthcoming). 39 Walter-Karydi 1973, pl. 123.996. 40 Paspalas, this volume, n. 68; Kerschner, this volume, n. 99; Akurgal et al. 2002, 93-4. 41 Ersoy 2003, and see above. 42 Several dozens of sherds are preserved in the British Museum. Cf. also a Grey ware fragment with a representation of a Sphinx once in the collection of von Bissing. Prins de Jong 1925, 23, 55 no. V.1. 43 Grey ware pottery was, in fact, produced all along the western coast of Asia Minor (and beyond), as far south as Miletos and Samos, although here the repertoire of shapes and kinds of decoration clearly distinguishes it from the northern production of Grey wares. For a survey, see Bayne 2000. 44 Möller 2000a, 173-4 no. 1; Johnston 1978, no. 1. 45 Lamb 1932, 10 fig. 4.10. 46 See Bayne 2000, 254. 47 Bayne 2000, 254. 48 Red painted floral decoration is found especially on 6th century bc Grey ware cups, jugs, cups with everted rim and stemmed plates from a tomb on Samos (Gercke and Löwe 1996, 68-70 nos 45.41-50), as well as on lids painted in Fikellura style at Miletos (Posamentir 2002, 20-21 fig. 5, 26 cat. no. 30). Grey ware cups with everted rims (Knickrandschalen) from Miletos have been shown through clay analysis to be local products (to be published shortly by U. Schlotzhauer), and analysis has shown that Grey ware from Berezan (Posamentir and Solovyov 2006: sample Bere 136) belongs to Milesian Group A, the Kalabaktepe workshops. For Grey ware pottery produced in Ephesos, see Kerschner 1997b, 209-10. 49 Cf. Möller 2000a, 173 no. 3. For a complete example of an open-work stand from Naukratis, see Lamb 1932, pl. 1.3. 50 E.g. Lamb 1932, 8 fig. 3.7, 15. Less close (though not unrelated to other fenestrated stands from Naukratis) are the examples from Larisa: Boehlau and Schefold 1942, 128 fig. 53, pl. 48.19-22. 51 Mommsen et al. 2001; Mountjoy and Mommsen (forthcoming). 52 P.A. Mountjoy in Mommsen, Hertel and Mountjoy 2001, 178: ‘This clay recipe is used at Troy from Troy I-VIII for unpainted wares and for Mycenaean pottery. The longevity of the group suggests local production.’ 53 Group B-Troy: Mommsen, Hertel and Mountjoy 2001, 200, 201 fig. 48 (Protogeometric-Early Geometric); Group D-Troy: ibid., 178, 179 figs 7-8 (Troy VI). For Archaic parallels, see also Lamb 1932, 10 fig. 4.1 (from Pyrrha), 3 (from Troy VIII), and Utili 2002, passim. 54 Cf. Spencer 1995, 304. For Lesbian Grey wares, see esp. Bayne 2000 and Spencer 1995, 301-3; for a summary of the evidence, see also Kerschner 2001, 88 with ns 165-6. 55 Bayburtluoðlu 1978; Dupont 1983, 41. 56 Cf. also the extensive discussion by Schlotzhauer in Schlotzhauer and Weber 2005, 81; 93; Schlotzhauer 2006, 311-13. 57 Cf. also Schlotzhauer 1999, 119-22; (forthcoming b).

East Greek Pottery from Naukratis: The Current State of Research 58 59 60 61 62

63 64 65 66 67 68

69 70

71

72

73

74 75 76

Posamentir and Solovyov 2006; Kerschner 2006; Dupont 1983, 35-6. Attula 2006. Schlotzhauer 2006, 301-7, 316 figs 4-6. Cf. Schlotzhauer 2001a, 115-16. Blinkenberg 1931, 738-9 nos 3172-4, fig. 74 and pl. 150.3171-3. One of them carries a graffito underneath the foot. They are dated late by Blinkenberg and compared with Terra Sigillata shapes but the inscription on the Naukratis piece contradicts this. That the type (Petrie’s fabric group D) is indeed early is supported also by at least one example reported to have come from early levels at Naukratis (Petrie 1886b, 21, 23), even though the ‘stratigraphy’ at Naukratis is, of course, notoriously unreliable. Even though they have sometimes been considered by scholars, only the various randomly published pieces have been collected: e.g. Möller 2001a, 142-3. Cf. Beazley and Payne 1929; Beazley et al. 1931; Venit 1984. Especially Lemos 1991. E.g. Cook 1933/4. E.g. the Tübingen group, Petrie group, Enman class or the Urla group etc.: Cook 1952 and 1998, 95-107. On top of the selection already made by the excavators, and thus further distorting the picture. On the criteria for keeping pottery and on the distribution of finds, see above, and Villing and Schlotzhauer, this volume. Schlotzhauer 2001a, 113-4, 122-4. Examples for misinterpretation through the undifferentiated use of published data on the Archaic pottery from Naukratis: Bowden 1991; Sørensen 2001. The cup is presumably that illustrated in Petrie 1886b, pl. 10.10. Petrie (ibid. 19) also points out that further examples of this type were found during the excavations. There are, indeed, several East Dorian cups with everted rim, even if they do not correspond to the precise variant represented by Fig. 24; cf. also Schlotzhauer 2006, 301 n. 64. The sites of Tell Sukas and Tocra (Taucheira), can be considered representative through their wealth of finds; cf. Schlotzhauer 2001b, 27 table 1, 30-6, 39-43, 297-8. G. Ploug 1973, 27-39, distinguishes ten groups at Tell Sukas on the basis of a selection of 250 cups with everted rim (overall c. 1500 such cups, mostly fragments, of the 6th century bc, were found; cf. ibid, 95-6). On present-day knowledge most of Ploug’s groups have to be attributed to the South Ionian type; this includes Ploug’s groups 2, 3, 5, and especially the (at Sukas) quantitatively strongest groups 6 and 9. Only group 1 and several examples of group 9, which includes mostly South Ionian types, can be attributed to the East Dorian variant. Groups 4 and 7-8 consist partly of singular pieces, and group 10 contains merely handles, which taken by themselves should not make up a group of their own and which, moreover, to date cannot be attributed to any production centres. In Tocra J. Hayes distinguished 14 types with several variants, and believed the majority of his types to be Rhodian (type I to XI: Hayes in Boardman and Hayes 1966, 111-5, 120-4) and only few examples Samian, i.e. South Ionian (Samian i-iii: ibid. 115-6, 124). On present knowledge, however, only type VI and several singular pieces, which he considered variants of type IX (Hayes in Boardman and Hayes 1966, 111-6, 120-4) are not South Ionian. Once more, then, the quantitatively strongest groups at Tocra in particular belong to South Ionia. Also at sites where so far only little East Greek pottery and few cups with everted rim have emerged East Dorian types are more rare than South Ionian ones, for example at Mez.ad H . ashavyahu, Tel Kabri or Ashkelon (Schlotzhauer 2001b, 298-303). Type 5 is in some respects close to or identical with Type II/2 of Boldrini 1994, 149-51 cat. nos 249-50, pl. 4.249-50. The example presented here belongs to a sub-type, Type 5.C (Schlotzhauer 2000, 410-1; 2001b, 86-7, 326-5).The typology of South Ionian cups used here is the new typology devised by the author on the basis of the study of several thousand cups with everted rim from Miletos, see Schlotzhauer 2001b; Schlotzhauer in Kerschner 1999, 21-3 with no. 71; Schlotzhauer 2000. Schlotzhauer Type 9.A is very similar to Type B1/B2 of Villard and Vallet 1955, 23-6 and to Type II/1 (253) of Boldrini 1994, 148-52 cat. no. 253, pl. 5.253. On Type 9.A, see Schlotzhauer 2000, 409-14; 2001b, 96-7, 332-3. For the role of Type 9 in the 6th century bc see Schlotzhauer 2000, 410-1; 2001b, 123. Cf. Petrie 1886b, pl. 13.1; Price 1924, 183 fig. 2; Venit 1988, 131 fig. 175, 174 fig. 175-6, pl. 41.175-6; Möller 2000a, 143. See Schlotzhauer 2001b, 392-6; Schlotzhauer (forthcoming b). On Ionian Little Master cups see esp. Kunze 1934; but also Cook 1998,

77 78 79 80 81 82

83

84

85

86 87 88 89 90 91

92 93 94 95 96 97

98 99

92-4; Schlotzhauer 2001b, 396-402; 2001a, 123-4; (forthcoming b); Shefton 1989;Walter Karydi 1973, 24-9. Cf. Price 1924, 183 fig 3; Kunze 1934, pl. 7.1-2, Beil. 7; Walter-Karydi 1973, pl. 49.422, pl. 50.445, pl. 53.448-9; Venit 1988, pl. 42.180; Möller 2000a, 142-3. Schlotzhauer Type 10 is more or less comparable to Type B1 of Villard and Vallet 1955, 23-6. The example presented here belongs to a subtype, Type 10,2.B (Schlotzhauer 2000, 409-14; 2001b, 99, 333-5). Schlotzhauer 2001b, 134-5, 503-5; 2001a, 410-1. Schlotzhauer 2001b, 340-1, diagrams 26-7. Boardman 1980, 49; Austin 1970, 51 no. 4; Sullivan 1996, 190; Möller 2000b, 747. As suggested by Schlotzhauer 2001a. This, of course, says nothing about the actual physical presence of Milesians at Naukratis. It has long been recognised that the place of origin of finds must not be equated with that of its carriers or users; see also Villing and Schlotzhauer, this volume. Proven beyond doubt by the kiln waster of a cup with everted rim from the area of the kilns at Kalabaktepe in Miletos (Kerschner 2002, 37-8, 114 cat no. 97, 175 fig. 63; provenance Group A, ‘Kalabaktepe workshops’) as well as a further cup, Fig. 21 (Nauk 4), which, however, belongs to provenance Group D, likely to be located at Miletos as well. P. Dupont’s analyses, too, associate the production of cups with everted rim with Miletos, see Dupont 1983, 34; 1986, 60-1; 2000, 451-2, and Dupont, this volume: Dupont Fig. 6 (sample NAU 61 = Group B1 = Miletos). Cf. Mommsen et al., this volume. Already the archaeometric analyses of Dupont 1983, 33, 40, had shown Samos to be one of the main places of production of cups with everted rim, especially of the highquality series with a myrtle wreath on the rim or relief appliqués on the shoulder-band. In the meantime, however, he himself (Dupont 2000, 451, and Dupont, this volume, Group E) questions this attribution. Schlotzhauer 2006, 311-3, 318 fig. 13. From an archaeological point of view the production of this special pottery with Hera-dipinti, otherwise known only from the mother sanctuary on Samos, has never been in doubt: see ibid. 311-3; Schlotzhauer in Schlotzhauer and Weber 2005, 81. For the full range of shapes with Hera-dipinti (including cups with everted rim) in the Heraion of Samos see Kron 1984, 1988 (with further references) and Furtwängler and Kienast 1989. In the first publication Group J could not yet be located; see Kerschner in Akurgal et al. 2002, 51, 108 cat. no. 68, pl. 5.68. Kerschner (forthcoming). Dupont, this volume, Group F, sample NAU 60. Dupont, this volume, Group E, samples NAU 62-3, 100-1. Williams, this volume, esp. n. 63. They are a different phenomenon from the earlier development of the class of cups with everted rim into the special Chian Chalice shape as it has been shown, e.g., by Boardman 1967, 103 fig. 60. As is discussed below, the 6th century bc saw the exchange and adoption of forms and motifs between several workshops in several production centres. Such a process must be presumed here. Sample Bere 11, cf. Mommsen et al., this volume; it will be discussed further by Kerschner 2006. On Group DD see also Mommsen et al., this volume. Posamentir in Posamentir and Solovyov 2006. Ibid. The assumption of a stopover is also supported by the character of the remaining East Greek finds from Tel Kabri, which is similar to that from Naukratis. For the so-called bird bowls, for example, which were produced in the North Ionian pottery centre B (also termed the ‘bird bowl workshops’ after this characteristic group of cups) and which only there display a long line of development, local imitations are attested for Miletos (provenance Group D), Kyme (provenance Group G) and the North Ionian production centres E (Klazomenai?) and F (Smyrna?) as well as Ephesos (provenance Group H), see Kerschner in Akurgal et al. 2002 and Kerschner and Mommsen, this volume. Posamentir 2006 and also Dupont 2000, 452, consider the Hellespont a possibility in this context, even if they only touch on this issue tentatively and briefly. In the following we will mainly talk of a workshop in the singular, even if the existence of several workshops – certainly over time but even at any one time – cannot be excluded. For the Archaic period, however, at least during the first few decades, historical arguments

Naukratis: Greek Diversity in Egypt | 67

Schlotzhauer and Villing and the small quantity of known locally produced pottery in Naukratis make it less likely that several workshops were responsible for the pottery of Group QANN. 100 The results have been made possible through the generous interest of the staff of the Department of Greek and Roman Antiquities of the British Museum, especially its Keeper D. Williams. In addition, further local pottery was discovered and analysed in the Museums of Bonn, Boston and Cambridge (see Table 1), which will be discussed more fully in the final publication of the Mainz Naukratis Project: Schlotzhauer and Weber (forthcoming). 101 Dupont 1983, 36 n. 38. Scholars had of course, for a long time, discussed and supposed the possibility of such local production, but mostly with regard to pottery groups now shown to be Greek / East Greek, such as Chian pottery. Some, however, correctly pinpointed local products, such as the amphora from Tell Defenneh (Fig. 39) considered by Cook 1954, 39, to have been made ‘in the locality of Tell Defenneh’, and now shown to be part of the local production Group QANN. 102 Dupont 1983, 36. 103 Dupont 1983, 36 n. 38. 104 Dupont, this volume, n. 13, moreover points out that some further samples share certain features characteristic for Nile Delta pottery. 105 The authors and also D. Williams have identified further products of the Naukratis workshops in several museums, but these have not been (or could not be) analysed, including pieces in Heidelberg, Alexandria and in The British Museum. Most of them will be discussed in Schlotzhauer in Schlotzhauer and Weber (forthcoming). 106 This amphora will be discussed by S. Weber in length in Schlotzhauer and Weber (forthcoming). 107 E.g. BM GR 1965.9-30.450. 108 One such cup with everted rim likely to be of local Naukratite production is BM GR 1886.4-1.777. Another cup with everted rim from Naukratis (BM GR 1886.4-1.1034) that was suspected by Oren 1984 (27 with note 19) to be of local Egyptian production, however, is surely of East Greek manufacture and merely dark grey because it was burnt. Oren, however, presumably correctly observed a local Egyptian imitation of an Ionian cup at T.21 (Migdol): Oren 1984, 27, fig. 23.2, fig. 24; perhaps this, too was produced at Naukratis? 109 For interconnections between the Archaic pottery of East Greece, the Greek mainland and Western Greece, see also Williams (forthcoming). 110 See Williams, in this volume. 111 What they considered as local is the type of pottery now identified as Chian, calling it ‘Naukratian’: Petrie 1886b, 19; Gardner 1888, 38-53; Edgar 1898/9, 57; see also Williams, this volume. See also n. 14, below. 112 See also Villing and Schlotzhauer, this volume, and in detail soon Schlotzhauer in Schlotzhauer and Weber (forthcoming). 113 Schlotzhauer 2006, 310-11. 114 Perhaps this piece falls into Petrie’s fabric group B1 (‘rough red brown clay, black stripes (earliest pottery from well, retrograde inscriptions)’), which, together with fabric groups B 5 (‘rough redbrown clay, coarse’) B6 (‘rough red-brown clay, very coarse red, white face’) and possibly other sub-groups of B, probably refer to locally produced vessels: Petrie 1886b, 17, 19, 21. 115 BM GR 1888.6-1.739. A modern incised inscription on the piece, ‘CEM’, puzzlingly seems to suggest the cemetery of Naukratis as the findspot, which is hard to reconcile with it bearing what looks like a votive inscription to Aphrodite. A second very similar rim fragment (BM GR 1888.6-1.291) with the graffito ... ]DIT[... does not join, even though it is very tempting to see them as belonging to the same bowl. 116 Harbottle, Hughes and Seleem 2005 avowedly set out to establish whether it could be determined if local Greek pottery workshop at Naukratis produced black-figure pottery, but the material chosen by them all emerged as Attic – perhaps hardly surprising given the rather obviously Attic nature of many of the pieces, including fragments attributed to well-known Attic vase-painters (see supra, n. 15). From the analyses by Dupont and especially Mommsen it is now clear that the local pottery of Naukratis looks very different from Attic pottery. 117 Boardman 1980, 133-41; Boardman, 1998b, 144 with 158 figs 305-6, 222 with 256 fig. 500; Boardman, this volume.

68 | Naukratis: Greek Diversity in Egypt

118 Cf. Boardman 1998b, 144; but now see Bailey, this volume; Mommsen et al., this volume (sample TbEgy 1). 119 See the discussion by Bailey, this volume; Weber in Schlotzhauer and Weber 2005, 86-91 esp. 93; Weber this volume; Weber (forthcoming); Weber in Schlotzhauer and Weber (forthcoming). The case is similar to that of the fragments of a black-figured amphora from Karnak in Oxford (Boardman 1998b, 220 with fig. 487). Here, too, it seems most likely that a North Ionian workshop produced the piece to order with the destination Egypt in mind. The iconography of the carrying of the boat of Amun is an Egyptian motif otherwise unknown in the Greek world, but the foreign elements are mixed with Greek ones and are shaped and interpreted against the background of Greek experience, presumably the carrying of a Dionysus ship in a Greek procession. 120 The production even of these pieces in Naukratis by East Greek potters and painters still cannot be excluded, of course, if one assumes that clay from their home cities was imported into Naukratis. This is still discussed particularly in connection with Chian pottery, yet the lack of evidence (the interpretation of the imported earth in the Elephantine palimpsest as potters’ clay, still upheld by Stager 2005, 251, is now discounted by many scholars; see also Mommsen and Kerschner, this volume, n. 7) and the diversity of styles and chemical clay groupings among the pieces considered in this context are potent arguments against this possibility. For a different opinion, however, see J. Boardman, this volume. 121 E.g. Boardman 1998b, 158 fig. 305. The amphora from Saqqara will be discussed in detail by S. Weber in Schlotzhauer and Weber (fortcoming). 122 Weber in Schlotzhauer and Weber (forthcoming). 123 ‘Excellent kylikes are also made in Naukratis, the native city of our boon-companion Athenaios. They are like phialai, made not as on the lathe but as if fashioned by the finger; moreover they have four handles and a broadly extended base (there are, by the way, many potters in Naukratis; from them also the gate which is near the potters’ workshops is called Keramike).’ 124 Petrie 1886b, pl. XLI. 125 Petrie 1886b, 22: ‘In the potters’ rubbish in the north-east of the town at 350 level were found B5 whorls [red-brown coarse fabric]; D, a fine-ribbed dish, smooth-faced; F2 [white-faced with orange lines and figures – i.e. Chiot], and same thicker; G2 [smooth unfaced, brown to red line; black inside with red and white lines (Eye bowls)]; L1 [plain buff and black bowls – cf. pl. x.4,5,6].’ 126 As Leonard 1997, 25-6, notes, potsherds and vitrified mudbrick fragments found on the hill of Kom Hadid were identified as waste products from a pottery kiln. Berlin 2001, 45-6, identifies much of the pottery from Naukratis as products of the nearby pottery workshops of Kom Dahab. On the local pottery of Hellenistic Naukratis, see Berlin 1997a; Berlin 2001. Egyptian production, in Alexandria, of high-quality West Slope and Gnathia pottery, by contrast, remains disputed: Alexandropoulou 2002, 196-7. 127 Leonard in Leonard 2001, 191-3 nos 31-2, fig. 3.6; see also Coulsen 1996, 79-81, nos 1359, 1427, 1542, 1639, fig. 44, pl. 13. For a full discussion see now Bailey (forthcoming), section ‘Miscellaneous objects’, cat. nos 3695-8. We are grateful to Donald Bailey for sharing this information with us pre-publication. 128 Cf. Bailey, ibid. 129 Oren 1984, 28, had already (correctly) claimed the existence of locally made pottery vessels shaped after Greek types among the unpublished material from Daphnae and Naukratis. 130 At least for Tell Defenneh this is certain, as visual observation clearly identifies the amphora BM AE 22333, of Samian shape, as local. For further discussion, see Weber in Schlotzhauer and Weber (forthcoming). 131 Holladay 2004; we owe this reference to S. Weber. 132 For workshops at Naukratis, see Möller 2000a, 148-54, 163-6. Tridacna shells are unlikely to have been produced at Naukratis (see Villing and Schlotzhauer, this volume, n. 51), but the manufacture of floral wreaths made of myrtle, marjoram or papyrus (?) is a possibility already from the time of Anakreon (Ath. 671e, 675f-676d; Pollux 6.107). For the ongoing debate on sculpture production in a Cypriot style at Naukratis, see most recently Höckmann in Höckmann and Koenigs (forthcoming) and Nick (forthcoming).

Neutron Activation Analysis of Pottery from Naukratis and other Related Vessels Hans Mommsen with M.R. Cowell, Ph. Fletcher, D. Hook, U. Schlotzhauer,A.Villing, S.Weber and D.Williams Abstract Abundances of 30 minor and trace elements of pottery sherds from mainly Naukratis and Tell Defenneh are presented. The compositions have been measured with the Neutron Activation Analysis (NAA), a procedure applied routinely in Bonn for many years. Many different elemental patterns often assigned to known production centres in western Asia Minor could be detected for the sherds excavated at Naukratis. All these imported vessels point to the importance of this Greek emporion in Egypt. In addition, the data reveal that pottery was also produced locally, presumably at Naukratis itself or in its vicinity. The archaeological results of these archaeometric studies are presented and discussed in separate contributions in this volume.* Neutron Activation Analysis (NAA) of archaeological pottery has been carried out routinely in Bonn for many years. The aim is to determine the production centres and places of pottery wares by comparing elemental compositions with reference material of known provenance. The assumption is that each paste prepared by the ancient potters according to a certain recipe using clay from one clay bed or mixtures of clays from several deposits has an unique elemental signature which can be traced to its origin. Our NAA method is described separately in this volume at length.1 About 80mg of pottery powder is needed, and it is taken from the sherds or vessels to be analysed with a pure sapphire drill. Our procedure determines up to 30 minor and trace elemental weight concentration values using an in-house pottery standard,2 which is calibrated with the well-known Berkeley pottery standard . The data evaluation method to search for samples of similar composition is given in Mommsen and Kerschner.3 Sample Choice Seventy-six samples from sherds excavated at Naukratis and 17 samples from sherds from Tell Defenneh, dating to the 7th and mainly to the 6th century bc and later, were taken and analysed during the years 2003–2006. In addition, nine samples of related material found at other sites have been included, one from Abusir (Abus 1), one from Thebes in Egypt (TbEgy 1), one from a non-specified site in the Nile Delta (DlEgy 1), one from Miletos (Milet 41), one from the area of Datça near Knidos (Knid 1), one from Kamiros, Rhodes (Kame 2), a further sample from Rhodes (Rhod 20), and two from Caria (Kari 1,2). A description of these samples and their current location is provided in Schlotzhauer and Villing Table 1 and an extended discussion of the choice of samples and the archaeometric results can be found in several contributions in this volume.4 Results Here, we summarize the results of our NAA measurements. This set of 102 samples is unusual in the respect that a very large

number of different concentration patterns appear, 48 in all. Twenty-two of these patterns belong to samples that are chemical loners and called ‘singles’. Each of them has an elemental pattern, which does not match any other of our patterns. Therefore, nothing can be concluded about these singles. A rate of about 15–20% of singles is often detected in NAA studies. The large number of the remaining patterns is, however, unusual especially for the site Naukratis. The pottery found there was imported from many different sites. This points, at the one hand, to the importance of this Greek emporion in Egypt having many trade or other contacts with numerous different sites, but, at the other hand, it might also be a consequence of a well-considered choice of samples for analysis or of a special selection of archaeologically questionable vessels. In our opinion, in this case, with many different patterns our in most cases univariate or, at choice (if sample numbers are large enough) multivariate statistical data evaluation procedure is especially important.5 Only with this procedure could all these many patterns be compared with our total databank, consisting of more than 6,500 samples from Greece and the Eastern Mediterranean, and similarities in compositions be detected. Such large databanks can be handled since our grouping is able to filter out all samples that are statistically similar to a pre-given composition, without any limitation of the number of samples. The total number of pottery groups consisting of more than two samples of similar composition now exceeds 200. All the samples in this study have been checked for matches with our sample and pattern databank. Calculations with such a large number of samples using Principal Component Analysis (PCA) or cluster analyses (CA), which generate dendrograms, are not feasible in practice. In addition to the already mentioned singles, only three groups of more than three sherds and five pairs of sherds have a composition that was new to us. All other remaining 18 patterns have been encountered before; the search for compositional matches can thus be considered successful. The reason is that for many years M. Kerschner has been collecting specific archaeological sample types for our analyses in order to build up our databank and to cover the most important workshops engaged in large scale and overseas trade. However, a definite production site cannot in all cases be assigned to these patterns. An overview of the groups encountered and, if known or made highly probable according to archaeological reasoning, the site of the producing workshops assigned to these groups are presented in Table 1. Table 2 gives the total numbers of group members including the new samples and the calculated average grouping values M and their spreads s (root mean square deviations) in % of M. Small differences to formerly published patterns are due to the increased number of group members. The measured individual concentration values of the samples described here can be found at our website homepage.6 Naukratis: Greek Diversity in Egypt | 69

Mommsen et al. In Figures 1 and 2 the results of discriminant analysis (DA) calculations are shown. All the samples of our databank that are members of the larger groups in Table 1 have been included in these calculations (Table 1, group nos 1–14, 19, and 21). In addition, one group X assigned to an Ephesian origin7 has been incorporated, too. The samples described in this study are represented by black dots. In Figure 1 only five clusters have been defined as input (the names given are the names of the groups in Table 1): all samples of the Cypriot groups (group numbers in Table 1: 1–3), of the Egyptian group QANN (21), of a group of unknown origin TD (19), of the Attic group KROP (14) and of all the groups assigned to western Asia Minor (4-13, X). The calculations are performed using all the elements given in the data tables except As, Ba, and Na.8 The clusters are well separated. The large cluster of the groups from western Asia Minor is treated separately in a second DA calculation with higher resolution. Figure 2 shows the result. The overlapping groups B–E and EMEB–EMEb are resolved in higher projections (not shown here). In the following only some archaeometric remarks concerning the different groups and their assignments to production places, if known, will be made in the sequence of the list of Table 1. A more extended discussion is given in the separate archaeological contributions in this volume.9 Patterns that can be exactly or very probably geographically located The four samples Nauk 35, 55, 67, and 68 (Villing Figs 2, 8, 21 and 22) are imports from Cyprus. Patterns CYPT (1) and EMEA (2), EMEa (3) have been discussed by Åström10 and Attula,11 respectively. These three patterns have a general Cypriot composition and a provenance of the members of these groups from Cyprus is without any doubt. Since we still do not have many reference samples from specific different sites of Cyprus, an accurate assignment to places of origin there cannot be made. Samples Nauk 51 (Schlotzhauer and Villing Fig. 24) and Knid 1 (Attula Figs 9–10) belong to the not very different groups EMEB (4) and EMEb (5), respectively, and are made most probably locally in the area of Emecik/Knidos.12 The three Hera mugs Nauk 1, 2, and 3 and sample Nauk 72 (Schlotzhauer and Villing Figs 14–16) have a composition J (5), which is assigned most probably to a Samian origin.13 Hughes using NAA had already measured the sample Nauk 2.14 According to our repeated grouping of these data this mug belongs like other sherds from Naukratis to his Samian group L.15 But this group L has concentration values that do not match in all elements our Samian group J. Only an extended interlaboratory study comparing the single steps and the correction procedures16 of the NAA methods applied in both our laboratories and checking the different standards used may explain these differences. A comparison of NAA data with data taken with other analytical analysis methods, e.g. Optical Emission Spectroscopy (OES), is even more difficult. Therefore we did not consult the OES results of sherds from Naukratis published by Jones.17 Members of several other patterns, which represent wellknown pastes used in different larger workshops in western Asia Minor, are also found in the wares of Naukratis. The pattern A (7) (Schlotzhauer and Villing Figs 18–20) is assigned to the 70 | Naukratis: Greek Diversity in Egypt

Kalabaktepe workshops, Miletos, and D (8) (Schlotzhauer Fig. 3; Schlotzhauer and Villing Figs 21–22) can also be assigned to a Milesian origin with high probability.18 Six vessels altogether are imports from Miletos to Naukratis and one to Tell Defenneh. Pattern B (9) belongs to the ‘bird bowl workshops’. They are assumed to be located somewhere in Northern Ionia, like the workshops using the paste of pattern E (10), which are also not definitely located.19 Seven samples excavated at Naukratis (Schlotzhauer and Villing Figs 1–5) and one sample from Thebes in Egypt (TbEgy 1) – a fragment joining the ‘Apries’ amphora (Bailey Figs 1–5) – have pattern B.20 One sample (Nauk 21 ass.) is associated to the group with pattern B. Samples are referred to as being ‘associated’ to a pattern, if they have a statistically similar composition in all the elements except for one or two. If such a small deviation is due to a measurement error, which is always possible for trace elements, or if it is real, but due to a singular contamination by the ancient potter, it can be added to the group. But it might also represent a quite similar clay paste of different origin. Pattern E (10) is represented in two samples at Naukratis (+ one associated sample Nauk 28 ass.) and in four samples at Tell Defenneh (Schlotzhauer and Villing Figs 6–10). Patterns G (11) and probably also g (12) can be traced to Kyme/Larisa or their vicinity in the Aiolis (Kerschner Figs 8–11, 13–30).21 The occurrence of pattern B-Troy (13)22 in samples Nauk 62, 63, and 65 (Schlotzhauer and Villing Figs 11–13) seemed at first astonishing. This pattern, and also a second pattern, D-Troy, formed with samples from Troy, can be assigned to the area of Troy, because clay samples from deposits near Intepe not far from Troy analysed recently, have very similar compositions.23 But since pattern D-Troy was already found in a sample of Milesian type at Troy,24 and since now both patterns occurring in Troy have also been detected in comparable material from Berezan,25 hitherto unknown workshops in the Troad termed ‘Hellespont workshops’ by Posamentir,26 must be considered as the origin of these wares. An Attic provenance emerged for samples Nauk 43, 57 [group KROP (14)], and 88 [group perb (15)].27 Although the membership of these samples to these chemically not very different groups is statistically not in doubt, we report this provenance with a certain reservation, since recent measurements of some first few samples from Chios28 also show for one sample pattern KROP and for a second one pattern perb. A provenance of these samples, an Archaic Chian chalice and a subgeometric Chian skyphos, from Attica contradicts archaeological knowledge. More samples of vessels locally produced on Chios are needed to see if the patterns of Chios are chemically separable from patterns assigned to Attica. Still not localized, but known patterns The sample Nauk 18 ass. (Villing Fig. 23) has a composition, which can be assigned to an Egyptian origin [group Marl (16)]. It is associated to samples in our databank from Egypt, which, according to petrographical investigations, are made of Egyptian Marl D.29 One sample (Nauk 6; Schlotzhauer and Villing Fig. 23) is the third sherd of a previously detected compositional pair. A sample from Tell Kabri (TeKa 3)30 and one from Berezan (Bere 11)31 are made of the same clay paste; the geographical position of the workshop is still unknown. This triple is now named

Neutron Activation Analysis of Pottery from Naukratis and other Related Vessels group DD (17). The occurrence of a third member to be added to this pair is a nice example of the stability of our measurements, since the three samples have been measured at different times in the years 1994, 2003 and 2004. Also of unknown origin is sample Nauk 53 (Schlotzhauer and Villing Fig. 25). It has the same composition as a rare group of Late Bronze Age sherds from Rhodes with silver mica inclusions. The members of this group RHc1 (18) are, according to archaeological theory, assumed to have been imported to Rhodes from one of the neighbouring islands, Kalymnos or Leros, where there are also clay deposits with silver mica inclusions.32 We have no reference material from these islands. Not localized, new patterns A group of six samples from Tell Defenneh (Weber Figs 16–21), three of them situlae, forms a new pattern TD (19). There are no comparable samples in our databank; archaeologically a provenance from western Asia Minor is probable.33 The same is true for a group of four samples, two from Naukratis (Nauk 73 and 74) (Schlotzhauer and Villing Fig. 26; Johnston Fig. 11), one from an unknown find spot in the Nile Delta region (DlEgy 1), and one from Abusir (Abus 1). They form a hitherto unknown pattern called ITAN (20) of unknown provenance. A third hitherto unknown pattern is QANN (21). Thirteen samples from Naukratis, one from the Delta and one from Tell Defenneh belong to it (Schlotzhauer and Villing Figs 30–39, 41–44). This pattern is not similar, but also not very different to a pattern QANM formed with four sherds from Qantir/Piramesse and assigned most probably to an Egyptian origin.34 This can be ascertained comparing the concentration values of both groups given in Table 2. The general similarity of these two patterns QANN and QANM can be seen also in Figure 3, where the differences of the concentration values normalised to the average spread values save of both groups are plotted as a bar diagram. After a best relative fit with a factor of 1.04 of group QANM with respect to group QANN, both patterns agree in all values inside ± 3save except for the Cr values. This assignment to Egypt, especially to Naukratis or to its vicinity, is strengthened by the archaeological classification and also by the fact, that the sherds of this group cover a time range of more than a 1000 years.35 Therefore, all members of group QANN are attributed to local workshops, the ‘Naukratis workshops’.36 Sample Nauk 33 was taken from the same sherd as the sample NAU 9 mentioned in the contribution by Dupont and Thomas in this volume (Dupont and Thomas Fig. 1),37 which belongs to Dupont’s group G. Dupont also assigns this group to an Egyptian origin, although without presenting any reference group. Pairs of samples of unknown provenance – preliminary grouping As given in Table 1 there are five compositional pairs of samples. Although the paired samples agree in nearly all concentration values with small spreads, sometimes one or two elemental values measured with small experimental error disagree by an amount that may exceed an acceptable range for larger groups. For example, the Rb values of the samples of pair 4 (Nauk 8 [111 ppm] and Emec 31 [88 ppm]) are quite different. Since Rb is measured with an experimental error of ±2.4 ppm,38 and since it can be assumed, that potters homogenised their pastes well, this large deviation may point to the fact that both samples have low

statistical probability to have been made from the same paste. Also the spread of the compositions of K for this pair (2.3 ppm and 1.6 ppm, respectively) measurable with an error of ±0.03 ppm is large. But since a deviation of these two alkali elements has been encountered before,39 this pair has been formed tentatively. One of the sherds (Nauk 85) of pair 85+86 is clearly of the same type as a sherd sampled by Dupont and Thomas (their NAU 55, Dupont and Thomas Fig. 5).40 Both sherds of this pair are difficult to classify archaeologically. Dupont calls NAU 55 member of his group C1 that is of unknown provenance like our pair Nauk 85+86. This study demonstrates that a large databank of many contemporaneous samples from western Asia Minor is needed to determine the provenance of pottery originating from there successfully. Notes *

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

16 17 18 19

20 21 22 23 24 25 26

27

28

The authors wish to thank the staff of the research reactor in Geesthacht for their technical support. Mommsen and Kerschner and references therein, this volume. Perlman and Asaro 1969. Mommsen and Kerschner, this volume. Schlotzhauer and Villing, Attula, Bailey, Kerschner, Villing, Weber, Williams and Villing, all this volume. Beier and Mommsen 1994; Mommsen and Kerschner, this volume. www.hiskp.uni-bonn.de/gruppen/mommsen/top.html . Badre et al. 2006, appendix by M. Kerschner. Mommsen 2004. Supra n. 4. Åström in Mommsen, Beier andÅström 2003, 5, 10. Data of group CYPT (samples HST 7a, 7b) given. Attula, this volume; Attula 2006; Mommsen, Schwedt and Attula 2006. Supra n. 11. Kerschner and Mommsen (forthcoming); Schlotzhauer 2006, 30814. BMRL no. 4543-46388-R. Hughes, pers. comm., data unpublished, available from British Museum, Department of Conservation and Science. We thank M. Hughes for sending these and additional data. Hughes et al. 1988. Group L is formed there with samples nos 43, 44, 45, and 46. According to our repeated evaluation of these data including dilution corrections also nos 129 (a Samian amphora, BM GR 1886.4-1.1291; Johnston Fig. 21) and 130 and Nauk 2 belong to L. Mommsen et al. 1987 summarise these corrections necessary for the NAA procedure in Bonn with specific consideration to low energy (Xray) photons. Jones 1986, 700-1. Akurgal et al. 2002; cf. also Schlotzhauer, this volume. Kerschner and Mommsen (forthcoming); Kerschner 2006 discusses the probability that the workshops using paste B are situated at or in the vicinity of Teos and that pattern E might have its origin at or in the vicinity of Klazomenai. Cf. Bailey, this volume. Kerschner and Mommsen, this volume; Kerschner, this volume. Mommsen et al. 2001; Mountjoy and Mommsen (forthcoming). Mountjoy and Mommsen (forthcoming). Mommsen et al. 2001. Mommsen, Kerschner, and Posamentir (2006 forthcoming). Posamentir and Solovyov 2006; Kerschner 2006. Groups B- and DTroy are called there TROB and TROD, respectively. As long as the workshop(s) using pastes B-Troy and/or D-Troy are not located exactly, we prefer to use the plural ‘workshops’, although both pastes could very well originate from the same workshop or an assemblage of workshops in a pottery production centre. Mommsen 2003; Harbottle, Hughes and Seleem 2005 present an Attic compositional pattern formed with sherds from Naukratis measured with NAA by the archaeometry group Brookhaven. We did not make an interlaboratory study with this laboratory, as, like in the case of the data of the British Museum of Hughes, these data can not be compared directly with our data. In collaboration with M. Kerschner, unpublished.

Naukratis: Greek Diversity in Egypt | 71

Mommsen et al. 29 Goren, pers. comm. The sample, from a mortarium, is discussed by Villing, this volume. 30 Kempinski 2002, 231, fig. 5.94:2, TeKa 3: Ionian cup reg. no. 5414/100. 31 Kerschner 2006. Bere 11 is a sample from a jug of Fikellura style (Louvre group). 32 Marketou et al. (forthcoming). 33 Weber, this volume.

34 35 36 37 38 39 40

Mountjoy and Mommsen 2001, 134, 139, group there called Mqan. Schlotzhauer and Villing, Table 1, this volume. Cf. Schlotzhauer and Villing, this volume. Dupont and Thomas, this volume. Mommsen and Kerschner, this volume. Mommsen et al. 1996. Dupont and Thomas, this volume.

Table 1 List of samples analysed, their assignment to the chemical groups, their provenance, and their individual fit factors [= dilution factors, in ( )] with respect to the groups (ass. = associated to the group, see text) 1. Group CYPT, 1 sample, Cyprus: Nauk 35 (1.08) 2. Group EMEA, 1 sample, Cyprus: Nauk 55 (0.96) 3. Group EMEa, 2 samples, Cyprus: Nauk 67 (1.03), 68 (0.99) 4. Group EMEB, 1 sample, (most probably) Knidian peninsula: Nauk 51 (1.02) 5. Group EMEb, 1 sample, (most probably) Knidian peninsula: Knid 1(0.98) 6. Group J, 4 samples, (most probably) Samos: Nauk 1 (0.95), 2 (0.92), 3 (0.96), 72 (1.01) 7. Group A, 4 samples, Kalabaktepe workshops, Miletos: Nauk 7 (1.02), 26 (0.98), 32 (1.21), 42 (0.94) 8. Group D, 3 samples, (most probably) Miletos: Nauk 4 (0.97), 39 (0.96), Defe 11 (0.96) 9. Group B, 8 samples (+ 1 ass.), Bird Bowl workshops, Northern Ionia (probably Teos): Nauk 10 (0.99), 21 ass. (1.08), 22 (1.06), 24 (0.99), 37 (0.94), 54 (1.08), 76 (1.03), 87 (0.98), TbEgy 1 (0.95) 10. Group E, 6 samples (+ 1 ass.), Northern Ionia (probably Klazomenai): Nauk 20 (0.94), 23 ass. (1.01), 58 (1.02), Defe 7 (0.96), 9 (0.96), 13 (1.00), 15 (0.97) 11. Group G, 2 samples, Aiolis, Kyme/Larisa: Nauk 12 (1.00), 13 (1.00) 12. Group g, 2 samples, Aiolis, Kyme/Larisa: Nauk 64 (1.02), 77 (0.95) 13. Group B-Troy, 3 samples, Troad (Hellespont workshops): Nauk 62 (1.04), 63 (0.95), 65 (1.09) 14. Group KROP, 2 samples, Attica (questionable, Chios(?), see text): Nauk 43 (1.00), 57 (0.89) 15. Group perb, 1 sample, Attica (questionable, Chios(?), see text): Nauk 88 (1.04) 16. Group Marl, 1 sample, general Egypt: Nauk 18 ass. (0.94) 17. Group DD, 3 samples, unknown: Nauk 6 (0.97), Bere 11 (1.05), TeKa 3 (0.99) 18. Group RHc1, 1 sample, unknown: Nauk 53 (0.85) 19. Group TD, 6 samples, unknown: Defe 1 (0.90), 2 (1.01), 3 (1.07), 4 (1.06), 5 (1.02), 8 (0.93) 20. Group ITAN, 4 samples, unknown: Nauk 73 (0.93), 74 (1.14), Abus 1 (1.32), DlEgy 1 (0.77) 21. Group QANN, 14 samples (+ 1 ass.), unknown, most probably local Egyptian (Naukratis workshops): Nauk 9 (0.99), 14 (1.03), 15 (1.00), 16 (1.06), 17 (1.03), 19 (0.95), 25 (0.98), 27 (0.96), 33 (1.02), 34 ass. (1.18), 79 (1.06), 81 (1.01), 82 (0.94), 83 (1.00), Defe 10 (0.96)

72 | Naukratis: Greek Diversity in Egypt

Neutron Activation Analysis of Pottery from Naukratis and other Related Vessels 22. Pair 1, unknown: Defe 16 (1.00), 17 (1.00) 23. Pair 2, unknown: Nauk 69 (1.01), 70 (0.99) 24. Pair 3, unknown: Nauk 85 (1.00), 86 (1.00) 25. Pair 4, unknown: Nauk 8 (1.00), Emec 31 (1.00) 26. Pair 5, unknown: Nauk 66 (1.00), Kari 2 (1.00) 27. Singles, 22 samples, unknown: Nauk 5, 11, 28, 29, 30, 36, 41, 44, 47, 52, 56, 59, 78, 80, 84 Defe 6, 12, 14, Kari 1, Kame 2, Milet 41, Rhod 20

Table 2 Patterns of the groups of this study with more than two samples Average concentrations of elements M measured by NAA in µg/g (ppm), if not indicated otherwise, and spreads s in percent of M. The individual data of each sample have been corrected for dilution with respect to M (best relative fit factors see Table 1). CYPT 4 samples M +/- s(%) As Ba Ca % Ce Co Cr Cs Eu Fe % Ga Hf K% La Lu Na % Nd Ni Rb Sb Sc Sm Ta Tb Th Ti % U W Yb Zn Zr

382. 11.5 36. 25.3 246. 2.25 1.01 4.9 14.1 2.84 1.5 18.6 0.41 1.18 16.1 169. 48. 0.72 23. 3.31 0.52 0.59 5.1 0.51 1.74 1.68 2.33 84.4 124.

13. 2.8 8.2 26. 7.1 15. 7.4 2.0 25. 9.5 9.2 8.7 4.4 8.1 11. 50. 5.3 23. 4.9 5.6 6.7 7.7 8.9 34. 13. 12. 4.4 5.4 20.

EMEA 32 samples M +/- s(%) 8.55 424. 10.3 34.7 27.2 291. 3.32 0.87 5.13 16.3 2.81 1.48 16.2 0.36 0.88 14.7 220. 53.4 0.7 21.7 3.08 0.53 0.52 5.49 0.53 1.66 1.56 2.08 96.7 87.5

19. 23. 18. 3.3 4.4 21. 9.7 4.2 3.5 11. 8.8 12. 3.5 5.5 14. 11. 16. 11. 13. 5.1 4.2 6.2 11. 4.3 21. 15. 16. 3.8 12. 38.

EMEa 5 samples M +/- s(%) 9.97 371. 12.1 37.7 29.9 326. 2.05 1.00 5.47 20.6 3.13 1.3 17.9 0.39 1.15 17.4 242. 38.4 0.87 23.8 3.49 0.53 0.53 5.73 0.7 1.97 1.57 2.3 107. 83.8

19. 23. 19. 2.9 5.1 18. 37. 3.5 8.0 15. 2.2 15. 4.2 5.2 32. 5.7 24. 24. 10. 2.2 2.4 9.1 6.3 6.7 9.9 8.6 16. 2.5 26. 50.

EMEB 30 samples M +/- s(%) 6.56 498. 5.59 69.8 39.3 394. 7.79 1.13 5.09 19.5 4.57 2.17 33.8 0.46 0.62 27.4 409. 113. 0.91 17.9 5.02 0.98 0.69 12.8 0.47 2.24 2.26 2.89 97.8 141.

28. 14. 20. 2.8 7.6 14. 15. 3.4 3.4 11. 4.5 8.2 3.7 7.5 13. 8.2 16. 13. 16. 3.1 5.1 4.2 7.3 2.6 18. 5.6 15. 3.3 15. 20.

EMEb 7 samples M +/- s(%) 6.7 466. 5.5 64.2 48.8 547. 6.76 1.09 5.29 19. 4.16 1.9 31.6 0.42 0.61 24.8 599. 99.1 0.89 17.7 4.48 0.89 0.67 11.7 0.49 2.02 2.19 2.69 97.4 119.

23. 15. 15. 2.8 7.9 3.0 11. 2.0 3.4 11. 4.2 4.4 2.7 6.4 14. 5.8 14. 7.8 19. 1.6 9.7 4.0 9.4 2.9 16. 4.6 6.8 2.0 25. 25.

J 29 samples M +/- s(%) 28.6 490. 5.54 82. 40.4 373. 13. 1.52 6.77 34.6 5.57 2.91 40. 0.58 0.64 33.9 378. 162. 3.24 25.8 6.57 1.23 0.93 15.7 0.74 2.9 3.87 3.58 132. 185.

33. 9.4 21. 2.4 6.0 4.6 16. 3.2 3.7 46. 5.5 8.0 3.5 4.3 18. 6.7 11. 7.6 17. 3.2 6.4 5.1 6.0 4.7 39. 6.6 17. 3.8 6.8 35.

Naukratis: Greek Diversity in Egypt | 73

Mommsen et al. Table 2 cont. Patterns of the groups of this study with more than two samples A 42 samples M +/- s(%) As Ba Ca % Ce Co Cr Cs Eu Fe % Ga Hf K% La Lu Na % Nd Ni Rb Sb Sc Sm Ta Tb Th Ti % U W Yb Zn Zr

16.8 550. 4.71 108. 19.9 143. 11.8 1.45 4.41 25.3 5.65 3.54 50.9 0.54 1.13 41.3 176. 219. 1.58 14.2 8.02 1.56 1.21 26.1 0.42 4.34 4.05 4.2 85.6 198.

34. 14. 25. 3.7 11. 18. 3.1 2.8 4.0 20. 11. 5.1 3.0 5.3 18. 6.7 32. 4.5 9.8 3.4 6.5 3.9 5.5 4.2 19. 7.6 21. 4.2 11. 38.

D 22 samples M +/- s(%) 21.5 596. 7.36 98.7 28.5 232. 11.1 1.48 5.13 24.3 4.88 3.05 46.5 0.52 0.95 39.3 323. 184. 2.32 16.9 7.65 1.37 1.11 22.1 0.46 4.03 3.15 3.87 94.9 192.

24. 2 20. 5 38. 5.7 10. 12. 10. 5.5 3.9 43. 17. 6.8 6.1 5.2 30. 7.7 17. 13. 32. 4.3 7.7 5.3 7.2 7.3 21. 9.7 10. 5.0 14. 31.

B 99 samples M +/- s(%) 1.2 41. 6.25 84.1 19.7 151. 19.4 1.32 4.51 22.1 6.29 2.65 39.8 0.48 0.66 31.4 101. 148. 1.04 20. 5.95 1.25 0.8 17.1 0.54 3.38 2.74 3.27 111. 213.

69. 2 17. 35. 6.0 13. 10. 16. 4.9 6.7 22. 4.6 6.1 4.6 5.2 22. 7.8 23. 7.6 23. 4.3 11. 7.1 6.5 4.4 18. 14. 11. 3.3 15. 32.

E 29 samples M +/- s(%) 1.7 497. 5.39 79.2 25.6 217. 15.1 1.32 5.13 24.2 5.79 2.55 37.9 0.5 0.94 30.5 192. 143. 1.43 20.8 5.75 1.14 0.84 15.4 0.59 2.68 2.46 3.34 112. 186.

94. 11. 16. 2.8 4.6 6.8 12. 2.4 4.2 10. 6.0 5.7 2.2 5.7 12. 4.6 18. 5.2 22. 4.1 4.0 5.1 6.7 2.7 22. 6.3 14. 2.4 13. 31.

G 61 samples M +/- s(%) 44.7 810. 4.82 121. 27.6 188. 23.8 1.99 6.04 29.1 5.79 3.12 56.1 0.6 0.96 51. 173. 178. 4.79 21.8 9.93 1.18 1.29 21.1 0.5 4.0 2.7 4.23 119. 173.

45. 12. 17. 2.9 3.8 6.4 12. 2.5 3.5 13. 9.7 3.9 2.0 3.9 13. 4.7 26. 3.6 12. 1.8 5.1 5.2 6.5 2.6 22. 15. 11. 2.7 5.6 33.

g 28 samples M +/- s(%) 45.2 739. 5.53 103. 26.8 211. 27.1 1.68 5.55 24.5 5.4 2.92 48.1 0.53 1.05 42.2 210. 172. 4.67 20. 8.18 1.13 1.08 18.9 0.5 3.44 2.69 3.69 107. 154.

65. 12. 15. 5.0 6.4 11. 11. 5.5 4.3 25. 9.5 4.9 3.6 4.6 33. 6.7 22. 6.0 18. 5.1 8.1 5.9 6.3 5.8 28. 9.3 9.7 5.1 9.3 39.

Table 2 cont. Patterns of the groups of this study with more than two samples B-Troy 96 samples M +/- s(%) As Ba Ca % Ce Co Cr Cs Eu Fe % Ga Hf K% La Lu Na % Nd Ni Rb Sb Sc Sm Ta Tb Th Ti % U W Yb Zn Zr

34.5 711. 4.93 71. 21.5 173. 9.1 1.24 4.31 18.7 4.95 2.76 34. 0.39 0.97 27.2 143. 130. 1.96 16.9 4.99 0.86 0.7 15.9 0.45 3.37 2.74 2.6 99.4 164.

78. 24. 37. 5.8 8.9 13. 12. 6.3 4.5 19. 12. 15. 6.4 6.7 25. 9.7 17. 11. 28. 6.1 8.0 6.6 7.7 12. 21. 12. 14. 6.1 10. 36.

KROP 89 samples M +/- s(%) 31.8 466. 6.88 67.6 35.5 500. 13.8 1.2 5.27 21.3 4.37 2.62 30.9 0.44 0.64 27.8 412. 139. 1.47 22.1 5.33 0.85 0.73 11. 0.48 2.51 2.18 2.8 121. 165.

74 | Naukratis: Greek Diversity in Egypt

74. 20. 48. 4.4 9.0 18. 24. 4.5 5.7 30. 11. 9.1 4.6 6.7 35. 8.9 14. 8.2 39. 4.2 6.4 6.8 6.8 5.7 21. 18. 18. 5.6 9.8 37.

perb 6 samples M +/- s(%) 22.8 312 . 7.67 47.8 35. 469. 10.2 0.98 5.02 24.5 3.5 1.28 21.7 0.38 1.08 18.4 452. 66.1 0.94 21.5 3.79 0.74 0.59 8.05 0.41 1.61 1.49 2.3 103. 119.

81. 18. 25. 2.9 6.1 7.7 27. 3.8 11. 83. 14. 16. 4.0 5.5 30. 6.5 8.3 12. 31. 8.0 7.0 6.7 13. 4.6 25. 12. 13. 2.7 18. 51.

Marl 5 samples M +/- s(%) 8.42 672. 12.8 66.5 18.6 102. 1.39 1.32 4.41 16.5 7.16 0.95 28.2 0.39 0.5 25.7 97. 34.5 0.97 14.6 4.95 1.21 0.71 6.4 0.61 1.72 1.24 2.43 80.2 213.

42. 21. 18. 2.6 5.3 14. 19. 2.8 3.9 7.9 18. 12. 2.5 4.0 34. 3.2 48. 9.9 97. 4.7 4.1 5.9 6.0 4.2 17. 9.4 13. 3.8 6.9 22.

DD 3 samples M +/- s(%) 20.4 648. 7.28 121. 40.1 313. 10.4 1.84 6.19 26.6 4.28 2.8 55. 0.58 0.96 49.8 474. 155. 2.99 21.3 9.73 1.16 1.24 24. 0.45 4.21 2.24 4.32 105. 121.

42. 5.6 16. 4.9 9.2 5.2 9.3 5.3 2.7 15. 4.9 5.5 2.7 12. 18. 4.6 3.6 4.5 10. 2.5 4.1 5.7 9.9 11. 19. 3.7 11. 13. 8.9 77.

Rhc1 22 samples M +/- s(%) 13.6 751. 3.68 98.9 18.3 278. 11.9 1.42 4.06 23. 7.29 2.93 50.4 0.4 1.37 37.3 225. 163. 1.77 13.4 6.43 1.45 0.79 25.8 0.47 4.95 3.47 2.97 81.8 152.

57. 24. 33. 5.9 14. 14. 9.0 5.3 7.0 18. 9.0 7.5 7.0 8.9 31. 11. 27. 4.9 14. 7.1 10. 4.4 9.4 7.5 18. 13. 13. 7.1 16. 41.

Neutron Activation Analysis of Pottery from Naukratis and other Related Vessels Table 2 cont. Patterns of the groups of this study with more than two samples TD 7 samples M +/- s(%) As Ba Ca % Ce Co Cr Cs Eu Fe % Ga Hf K% La Lu Na % Nd Ni Rb Sb Sc Sm Ta Tb Th Ti % U W Yb Zn Zr

4.37 170. 6.97 44.2 51.6 563. 4.83 0.75 4.81 12.9 2.68 1.44 21.6 0.29 0.58 13.4 817. 72.1 0.41 14.8 2.69 0.64 0.47 7.68 0.38 1.55 1.45 1.79 88.1 106.

22. 36. 28. 2.2 9.5 15. 13. 5.6 8.6 18. 5.3 6.9 3.8 5.2 16. 18. 17. 13. 18. 6.3 5.2 7.6 8.9 7.4 17. 16. 18. 3.6 17. 29.

ITAN 4 samples M +/- s(%) 12.7 361. 7.12 77.7 15.9 86.2 8.46 1.23 4.39 24.6 4.65 2.84 36.9 0.43 0.82 30.1 91.6 151. 1.45 16.5 5.56 1.05 0.78 13.9 0.47 3.12 2.61 2.92 98.3 116.

12. 5.2 44. 1.4 6.9 2.1 7.1 2.2 2.5 7.9 2.6 6.8 1.1 3.3 23. 6.3 35. 3.4 14. 2.9 1.6 5.1 9.3 2.8 13. 13. 14. 2.0 9.1 35.

QANN 14 samples M +/- s(%) 2.95 547. 3.71 69.6 36.0 160. 1.52 1.98 7.39 25.0 7.55 1.18 30.4 0.54 1.37 31.6 136. 53.7 0.34 25.0 6.63 1.34 0.96 6.40 1.06 1.58 1.76 3.35 111. 281.

56. 13. 22. 3.2 5.7 3.8 5.6 2.4 2.7 11. 8.3 7.4 3.0 5.1 21. 6.5 29. 5.0 21. 2.6 5.2 4.3 5.8 5.6 6.4 14. 23. 3.9 6.3 13.

QANM 4 samples M +/- s(%) 7.30 666. 1.82 68.2 32.9 132. 1.50 1.83 6.91 21.0 6.28 1.23 31.2 0.51 1.23 32.6 100. 51.3 0.45 23.7 6.82 1.36 0.95 6.74 1.00 1.66 1.29 3.04 112. 312.

66. 15. 40. 1.4 9.1 3.3 8.0 2.0 1.6 12. 4.0 13. 2.3 33. 11. 8.5 10. 4.6 46. 3.6 3.8 4.1 10. 2.6 9.3 6.8 9.0 6.3 7.6 9.

Table 2 cont. Patterns of the sample pairs of this study pair 1 2 samples M +/- s(%) As Ba Ca % Ce Co Cr Cs Eu Fe % Ga Hf K% La Lu Na % Nd Ni Rb Sb Sc Sm Ta Tb Th Ti % U W Yb Zn Zr

21.8 804. 2.56 89.8 24.8 160. 13.7 1.43 4.71 17.7 4.53 2.57 40.0 0.42 1.74 40.5 166. 124. 3.61 19.7 6.77 0.92 0.83 21.8 0.42 3.65 4.24 2.80 112. 140.

9.7 6.9 9. 2.4 4.2 3.1 4.5 1.8 1.4 75. 7.7 6.7 1.5 4.1 3.5 9.0 26. 2.2 8.0 2.0 0.7 3.6 7.0 7.3 25. 5.8 9.8 2.1 2.2 20.

pair 2 2 samples M +/- s(%) 4.07 430. 6.14 87.6 25.4 204. 14.0 1.39 4.47 21.3 6.64 2.16 40.6 0.52 0.95 32.3 91.9 123. 1.11 18.7 6.03 1.13 0.79 14.7 0.66 2.74 2.21 3.29 90.9 141.

12. 3.7 3.5 3.7 12. 4.0 6.6 2.2 1.8 8.5 0.9 1.0 1.7 19. 0.5 2.8 63. 2.3 15. 0.5 1.5 3.6 3.3 0.4 11. 7.4 6.2 1.6 27. 21.

pair 3 2 samples M +/- s(%) 4.77 388. 9.68 81.8 20.2 125. 7.50 1.26 4.33 23.7 4.92 2.39 39.1 0.41 0.82 31.5 276. 144. 0.76 16.9 5.94 1.11 0.76 12.7 0.43 3.58 2.31 2.92 109. 73.0

12. 4.0 2.4 3.2 30. 4.2 11. 1.8 1.8 3.8 3.6 2.0 1.8 24. 3.0 2.9 37. 4.2 14. 0.6 0.8 3.7 3.4 0.5 11. 2.6 4.3 4.8 24. 50.

pair 4 2 samples M +/- s(%) 16.0 474. 7.79 81.7 34.1 329. 9.70 1.50 5.95 25.4 5.53 1.96 39.9 0.55 1.06 35.8 303. 99.9 2.25 23.4 6.86 1.19 0.84 15.0 0.60 2.59 3.23 3.51 114. 186.

35. 9.7 25. 2.1 1.6 5.4 1.1 1.8 5.8 13. 3.8 26. 2.8 3.1 9.5 6.7 12. 16. 12. 2.1 7.8 4.9 6.2 2.6 14. 3.9 12. 1.7 2.3 16.

pair 5 2 samples M +/- s(%) 6.21 479. 4.65 101. 31.2 262. 8.82 1.55 5.37 23.9 7.18 2.61 47.1 0.59 1.09 39.3 453. 153. 1.25 19.2 7.24 1.43 0.99 19.7 0.74 2.68 2.50 4.08 123. 162.

14. 5.3 24. 6.0 3.4 7.7 8.5 7.8 4.6 5.8 3.8 0.7 5.6 2.9 4.7 4.3 8.4 6.4 38. 2.4 2.4 3.6 9.6 0.9 53. 2.8 7.0 1.2 1.8 25.

Naukratis: Greek Diversity in Egypt | 75

Mommsen et al.

Figure 1 Result of a discriminant analysis (DA) calculation of 600 samples assuming 5 clusters (see text). Samples included in this study are shown as black dots. Plotted are the discriminant functions W1 and W2 which cover 93 (76 + 17) % of the between-group variance.The ellipses are the 2s boundaries of the groups.The Egyptian cluster QANN is well separated from clusters originating from other regions (Cyprus,Attica, western Asia Minor, unknown cluster TD).The large cluster of western Asia Minor is treated in a separate DA, see Figure 2.

Figure 2 Result of a discriminant analysis (DA) calculation of 450 samples of the cluster ‘western Asia Minor’ in Figure 1, assuming now 10 separate clusters as named in the Figure (TROB = B-Troy). Plotted are the discriminant functions W1 and W2, respectively, which cover 76 % and 11 % of the between-group variance. The ellipses are the 2s boundaries of the groups.All clusters are well separable, overlapping clusters (e. g. B and E) are resolved in other projections.

Figure 3 Graphical comparison of chemical compositions of the two groups QANN and QANM given in Table 2. Plotted are the differences of the average concentration values of the two groups normalised to the averaged standard deviation (spread) save.The values of group QANM have been multiplied by the best relative fit factor 1.04 with respect to group QANN. Both groups have a generally similar composition except for Cr and can be assigned with high probability to an Egyptian origin (see text).

76 | Naukratis: Greek Diversity in Egypt

Naukratis: Les Importations Grecques Orientales Archaïques Classification et détermination d’origine en laboratoire

Pierre Dupont et Annie Thomas Abstract Some 30 years ago a batch of 85 specimens of Archaic East Greek wares from Naukratis, mostly in the collection of the Museum of Classical Archaeology in Cambridge, were sampled for archaeometric analysis under the guidance of the late R.M. Cook. The aim of these analyses was not to cover the whole range of Archaic wares from the site, but to test the validity of the general classification of results obtained from Archaic East Greek finds from Istros. The batch of samples included the main styles of painted pottery: Chian, Late Wild Goat, North-Ionian black-figure and Fikellura, supplemented by some other specimens of Aiolian Wild Goat, Ionian bowls, Ionian and Vroulian cups, banded ware, grey ware, lamps as well as a single piece of a situla said to come from Tell Defenneh. The results of the archaeometric analysis of these samples are reported here. Les céramiques ioniennes de Naukratis monopolisent depuis longtemps l’attention face aux autres catégories importées, en raison du rôle déterminant joué par onze des principales cités de Grèce de l’Est (Milet, Samos, Chios, Clazomènes, Téos, Phocée, Mytilène, Cnide, Halicarnasse, Phasélis, Rhodes), dans la fondation et le développement de ce comptoir commercial hellénique en terre d’Egypte.1 Le propos du présent travail va être d’examiner, à la lumière de quelques séries d’analyses physico-chimiques effectuées au Laboratoire de Céramologie de Lyon, la provenance de ces trouvailles céramiques de type grec oriental au cours de l’époque archaïque, qui a constitué l’âge d’or du site sous les e pharaons de la 26 Dynastie. Les résultats dont nous allons rendre compte ont été obtenus il y a de nombreuses années déjà, dans le cadre d’une vaste enquête sur les centres producteurs de la Grèce de l’Est archaïque à partir des trouvailles d’un site colonial du PontEuxin – en l’occurrence Histria – et d’échantillonnages de référence collectés sur les principaux sites de fabrication potentiels de Grèce d’Asie (Phocée, Pergame, Çandarlý, Kymè, Myrina, Smyrne, Clazomènes, Erythrées, Colophon, Milet, Ephèse) et sur quelques sites de consommation (Larisa / Hermos) de Grèce d’Asie et des îles proches (Lesbos, Chios, Samos, Rhodes, Cos).2 A cette époque là, l’intérêt d’analyser une série d’échantillons de Naucratis résidait surtout dans une comparaison avec les approvisionnements d’Histria. Or, l’échantillonnage dont j’ai pu disposer, grâce à la bienveillance du regretté Robert Cook, alors en charge de la collection du Museum of Classical Archaeology de Cambridge, était loin d’être aussi diversifié que celui dont j’avais pu bénéficier pour Histria. En effet, il consistait essentiellement en représentants des principaux styles peints de la Grèce de l’Est: style de Chios (alias ‘Naucratite’), style des Chèvres Sauvages, style nord-ionien à figures noires et style de Fikellura, le reste des échantillons étant assez disparate (coupes ioniennes, céramique grise, lampes...).

Pour cette raison, notre échantillonnage de Naukratis ne pouvait a priori prétendre, en l’absence de céramiques communes et d’amphores-conteneurs, au même degré de représentativité que celui d’Histria, y compris d’un point de vue chronologique, du fait de l’antériorité du site pontique. Les analyses ont été effectuées en spectrométrie de fluorescence X au laboratoire de céramologie de Lyon (CNRSUMR 5138, Maison de l’Orient). Les premières séries de mesures n’ont porté que sur les huit éléments chimiques majeurs (CaO, Fe2O3, TiO2, K2O, SiO2, Al2O3, MgO, MnO), les dernières sur 13 éléments supplémentaires. Par suite, l’exploitation informatisée des résultats n’a pu porter que sur les huit éléments majeurs, communs à toutes les séries. Cette exploitation s’est faite en classification automatique par la méthode hiérarchique ascendante dite de l’analyse des grappes (‘cluster analysis’), qui calcule la distance mathématique entre les échantillons: plus celle-ci est faible entre deux individus, plus leur indice de similarité est élevé. La partition obtenue se présente sous la forme d’un diagramme arborescent ou dendrogramme, qui n’est jamais rien qu’une sorte d’arbre généalogique à l’envers, dont les échantillons engagés forment les extrémités: plus l’indice de similarité entre individus est élevé, plus les ramifications se trouvent placées bas sur le dendrogramme. L’interprétation de tels diagrammes, simple dans son principe, s’avère en réalité délicate: il faut déterminer, d’un bout à l’autre du diagramme, à quel niveau de ramification les groupements géochimiques correspondent le mieux à la réalité archéologique; d’autre part, lorsque les dispersions des teneurs sont trop larges pour certains éléments (cas fréquent du calcium), l’ordinateur a tendance à faire éclater un même groupe géochimique en plusieurs sousgroupes purement artificiels, en réalité complémentaires; il faut être particulièrement attentif aussi aux individus présentant des valeurs extrêmes au sein de chacun des groupements obtenus, car ceux-ci revèlent souvent des mal classés ou des classés ‘faute de mieux’, notamment dans le cas d’individus appartenant à des groupes mal représentés au sein de l’échantillonnage; en général, de tels individus sont en position instable et leur insertion au sein du dendrogramme peut varier à la moindre modification de l’effectif soumis à la classification automatique; d’où l’importance des recoupements systématiques de tris. Quant aux marginaux, ils sont rejetés du côté droit des principaux groupes et, pour les plus déviants, à l’extrême-droite du dendrogramme, où ils se rattachent très haut au tronc de l’arborescence. On obtient de la sorte un premier aperçu de la partition d’ensemble de l’échantillonnage. Toutefois, les classifications obtenues par analyse des grappes sont surtout valables au niveau du groupe. Pour affiner les résultats et parvenir à des attributions individuelles plus fiables, on a recours à l’analyse discriminante quadratique, de maniement plus délicat, laquelle va fournir une probabilité d’appartenance à chacun des groupes du réseau de références Naukratis: Greek Diversity in Egypt | 77

Dupont et Thomas locales. Quelle que soit la méthode de tri utilisée, les déterminations d’origine ne peuvent être obtenues qu’indirectement, par confrontation avec un réseau de références locales sûres, adapté au problème à traiter. De la qualité et de la densité de ce réseau dépendra en grande partie la fiabilité des attributions. A Lyon, notre banque de données sur la Grèce de l’Est rassemble plusieurs centaines d’échantillons de référence collectés sur les sites mentionnés plus haut. Même si elle comporte encore certaines lacunes, on peut estimer qu’elle assure déjà une couverture satisfaisante des principales zones de production potentielles. Après avoir passé en revue les différentes catégories de matériels composant notre échantillonnage de Naukratis, nous en examinerons un dendrogramme de tri, dont le dépouillement va nous livrer les principaux groupes présents sur place. Puis, à partir de ce canevas, nous rattacherons, dans la mesure du possible, ces différents groupes à des centres producteurs, à la lumière des confrontations recoupées avec notre réseau de références locales pour la Grèce de l’Est. Enfin, nous nous livrerons à une comparaison sommaire avec la situation observable en mer Noire sur le site d’Histria. Notre échantillonnage de Naukratis est composé essentiellement de spécimens des collections du Museum of Classical Archaeology de Cambridge, augmenté de quelques autres pièces d’origine diverse. L’effectif total des tessons analysés est de soixante-dix-huit échantillons à ce jour. Le style des Chèvres Sauvages est représenté par une série de quatorze

fragments, pour la plupart du ‘Late Wild Goat’ canonique de R.M. Cook (Fig. 1: NAU 1-10; Fig. 4: NAU 52; Fig. 6: NAU 76), plus un de faciès éolien (Fig. 4: NAU 74) et un autre de faciès dorien (Fig. 4: NAU 53); le style clazoménien à figures noires par quinze pièces (Fig. 1: NAU 11-19; Fig. 4: NAU 51); celui de Chios par vingt-trois individus, essentiellement des fragments de calices variés (Fig. 2: NAU 20–39), plus un à décor ‘Middle Wild Goat II’ (Fig. 4: NAU 54) et deux du ‘Polychrome Style’ de Boardman (Fig. 4: NAU 56–57); celui de Fikellura par un assortiment varié de 11 également (Fig. 3: NAU 40–50); cet échantillonnage a pu être complété de quelques coupes ioniennes des types de Vallet-Villard (deux de B1 et deux de B3) (Fig. 5: NAU 60–63, 100–101) et de Vroulia (deux exemplaires) (Fig. 5: NAU 58–59), d’un fragment de coupe vroulienne du ‘style ancien’ de Kinch surcuite (ou, selon la suggestion de D. Williams and A. Villing, d’un skyphos de type, mais pas production, corinthien) (Fig. 5: Nauk 55), d’un fragment de bol ionien du type à rosettes (Fig. 5: NAU 64) et d’un autre à décor ‘Late Wild Goat’ (Fig. 5: NAU 65), d’un tesson d’épaule d’œnochoé ionienne fine de type ‘schwarzbunt’ à frise de languettes incisées sur fond de vernis noir et à rehauts grenat, proche du style de Vroulia (Fig. 6: NAU 66); d’un autre, à pâte siliceuse et décor de bandes (NAU 71), de deux fragments de lampes ioniennes (Fig. 6: NAU 72-73) et de cinq fragments de céramique grise (Fig. 6: NAU 67, 75); enfin, nous avons inclus à l’effectif un tesson de situle du type dit ‘de Daphnae’ donné comme provenant d’Egypte et appartenant au groupe C de Cook3 (Fig. 6: DEF 1).

NAU 21

NAU 20

NAU 1

NAU 24

NAU 4

NAU 5

NAU 6

NAU 9

NAU 10

NAU 25

NAU 27

NAU 26

NAU 7

NAU 28

NAU 8

NAU 23

NAU 22

NAU 3

NAU 2

NAU 29

NAU 31 NAU 30

NAU 11

NAU 33 NAU 34

NAU 32

NAU 13 NAU 12

NAU 35

NAU 15

NAU 14

NAU 37

NAU 16

NAU 17

NAU 18

Figure 1 Echantillons NAU 1-19

78 | Naukratis: Greek Diversity in Egypt

NAU 19

NAU 36 NAU 38

Figure 2 Echantillons NAU 20-39

NAU 39

Naukratis: Les importations grecques orientales archaïques

Figure 3 Echantillons NAU 40-50

Figure 5 Echantillons NAU 55, 58-65, 100-101

Figure 4 Echantillons NAU 51-54, 56-57, 74

Figure 6 Echantillons NAU 66-67, 72-73, 75-76, DEF 1

Après élimination de quelques individus marginaux (la plupart du temps par fixation de manganèse, un type de pollution assez fréquent en milieu d’enfouissement de type réducteur humide) (NAU 33, 34, 68, 75), le dendrogramme de classification des données d’analyse de cet ensemble (Fig. 7) a traduit une partition en 11 groupes ou sous-groupes, qui peut être interprétée comme suit: A1 rassemble la plupart des échantillons du style de Chios, y compris les deux spécimens du ‘Polychrome style’ de Boardman (NAU 19, 21, 22, 23, 24, 25, 26, 27, 28, 29, 31, 32, 35, 36, 37, 38, 39, 54, 56, 57), plus un fragment d’œnochoé du style nord-ionien à figures noires (NAU 19). Il s’agit là d’un groupe particulièrement homogène, où les fusions entre individus se font très bas. Les

compositions de ce groupe coïncident d’assez près avec celles de notre principal groupe de référence de Chios. L’attribution à ce groupe d’un tesson isolé du style nord-ionien à figures noires apparaît très incertaine, du fait ses teneurs sensiblement plus élevées en potassium et aluminium, plus proches de celles des productions d’Ionie du Nord. B1 regroupe le gros du style de Fikellura (NAU 40, 41, 44, 46, 47, 48, 50), plus un fragment de coupe ionienne Villard B1 (NAU 61). Les compositions de ce groupe correspondent à celles de l’un de nos trois principaux groupes de référence de Milet. Les fouilles récentes de Kalabaktepe ont bien confirmé les résultats de laboratoire obtenus à Lyon: le style de Fikellura est omniprésent à Milet et tous les groupes de Cook y sont

STYLE DE CHIOS STYLE "LATE WILD GOAT"

Figure 7 Dendrogramme de classification automatique

STYLE NORD-IONIEN A F.N. STYLE DE FIKELLURA BOL IONIEN COUPES IONIENNES COUPES VROULIENNES

* SITULES DE DAPHNAE

A1

B1

C1 A 2

D

B2 B3

E

F

*

C2 G

Naukratis: Greek Diversity in Egypt | 79

Dupont et Thomas représentés. Quant au spécimen de coupe ionienne B1, il correspond à une variante bien attestée sur place, parfois même avec un décor peint dans le style de Fikellura. Milet a également produit et exporté d’autres formes de coupes ioniennes (Villard A2, B2 et B3), mais moins typées. C1 se réduit à deux échantillons: l’un de pinax du style des Chèvres Sauvages attribué à la Doride (NAU 53), l’autre de cette coupe vroulienne ancienne (ou skyphos?) surcuite (NAU 55).4 Les compositions de ces deux pièces ne cadrent avec aucune de nos références locales de la Grèce de l’Est, en particulier de celles du groupe principal de Rhodes. On n’a pas affaire non plus, semble-t-il, à des productions de la vallée du Nil: le surcuit susmentionné ne devrait donc pas correspondre à un raté de fabrication, mais résulter d’une fusion accidentelle à l’occasion d’un incendie. Il est possible que l’on ait affaire ici à des productions d’un centre de Doride continentale. Il est difficile d’en dire plus, le groupe d’appartenance de ces deux tessons étant de toute évidence trop mal représenté dans notre échantillonnage. A2 ne comporte également que deux échantillons, tous deux du style de Chios (NAU 20, 30). Ceux-ci n’ont été rejetés manifestement qu’en raison de leur teneur très élevée en calcium par rapport au groupe A1, mais leurs autres caractéristiques de composition ne s’écartent pas de celles de nos références de Chios. D mêle essentiellement des échantillons du style ‘Late Wild Goat’ (NAU 1, 2, 3, 4, 5, 52) et du style nord-ionien à figures noires (NAU 11, 14, 16, 51), plus un fragment de céramique grise (NAU 69). Il s’agit là d’un groupe homogène, les analyses de laboratoire ayant clairement établi que, sur les marchés d’exportation, les deux styles sont attribuables à des ateliers d’Ionie du Nord. Toutefois, les caractéristiques géochimiques de notre groupe D s’écartent de celles de Clazomènes, même si les gammes de productions sont très voisines. Elles correspondent en revanche à celles d’un groupe bien attesté parmi les trouvailles d’Histria et dont l’origine semble devoir être recherchée du côté de Téos d’après certaines analyses préliminaires d’échantillons de ce site. B2 (NAU 45, 49) et B3 (NAU 42, 43) renferment chacun deux échantillons du style de Fikellura. Les compositions de ces deux paires distinctes se rattachent à celles de deux autres groupes locaux de Milet. Il y aurait lieu de renforcer les effectifs de B2 et B3 pour conforter ces attributions. E est beaucoup plus fourni et composite à la fois, puisqu’on y trouve côte à côte des échantillons du style ‘Late Wild Goat’ canonique de Cook (NAU 65, 66) et de sa variante éolienne (‘Atelier du Deinos de Londres’ de Kardara5 – cf. Kerschner, ce volume) (NAU 74), d’autres du style nord-ionien à figures noires (NAU 12, 13, 17, 18), plusieurs exemplaires de coupes ioniennes B2 et B3 (à ornements d’applique) (NAU 62, 63, 100, 101), ainsi qu’un fragment de lampe (NAU 73). Après confrontation avec notre réseau de références locales, il est apparu que ce groupe est constitué en réalité de deux entités distinctes, de compositions très voisines: l’une, correspondant manifestement à un centre de fabrication d’Eolide, distinct de Phocée, rassemblant deux spécimens ‘Late Wild Goat’ (NAU 65, 66) et d’autres du style nord-ionien à f.n. (NAU 12, 13, 17, 18); l’autre, attribué traditionnellement à Samos, comme renfermant surtout des coupes ioniennes fines de grande diffusion (NAU 62, 63, 100, 101). S’il s’avérait que ces deux entités forment deux 80 | Naukratis: Greek Diversity in Egypt

sous-groupes originaires de la même aire régionale, l’attribution des coupes ioniennes fines de grande diffusion à Samos s’en trouverait ipso facto remise en question. En effet, celle-ci repose encore sur des bases bien fragiles, puisque les coupes ioniennes en question ne forment qu’un groupe géochimique secondaire, même au sein des seules trouvailles de l’Héraion. On imagine les répercussions d’une réattribution du gros des coupes ioniennes fines à l’Eolide: se trouverait par exemple expliqué la particulière fréquence de ce type de vase à boire en Méditerranée occidentale, zone de colonisation phocéenne par excellence. F combine des échantillons des styles ‘Late Wild Goat’ (NAU 6, 7, 76) et nord-ionien à figures noires (NAU 15), ainsi qu’un bol ionien du type à rosettes (NAU 64), une coupe ionienne fine du type B1 ‘Lambrino’ d’Alexandrescu (NAU 60) et un fragment de céramique grise (NAU 70). Les compositions de ce groupe se rattachent à celles des ateliers de Clazomènes. La présence de cette coupe ionienne fine à vernis noir et filets grenat est beaucoup moins incongrue qu’il ne paraît: nous avons déja pu mettre en évidence une composition de ce type sur une autre coupe B1 ‘Lambrino’ de Bayraklý à décor ‘Late Wild Goat’.6 Cette pièce est donc à verser aussi au dossier des grands ateliers spécialisés primordiaux de coupes ioniennes. C2 associe deux coupes vrouliennes (NAU 58, 59) et le supposé fragment de situle de Daphnae du groupe C de Cook (DEF 1). Ces trois échantillons présentent en commun des teneurs en magnésium très élevées, qu’on ne trouve guère qu’à Rhodes. L’attribution de la situle de Daphnae à des ateliers rhodiens plutôt qu’helléno-égyptiens demanderait naturellement à être confirmée à l’aide d’échantillons supplémentaires. Si l’attribution rhodienne de notre échantillon se vérifiait, elle n’exclurait pas pour autant l’éventualité d’imitations manufacturées en Egypte même. G est formé d’un fragment de fruit-stand ‘Late Wild Goat’ (NAU 9) et d’un autre d’œnochoé ionienne à bandes (NAU 71). Les deux présentent en commun des pâtes sablonneuses rappelant celles des productions de la Vallée du Nil, avec des teneurs anormalement élevées en titane et particulièrement basses en potassium. Le style du fruit-stand évoque une variante éolienne du ‘Late Wild Goat’.7 L’échantillonnage dont nous avons disposé n’est certainement pas représentatif du faciès céramique de Naukratis, les proportions réelles des catégories importées n’étant pas respectées. Les comparaisons avec la situation observable à Histria ne peuvent donc avoir qu’une valeur indicative. Néanmoins, il est possible, pour les catégories principales, de formuler les constatations suivantes. • Le groupe chiote présente des caractéristiques de composition identiques à celles de notre principal groupe de référence de l’île et tout à fait comparables à celles rencontrées aux antipodes du monde colonial, sur les importations chiotes d’Histria par exemple. Les formes exportées correspondent essentiellement à des vases à boire: calices et tasses notamment. Le fait qu’à Naukratis, même le ‘Grand Style’ de Boardman présente lui aussi des compositions chiotes et non naukratites doit être signalé; si aucun exemplaire de cette variante polychrome du style de Chios n’est encore attestée à Histria, quelques fragments ont été exhumés à Bérézan.8 • Les productions de l’Ionie du Nord forment, comme à Histria, deux entités séparées, dont l’une correspond à

Naukratis: Les importations grecques orientales archaïques Clazomènes et l’autre, plus importante encore sur le plan quantitatif, à un centre non identifié de la même région,9 situé sans doute plutôt du côté de Téos qu’en direction d’Erythrées ou de Smyrne, dont les compositions sont différentes. Les officines nord-ioniennes, dont les exportations outre-mer ont démarré plus tard que celles de l’Ionie du Sud, ont choisi la voie, tracée par Corinthe, de la fabrication en grande série, qui connaîtra le succès commercial que l’on sait: ‘Late Wild Goat’ bâclé, fruitstands et assiettes à décor simplifié de grecques et motifs lotiformes, bols ioniens... . • Souvent proches des précédentes, les productions de l’Eolide, d’allure souvent plus ‘provinciale’, ne semblent pas avoir eu pour siège Phocée. Leur diffusion paraît avoir été beaucoup plus restreinte que celle des exportations nordioniennes, sauf si une connection se confirmait avec les coupes ioniennes fines. • A Naukratis comme à Histria, ce sont les mêmes groupes milésiens qui sont représentés, l’un d’entre eux se détachant nettement des deux autres sur le plan quantitatif. Du fait d’une fondation plus tardive, Naukratis n’a pas livré de ‘Middle Wild Goat II’, mais les groupes Fikellura de Cook sont massivement attestés. • Les grands centres exportateurs de coupes ioniennes n’étaient peut-être pas tous localisés en Ionie du Sud. Divers indices donnent à penser au contraire qu’une partie des formes fines a pu être fabriquée par des ateliers d’Ionie du Nord / Eolide. • Par rapport à Histria et au reste de la mer Noire, où elles font totalement défaut, les rares productions archaïques imputables à l’île de Rhodes ou à sa pérée sont bien présentes à Naukratis, sous la forme d’un petit nombre de coupes vrouliennes et, semble-t-il de situles de Daphnae.10 • Il reste enfin à évoquer le cas des productions céramiques coloniales fabriquées sur le sol égyptien. Elles ne paraissent pas avoir connu un développement très considérable, du moins en ce qui concerne celles à décor peint, et les rares spécimens que nous avons pu identifier par les analyses n’ont pas été nécessairement produits à Naukratis même. Toutefois, il existe de fortes présomptions pour que ce soit le cas, à en juger d’après le témoignage, bien postérieur certes, d’Athénée (Deipn. 9.480).11 Contrairement à une opinion tenace,12 sous-estimant les capacités d’adaptation des artisans potiers installés à demeure, il n’était point besoin d’importer de l’argile de Grèce de l’Est, les vases exhumés sur place ne nécessitant pas de caractéristiques de pâte particulières, à l’exception de la poterie à feu: les spécimens que nous avons pu identifier présentent des compositions qui sont celles de la basse vallée du Nil.13 Par ailleurs, alors qu’à Histria les productions locales à décor peint sont dominées par le style de Fikellura milésien, à Naukratis, les quelques imitations identifiées par les analyses sont à dominante nord-ionienne (‘Late Wild Goat’). Toutefois, il ne faut y voir, semble-t-il, que l’effet du hasard, car la vaisselle commune produite à Histria et, plus généralement, au nord de la

mer Noire présente elle aussi un faciès dominant du type Ionie du Nord – Eolide. Telles sont donc, rapidement esquissées, les interprétations archéologiques que l’on peut tirer des données d’analyse chimique de cet échantillonnage de Naukratis. Mais il est aussi une leçon sous-jacente ou sub-liminale à tirer de la démarche archéométrique utilisée en ce qui concerne les critères de différenciation, dont la fiabilité peut s’avèrer très variable: élevée dans le cas des ressemblances et dissemblances de composition (pour autant, bien sûr, que le réseau de références locales soit assez complet), mais beaucoup plus aléatoire selon la part tenue par les probabilités a priori, d’ordre archéologique notamment. Ceci explique que les attributions d’origine en laboratoire puissent parfois être remises en cause ultérieurement, au fur et à mesure des compléments apportés au réseau de références ou des progrès réalisés dans l’étude typologique traditionnelle des catégories céramiques entrant dans la composition des échantillonnages. Faute pour les archéologues (comme pour certains archéomètres) d’en prendre conscience, les analyses de laboratoire, telles les langues du vieil Esope, peuvent s’avérer capables du meilleur comme du pire. Notes 1

2 3

4

5 6 7

8 9 10

11 12 13

Sur ces matériels, cf. la rétrospective récente de Möller 2000a, 127-47, et, surtout, l’excellente discussion de fond de Kerschner 2001, 69-94, pls 7-10. Dupont 1983, avec biblio. antérieure. Cook 1954, II. D. m, 29-32; notre échantillon porte le même motif de palmette à incisions et rehauts grenat que les spécimens illustrés pl. 10.2-3, mais il pourrait s’agir aussi d’un fragment de stamnos, tel celui reproduit pl. 10.5. Sur les situles de Daphnae et la ‘Dark Ground Ware’, cf. aussi Schaus 1995, 25-9, pls 11-12; Weber Figs 10-13, 20-22. C’est aussi le cas pour deux autres échantillons de coupes / skyphoi apparemment similaires du British Museum analysés par le laboratoire de Bonn (Nauk 85-86), qui ne se rattachent à aucun des groupes de référence répertoriés de Grèce de l’Est. Kardara 1963, 276 no. 3. Cf. aussi la contribution de M. Kerschner dans ce même volume. Dupont 2000, 452 fig. 317. Le même tesson a fait l’objet d’une analyse de la part du laboratoire de Bonn (Nauk 33), qui l’attribue à un atelier du Delta du Nil, baptisé ‘Naukratis workshop’ (Cf. contribution de H. Mommsen dans ce volume). Korpusova 1987, 45 fig. 18. Groupe ‘Ionie du Nord 2’ apud Dupont 1983, 31-3. Naturellement, le résultat obtenu sur un unique échantillon du groupe C de Cook demandera à être validé par l’analyse de pièces supplémentaires et ne saurait être étendu à l’ensemble de cette classe disparate. Sur l’éventualité d’un production céramique sur place, cf. dernièrement: Möller 2000a, 136-45; Piekarski 2001b. Cf. encore, récemment, Kreuzer 1992, 54. Sur les ressources argileuses de la vallée du Nil, cf. Hope 1977, 72-4; Nordström et Bourriau 1993, 157-61; Aston 1996, 2-9. Sur les compositions des argiles de la vallée du Nil: Hancock, Aufreiter et Elsokkary 1986/7, 61-71. A en juger d’après les références de la banque de données du laboratoire de Lyon, le Delta, du fait de la densité des dépôts anthropogènes, constitue un milieu en moyenne plus réducteur que le reste de la vallée du Nil, avec de fréquentes pollutions par les phosphates et le manganèse. C’est un peu le cas pour nos échantillons NAU 9 et 71 et manifestement le cas pour NAU 34, 67-8 et, surtout, NAU 33.

Naukratis: Greek Diversity in Egypt | 81

Dupont et Thomas

Appendice Inventaire des echantillons analysés NAU 1 Cambridge, Arch. Museum, NA 7. Deinos. Style ‘LateWild Goat’. NAU 2 Cambridge, Arch. Museum, NA 13. Deinos. Style ‘Late Wild Goat’. NAU 3 Cambridge, Arch. Museum, NA 26. Œnochoé. Style ‘Late Wild Goat’. NAU 4 Cambridge, Arch. Museum, NA 30. Deinos. Style ‘Late Wild Goat’. NAU 5 Cambridge, Arch. Museum, NA 31. Deinos. Style ‘Late Wild Goat’. NAU 6 Cambridge, Arch. Museum, NA 32. Deinos. Style ‘Late Wild Goat’. NAU 7 Cambridge, Arch. Museum, NA 34. Œnochoé. Style ‘Late Wild Goat’. NAU 8 Cambridge, Arch. Museum, NA 47. Fruit stand. Style ‘Late Wild Goat’. NAU 9 Cambridge, Arch. Museum, NA 48. Fruit stand. Style ‘Late Wild Goat’. NAU 10 Cambridge, Arch. Museum, NA 50. Fruit stand. Style ‘Late Wild Goat’. NAU 11 Cambridge, Arch. Museum, NA 24. Œnochoé. Style nord-ionien à f. n. NAU 12 Cambridge, Arch. Museum, NA 60. Deinos. Style nord-ionien à f. n. NAU 13 Cambridge, Arch. Museum, NA 133. Lékanè. Style nord-ionien à f. n. NAU 14 Cambridge, Arch. Museum, NA 134. Amphorette. Style nord-ionien à f. n. NAU 15 Cambridge, Arch. Museum, NA 135. Amphorette. Style nord-ionien à f. n. NAU 16 Cambridge, Arch. Museum, NA 136. Œnochoé. Style nord-ionien à f. n. NAU 17 Cambridge, Arch. Museum, NA 137. Amphorette. Style nord-ionien à f. n. NAU 18 Cambridge, Arch. Museum, NA 138. Amphorette. Style nord-ionien à f. n. NAU 19 Cambridge, Arch. Museum, NA 140. Œnochoé. Style nord-ionien à f. n. NAU 20 Cambridge, Arch. Museum, NA 68. Calice. Style de Chios. NAU 21 Cambridge, Arch. Museum, NA 66. Calice. Style de Chios. NAU 22 Cambridge, Arch. Museum, NA 70. Phiale. Style de Chios. NAU 23 Cambridge, Arch. Museum, NA 73. Calice. Style de Chios. NAU 24 Cambridge, Arch. Museum, NA 74. Calice. Style de Chios. NAU 25 Cambridge, Arch. Museum, NA 76. Calice. Style de Chios. NAU 26 Cambridge, Arch. Museum, NA 77. Calice. Style de Chios. NAU 27 Cambridge, Arch. Museum, NA 78. Calice. Style de Chios. NAU 28 Cambridge, Arch. Museum, NA 79. Calice. Style de Chios. NAU 29 Cambridge, Arch. Museum, NA 80. Calice. Style de Chios. NAU 30 Cambridge, Arch. Museum, NA 82. Calice. Style de Chios. NAU 31 Cambridge, Arch. Museum, NA 83. Calice. Style de Chios. NAU 32 Cambridge, Arch. Museum, NA 84. Calice. Style de Chios. NAU 33 Cambridge, Arch. Museum, NA 86. Calice. Style de Chios. NAU 34 Cambridge, Arch. Museum, NA 88. Tasse. Style de Chios. NAU 35 Cambridge, Arch. Museum, NA 89. Calice. Style de Chios. NAU 36 Cambridge, Arch. Museum, NA 91. Calice. Style de Chios. NAU 37 Cambridge, Arch. Museum, NA 93. Calice. Style de Chios. NAU 38 Cambridge, Arch. Museum, NA 96. Phiale. Style de Chios. NAU 39 Cambridge, Arch. Museum, NA 97. Bol (?). Style de Chios. NAU 40 Cambridge, Arch. Museum, AG 232B. Œnochoé. Style de Fikellura. NAU 41 Cambridge, Arch. Museum, NA 118. Œnochoé. Style de Fikellura. NAU 42 Cambridge, Arch. Museum, NA 119. Œnochoé. Style de Fikellura. NAU 43 Cambridge, Arch. Museum, NA 121. Œnochoé. Style de Fikellura. NAU 44 Cambridge, Arch. Museum, NA 122. Œnochoé. Style de Fikellura. NAU 45 Cambridge, Arch. Museum, NA 126. Œnochoé. Style de Fikellura. NAU 46 Cambridge, Arch. Museum, NA 128. Œnochoé. Style de Fikellura. NAU 47 Cambridge, Arch. Museum, NA 129. Œnochoé. Style de Fikellura. NAU 48 Cambridge, Arch. Museum, NA 130. Œnochoé. Style de Fikellura. NAU 49 Cambridge, Arch. Museum, NA 131. Œnochoé. Style de Fikellura. NAU 50 Cambridge, Arch. Museum, NA 132. Œnochoé. Style de Fikellura. NAU 51 Louvre, AM 1479. Pinax. Style nord-ionien à f. n. NAU 52 Cambridge, Arch. Museum, NA 51. Segmentteller. Style nord-ionien à f. n. NAU 53 Cambridge, Arch. Museum, NA 45. Pinax. Style des Chèvres Sauvages. NAU 54 Calice. Style de Chios. ‘Middle Wild Goat II’. NAU 55 Coupe vroulienne ‘ancien style’ de Kinch ou skyphos indéterminé (surcuite). NAU 56 Calice. Style de Chios. ‘Polychrome Style’. NAU 57 Calice. Style de Chios. ‘Polychrome Style’. NAU 58 Coupe vroulienne. NAU 59 Coupe vroulienne. NAU 60 Coupe ionienne fine. Forme Villard B1. Type ‘Lambrino’. NAU 61 Coupe ionienne fine. Forme Villard B1. Type milésien. 82 | Naukratis: Greek Diversity in Egypt

Naukratis: Les importations grecques orientales archaïques NAU 62 NAU 63 NAU 64 NAU 65 NAU 66 NAU 67 NAU 68 NAU 69 NAU 70 NAU 71 NAU 72 NAU 73 NAU 74 NAU 75 NAU 76 NAU 100 NAU 101 DEF 1

Coupe ionienne fine. Forme Villard B3 . Décor d’applique (tête de bélier). Coupê ionienne fine. Forme Villard B3. Lèvre: frise de feuilles de laurier. Bol ionien. Type à rosettes de points. Bol ionien. Style ‘Late Wild Goat’. Amphorette à v. n. Frise d’épaule: languettes incisées avec rehauts grenat. Céramique grise. Rotella d’anse d’œnochoé. Céramique grise indéterminée. Céramique grise indéterminée. Céramique grise indéterminée. Oenochoé ionienne à bandes. Pâte siliceuse. Lampe type Howland 19A. Lampe ionienne massive a flancs en léger dévers. Deinos. Style des Chèvres Sauvages éolien. Alabastron fusiforme cotelé, à pâte grise et couverte noire. Cf. Samos VI.1, pl. 35 n° 271-5: ‘spezifisch samisch’. Amphorette Style ‘Late Wild Goat’ à frise lotiforme incisée / v. n. / épaule. Coupe ionienne Villard B2 / Hayes VIII. Coupe ionienne Villard B2 / Hayes VIII. Situle du type dit ‘de Daphnae’.

Resultats d’analyse (en %, sauf pour MnO en ppm)

N° NAU 01 NAU 02 NAU 03 NAU 04 NAU 05 NAU 06 NAU 07 NAU 08 NAU 09 NAU 10 NAU 11 NAU 12 NAU 13 NAU 14 NAU 15 NAU 16 NAU 17 NAU 18 NAU 19 NAU 20 NAU 21 NAU 22 NAU 23 NAU 24 NAU 25 NAU 26 NAU 27 NAU 28 NAU 29 NAU 30 NAU 31 NAU 32 NAU 33 NAU 34 NAU 35 NAU 36 NAU 37 NAU 38 NAU 39 NAU 40 NAU 41

CaO 09. 8 09. 5 09. 9 09. 4 08. 8 08. 8 06. 9 08. 2 04. 7 05. 3 07. 6 09. 0 06. 5 07. 9 06. 3 04. 7 07. 2 07. 6 08. 3 13. 2 10. 8 11. 5 10. 3 07. 7 09. 0 09. 0 11. 7 09. 0 06. 9 13. 0 10. 0 10. 1 16. 1 08. 2 10. 1 07. 6 13. 2 08. 7 12. 4 11. 0 11. 3

Fe2O3 06. 40 06. 30 05. 85 06. 35 06. 35 06. 70 07. 70 06. 80 10. 80 06. 30 06. 10 09. 00 08. 05 07. 00 07. 50 07. 45 08. 25 08. 80 07. 85 06. 45 06. 85 07. 50 07. 05 07. 50 07. 45 07. 25 06. 55 07. 60 07. 55 06. 47 07. 00 07. 25 06. 85 06. 25 06. 45 08. 05 07. 25 07. 85 07. 25 07. 80 07. 50

TiO2 0. 94 0. 91 0. 84 0. 95 0. 91 0. 87 0. 91 0. 91 2. 29 0. 92 0. 93 0. 87 0. 91 0. 96 0. 97 1. 00 0. 95 0. 83 0. 88 0. 72 0. 74 0. 77 0. 75 0. 81 0. 79 0. 78 0. 70 0. 80 0. 88 0. 67 0. 82 0. 84 0. 71 0. 84 0. 74 0. 82 0. 75 0. 85 0. 72 0. 79 0. 76

K2O 2. 88 2. 98 2. 83 2. 88 2. 88 3. 36 2. 93 3. 12 1. 30 2. 98 3. 17 3. 07 3. 50 3. 12 3. 07 3. 60 3. 22 3. 41 3. 07 2. 21 2. 11 2. 02 2. 02 2. 02 2. 02 2. 21 2. 06 2. 21 2. 45 2. 10 2. 21 2. 30 2. 06 2. 26 2. 11 2. 11 2. 11 2. 45 2. 06 3. 41 2. 88

SiO2 58. 4 58. 5 59. 6 58. 8 59. 2 58. 6 58. 0 58. 1 59. 3 62. 4 60. 4 52. 8 55. 7 57. 4 58. 4 58. 2 54. 6 53. 4 56. 4 55. 5 58. 1 55. 9 57. 1 58. 9 58. 3 58. 2 58. 6 54. 9 56. 1 57. 0 57. 5 56. 3 53. 2 60. 3 60. 5 58. 6 55. 0 56. 0 54. 5 51. 8 51. 7

Al2O3 19. 8 19. 4 18. 6 18. 8 19. 3 19. 4 19. 4 20. 0 17. 1 20. 3 20. 1 19. 9 19. 7 21. 8 19. 8 22. 5 21. 3 21. 3 17. 5 13. 8 14. 1 14. 7 14. 7 15. 1 15. 1 15. 6 13. 6 15. 3 16. 9 14. 8 15. 1 16. 2 13. 8 15. 3 13. 8 15. 9 14. 5 17. 1 14. 3 16. 7 16. 4

MgO 02. 60 01. 90 02. 30 02. 05 02. 35 03. 10 03. 30 02. 70 03. 80 02. 75 01. 95 05. 40 01. 85 02. 60 03. 10 02. 35 03. 80 04. 60 05. 40 05. 85 06. 00 07. 20 06. 90 06. 60 05. 55 05. 70 05. 45 05. 70 05. 00 05. 65 06. 25 05. 90 05. 45 05. 15 05. 50 06. 20 06. 30 06. 05 07. 35 06. 70 07. 35

MnO 0740 0840 0880 1260 1020 1620 1740 1000 1920 0880 0940 1480 1040 0880 1480 1020 1600 1200 1000 1760 1260 1300 1160 1180 1340 1400 1100 1500 1220 2090 1140 1060 6360 2960 1520 1280 1280 1040 1200 1300 1240 Naukratis: Greek Diversity in Egypt | 83

Dupont et Thomas N°

CaO

Fe2O3

TiO2

K2O

SiO2

Al2O3

MgO

MnO

NAU 42 NAU 43 NAU 44 NAU 45 NAU 46 NAU 47 NAU 48 NAU 49 NAU 50 NAU 51 NAU 52 NAU 53 NAU 54 NAU 55 NAU 56 NAU 57 NAU 58 NAU 59 NAU 60 NAU 61 NAU 62 NAU 63 NAU 64 NAU 65 NAU 66 NAU 67 NAU 68 NAU 69 NAU 70 NAU 71 NAU 72 NAU 73 NAU 74 NAU 75 NAU 76 NAU 100 NAU 101 DEF 01

04. 1 03. 6 10. 9 08. 1 09. 3 09. 0 08. 6 07. 1 07. 8 06. 2 08. 9 12. 4 12. 1 12. 9 08. 5 09. 3 08. 9 03. 9 07. 6 09. 5 08. 6 05. 5 07. 7 06. 9 07. 1 04. 3 05. 3 07. 6 05. 5 08. 6 13. 5 07. 3 06. 30 14. 18 06. 10 05. 8 05. 5 12. 1

06. 10 06. 10 07. 45 05. 90 08. 05 07.35 07. 40 06. 30 06. 80 07. 00 06. 74 06. 63 07. 05 06. 46 07. 66 07. 40 08. 52 08. 63 07. 77 06. 83 08. 65 08. 22 07. 92 09. 24 08. 40 06. 72 06. 76 06. 83 07. 30 09. 76 08. 10 08. 85 08. 69 07. 75 07. 22 08. 59 08. 57 08. 13

0. 80 0. 81 0. 78 0. 72 0. 81 0. 80 0. 79 0. 76 0. 78 0. 94 0. 93 0. 86 0. 76 0. 80 0. 82 0. 81 0. 77 0. 84 0. 95 0. 74 1. 00 0. 98 0. 94 0. 89 0. 90 0. 93 0. 89 0. 87 0. 89 1. 83 0. 86 0. 82 0. 87 0. 84 0. 86 0. 96 0. 95 0. 60

4. 51 4. 51 3. 17 3. 65 3. 26 3. 46 3. 31 3. 89 3. 65 3. 35 2. 96 3. 11 1. 61 2. 83 2. 08 2. 30 2. 62 2. 59 3. 48 3. 80 3. 82 3. 85 3. 04 3. 46 3. 49 2. 82 2. 75 2. 83 3. 26 1. 37 3. 22 3. 69 3. 61 3. 35 3. 17 4. 04 4. 24 1. 68

59. 7 60. 1 51. 7 56. 9 51. 5 53. 3 54. 0 57. 0 55. 9 57. 0 58. 6 53. 8 53. 2 53. 0 56. 7 53. 6 44. 3 53. 9 56. 1 53. 1 48. 6 54. 1 56. 2 49. 4 53. 6 63. 6 62. 9 57. 7 57. 3 56. 4 50. 4 53. 8 54. 3 51. 3 60. 3 55. 4 56. 4 50. 3

20. 0 19. 8 17. 3 17. 2 17. 2 17. 9 17. 3 18. 3 17. 9 21. 6 19. 7 16. 4 13. 6 17. 2 14. 7 15. 3 13. 6 15. 0 19. 1 17. 9 21. 3 21. 4 19. 2 21. 1 20. 4 17. 5 15. 5 17. 2 17. 9 15. 5 18. 5 19. 6 20. 4 16. 8 19. 1 21. 0 20. 4 10. 5

03. 00 03. 55 06. 90 04. 40 06. 40 06. 95 06. 75 04. 65 05. 85 02. 85 02. 20 05. 00 07. 06 03F. 57 06. 84 07. 22 12. 20 11. 86 03. 47 04. 31 05. 09 03. 57 03. 37 04. 82 04. 20 02. 70 03. 63 03. 02 05. 20 03. 60 03. 79 03. 86 03. 98 03. 87 01. 99 03. 06 02. 91 18. 00

1140 0900 1160 1200 1180 1060 1100 0820 1060 0940 0850 1050 1150 0910 1190 1130 0960 0650 1530 1250 1290 1110 1720 1270 1180 1500 2940 0860 1450 1500 1904 1326 1191 3391 1732 1202 1499 1285

84 | Naukratis: Greek Diversity in Egypt

Archaic Greek Plates from the Apollo Sanctuary at Emecik, Knidia Results and Questions Concerning Dorian Pottery Production

Regina Attula Abstract New finds from the recent excavations at the sanctuary of Apollo at Emecik on the Knidian peninsula add much to our understanding of East Dorian pottery. A large group among the finds are painted plates. Some of them are decorated with marine or mythological subjects and functioned as votive plaques. Others, with floral decoration, for the first time attest the existence of a local, East Dorian Fikellura production.* The Turkish-German excavations of 1998–2003 in the extraurban sanctuary of Apollo Karneios at Emecik village on the Knidian peninsula produced a great amount of Archaic Greek material from the 7th and 6th centuries bc.1 For the first time clays from the Knidia were analysed by Neutron Activation Analysis (NAA), with 137 samples in all from terracotta figurines and ceramic vessels.2 The main result of the NAA are seven new and hitherto unknown chemical groups, named EMEA, B, C, D, E, F and G (Fig. 1).3 These new results substantially enrich our level of knowledge concerning the spectrum of Archaic Greek finds from the Knidian peninsula and East Dorian pottery production and its relations to Ionian and other workshops.4 As a result, a new facet of the production of Fikellura pottery begins to emerge.

Archaic plates from the Knidan peninsula One of the most important groups of material from the excavation at Emecik are the plates and stemmed plates (‘fruitstands’) or flat bowls, with a minimum of 40 to 45 pieces in all.5 No complete vessels are preserved. Very similar plates and stemmed plates are known also from the Archaic settlement layers in Burgaz/Datça to the west of Emecik, the closest Knidian findspot for Archaic East Greek pottery.6 These shallow shapes are about 30cm in diameter and are distinguished by wide rims, several of them with handles or with spool-shaped lugs. Their undersides can be completely glazed or streaky; some have ring-bases. Two groups can be distinguished: plates with patterned decoration and plates with figured decoration, and in both groups several examples are designed as segment plates.7 Figure-decorated plates: mythological and narrative scenes Two plates from Emecik show single animals between various filling ornaments: a bird with filling rosettes on cat. no. 7 and the bull on no. 8 (Figs 2–3). Their provenance from a single workshop is indicated by the use of the same bright reddishbrown added colour, which was only observed on these two fragments.

Figure 2 Cat. no. 7 (not sampled)

Figure 1 Results of Neutron Activation Analysis with 112 samples

Naukratis: Greek Diversity in Egypt | 85

Attula

Figure 3 Cat. no. 8 (not sampled)

Figure 4 Cat. no. 6 (Emec 71, EMEB)

On the fragmentarily preserved flat plate cat. no. 6 (Fig. 4) the figure of a Potnia Theron is depicted. She wears a long belted garment, which on the lower part is decorated with vertical geometric patterns of zigzags and filled squares. Her uplifted arm is drawn in outline technique, the other is not preserved. Her head and the upper body as well as the animals that are her attributes cannot be reconstructed with certainty; perhaps figures of lions or geese originally completed the composition. Similar East Greek plates with mythological figures include well-known examples of segment plates, such as the Euphorbos plate (Fig. 5)8 and the Gorgon plate in the British Museum,9 or the Perseus plate in Berlin,10 all three from Kameiros. Elena Walter Karydi admired them for belonging ‘to the heyday of Archaic East Dorian vase-painting’.11 The movement of these figured schemes, the frontality of the Gorgon face and the added colours12 all add to the monumental effect of these pinax-like painted plates.13 To this small list of mythological figures on Archaic East Dorian ceramics furthermore may be added the depictions of Typhon and a Boread on a situla from Tell Defenneh in the British Museum (Weber Fig. 8).14 Among these early mythological pictures15 the Euphorbos plate (c. 600 bc)

occupies a special position because of the complex depiction of a scene with three fully armed men.16 Two of them are fighting with their lances and round shields while the corpse of the third is lying on the earth between them. All three hoplites are individualized and recognizable by their name inscriptions. Their names Menelas, Hector and Euphorbos are known from the Iliad, although the picture and the epic text are not congruent (Il. 17.70–89).17 The other mythological figures on East Greek plates, which are single figures without name inscriptions, are recognizable from their individual appearance (Gorgo Medusa, Potnia Theron) or from their attributes (Perseus, Potnia Theron).18 Here we find close parallels to the representations of Athena Promachos on Chian pottery.19 All these depictions of an epic scene or of a mythological figure have both a decorative and a special narrative content. Like the three above-mentioned examples from Kameiros, many of the East Greek painted plates come from Rhodian graves or from neighbouring island sites on Kos or Kalymnos. In general, the date of the Dorian plates can be given as between the middle/late 7th and the early 6th century bc. The problem of their provenance is still unsolved.20 The archaeometric analysis of the clay sample from plate cat. no. 6 (sample Emec Figure 5 Euphorbos plate, British Museum GR 1860.4-4.1

86 | Naukratis: Greek Diversity in Egypt

Archaic Greek Plates from the Apollo Sanctuary at Emecik, Knidia 71, Fig. 4) shows that they belong to the main chemical group EMEB, which contains most of the Archaic painted fine ceramics from Emecik.21 In all probability, this group EMEB is to be localized in an East Dorian pottery workshop, presumably in the Knidia. Karl Schefold already localized the London Euphorbos plate and the whole ‘Euphorbosstil’ in the Knidian peninsula.22 Elena Walter-Karydi, on the other hand, supported a Koan or Kalymnian origin for this plate and generally described the Knidian plates as of poor quality and much more provincial than the painted plates from the Dorian islands.23 It led her to the assumption ‘that Cos played a leading role in this school’.24 The role of the Koan workshops will become clearer in future studies on the Archaic material, which at present is little known. The placing of the Potnia Theron plate, cat. no. 6 (Fig. 4), directly into the Euphorbos group is not possible, since unfortunately this piece is too fragmentarily preserved. Many formal features match, but the shape of the rim is unknown and therefore the profile is incomplete. On our cat. no. 6 there seem to be no incisions such as are found on Hector’s shield-device on the Euphorbos plate, or on the figure of Perseus on the plate in Berlin, nor is there evidence for polychromy as it is found on the Euphorbos plate. Moreover, no filling ornaments are preserved. The best thematic correspondences are perhaps with the London Gorgon plate, in so far as both depict a long-garmented mythological female figure in partial frontality. What emerges, then, from this comparison of formal and technical features is that the plates with figured decoration are not a uniform group. This would suggest caution in the use of the conventional terms ‘Euphorbosstil’ or Euphorbos group, or at least its limitation to iconographic features. In fact, the name seems hardly suitable for the classification or localization of the figure-decorated plates, as regards either form, technique or content. Furthermore, it is significant that many East Dorian plates with figured decoration are comparable to Cycladic examples, such as the Bellerophon plate from Thasos, which is probably of Naxian origin.25 The similarities extend to both formal and stylistic features. In order to establish a general model of the various Archaic traditions in East Greek plate painting, we still lack an overview of the figure-decorated plates from East Greek coastal findspots and from the Aegean islands, with their evidently Cycladic influences. Archaeometric investigations of selected samples from Dorian, Ionian and Cycladic plates, with a distinct archaeological question to be answered, would be very

Figure 6 Cat. no. 1 (Emec 1, EMEB)

Figure 7 Cat. no. 2 (Emec 60, EMEB)

Figure 8 Cat. no. 3

worthwhile. Three other plates from Emecik, cat. nos 1–2 (Figs 6–7) and 5 (Fig. 11), show marine and submarine subjects. Best preserved is fragment cat. no. 1 (Fig. 6) with the detailed depiction of a rowing ship. It was painted in diluted glaze without incisions. We see the bow with an apotropaic eye and the row of side oars. The armed crew on deck is to be assumed behind the horizontal line of overlapping semicircles, only four of which are preserved, which represent the large hoplite shields. Around this ship a dolphin and several filling ornaments are grouped. Direct parallels for the plate cat. no. 1 (Fig. 6) are the two segment plates cat. nos 3 (Figs 8–9) and 4 (Figs 9–10) kept in the British Museum. They are said to have been ‘found in a small temenos at Datcha near Cnidus’ and were probably in fact found in the Apollo sanctuary at Emecik.26 As a result of the NAA we find the sample from no. 4 (sample Knid 1) in the chemical group EMEb, which is very close to the main local group EMEB.27 Numbers 3 and 4 each show the ship in the lower segment of the plate. Fragment no. 4 depicts the bow side of a ship with the long row of side oars and with two seamen on deck. Their big eyes are drawn in outline technique, and in general their heads are depicted with only few lines, which adds special expressiveness to these figures. They are similar to the depiction of the helmsman on a warship that is shown on a votive pinax from the Athena sanctuary at Sounion;28 this example stands in the strong iconographic tradition of ship depictions on Attic Geometric vessels.29 The ship on plate no. 4 (Figs 9–10) is accompanied by a dolphin. Remarkably, it faces to the right, the opposite direction from the other ship depictions on plates. The second ship plate in the British Museum, cat. no. 3 (Figs 8–9), shows a nearly fully preserved ship facing left. Beside the long

Figure 9 Cat. nos 3 and 4, profiles

Naukratis: Greek Diversity in Egypt | 87

Attula

Figure 11 Cat. no. 5 (Emec 59, EMEB)

Figure 10 Cat. no. 4 (Knid 1, EMEb)

row of side oars the two steering oars are also preserved, as is the aphlaston, with its high curved but not yet fan-like, preClassical shape.30 Exactly the kind of steering oars as on cat. no. 3 and part of the ship’s hull also are depicted on a small plate fragment from Emecik, cat. no. 2 (Fig. 7). This fragment, therefore, should be restored to a very similar depiction as on plate no. 3. These four ship plates, cat. nos 1–4, all from the Knidian peninsula – two from the Apollo sanctuary at Emecik and two in the British Museum – represent the most detailed East Greek depictions of ships on ceramics.31 Presumably they were all produced in the same pottery workshop, at least nos 1, 2 and 4. The smaller plate no. 3 differs from the other three in shape, colour and the quality of the bright reddish-brown glaze, but not in its fabric or in the artistic subject. Maybe this workshop designed such ship plates in several sizes and produced them in different varieties of glaze. To this small number of East Dorian ship plates may, on stylistic grounds, be added also an example from Cyrene, on which only the bow of the ship is preserved.32 Furthermore, there is a plate from Delos with a much more stylized depiction of a ship between a large lotus flower and small rosettes.33 Another plate fragment with a marine subject from Emecik shows an underwater scene (cat. no. 5, Fig. 11). Beneath two dolphins a snake-like or Hydra-like sea-monster with sharp triangular spikes seems to be represented. In Archaic East Greek vase-painting this remarkable subject is without parallels, but it is close to a single depiction on an Early Iron Age plate from Cyprus.34 As a result of the NAA we find the sample from cat. no. 5 (Emec 59) in the main group EMEB, together with the two ship plates from Emecik cat. nos 1–2 (Figs 6–7), the ship plate in the British Museum cat no. 4 (Figs 9–10), and the Potnia Theron plate cat. no. 6 (Fig. 4).

Figure 12 Cat. no. 9 (Emec 62, single)

88 | Naukratis: Greek Diversity in Egypt

The similarities between these plates with figured decoration and painted pinakes (or plaques) have been mentioned above. Maybe the ship plates from Emecik, and the examples from Cyrene and Delos mentioned above, which were also found in sacral contexts, had a similar function as dedications in a sanctuary35 as did the Attic votive pinax from Sounion.36 A strong further argument for this are two small prefiring drilled holes in the central upper part of many Archaic East Greek plates, as can be seen on plate cat. no. 3 (Fig. 8).37 These holes served for hanging or otherwise attaching these picture plates. Even if on the Emecik fragments no such holes are preserved, such a representative use of the ship plates seems suggestive. Pattern-decorated plates: a further instance of local Fikellura production? In the group of plates or stemmed plates with patterned decoration, cat. nos 9–12 (Figs 12–15) there is no less variety than among the plates with figured decoration. One of the main types here is the plate with concentrically decorated surface. In addition to some plates with purely banded decoration there are several with concentric decoration around a central floral motif, often a big star rosette. As in the figure-decorated plates, some pattern-decorated ones are also designed as segment plates. Furthermore, added colours are rare and there are no incisions. Many filling ornaments are in common with those on the figuredecorated plates and also on the closed vessel forms (amphorae, jugs). As with the figure-decorated plates cat. nos 1–2 and 5–6, some of the samples from pattern-decorated plates fall into the main chemical group EMEB.38 The fragment cat. no. 10 (Fig. 13) from the centre of a stemmed plate deserves our particular attention, since it shows a certain connection between Emecik and Naukratis. This is now the third such link to be established via the Naukratis material kept in the British Museum, in addition to an East Dorian cup (Schlotzhauer and Villing Fig. 24)39 and Cypriot pottery.40 The fine and thick white-ground surface of plate no. 10 contrasts with the concentric black-glazed bands framing a row of

Figure 13 Cat. no. 10 (Emec 63, EMEE)

Archaic Greek Plates from the Apollo Sanctuary at Emecik, Knidia

Figure 14 Cat. no. 11 (Nauk 8)

Figure 15 Cat. no. 12 (Emec 58, EMEE)

meander hooks. From the frieze of alternating lotus flowers and buds between the glazed zones only the lower half of a single bud is preserved, but the reconstruction to a lotus frieze seems unequivocal. Stylistically very close to our no. 10 is a rimfragment from a plate or a stemmed plate, cat. no. 11 (Fig. 14), from Naukratis kept in the British Museum.41 Its colour and quality of clay, containing mica, as also the quality of the slip and glaze are closely comparable. Furthermore, the lotus buds show the same form as those on fragment no. 10. This slim shape of the buds without a contour-line and with the glazed dot at the bottom are known to be typical of South Ionian Fikellura lotusfriezes (SiA II, MileA II).42 The band of meander hooks is also present: this motif appears twice on no. 11, below the lotus frieze and also on the rim. In comparing common stylistic features from both plates no. 10 and no. 11, one would attribute them without hesitation to the same South Ionian workshop. The result of the NAA places cat. no. 10 (Emec 63) in the group EMEE, which is a small and very heterogeneous group of only five samples, including the plate fragment cat. no. 12 (Fig. 15).43 The fragment from Naukratis (Nauk 8) now finds a chemical partner in another sample from Emecik (Emec 31), which belongs to an unglazed base sherd, perhaps from an amphora.44 Nauk 8 with Emec 31 form a chemical pair with a still unknown provenance. Since there is no chemical connection between the Fikellura-style plate no. 10 (Emec 63) and the unspecific base sherd Emec 31, the provenance of both Fikellurastyle plates nos 10 (Emec 63) and 11 (Nauk 8) still cannot be localized. On the present state of knowledge we have to focus on two possibilities for an interpretation model: Either the assumed local East Dorian (maybe Knidian) production of such Fikellurastyle stemmed plates was so close to South Ionian (Milesian) that we should speak here of excellent copies – or maybe, on the other hand, the group EMEE was produced not by an East Dorian, but by a hitherto unknown South Ionian pottery workshop, which perhaps is to be placed at Miletos or in the surrounding region (SiA II, MileA II).45 Recently Richard Posamentir described a similar phenomenon when publishing an important plate fragment from Berezan kept in the Hermitage Museum.46 This plate shows a lotus-frieze very similar to those on plates nos 10 (Emec 63) and 11 (Nauk 8), but painted in a bright reddish-brown glaze. From the technical and stylistic features one would identify this piece as South Ionian or Milesian (MileA II). Surprisingly, the sample of this plate (Bere 125) fits into the chemical group DTroy, which contains samples from Bronze Age and Iron Age ceramics from Troy and its surrounding region.47 On strong topographical and historical arguments Posamentir (as well as Mommsen et al. this volume) assumes the localization of this group D-Troy in a pottery workshop at Abydos (‘Hellespont

workshops’). Furthermore, Udo Schlotzhauer has now established Fikellura-style production at Naukratis (‘Naukratisworkshop’), which included plates with a lotus frieze, too (samples Nauk 25 and 33 [Dupont and Thomas Fig. 1].48 We should not be surprised by the discovery of further local production centres of Fikellura-style pottery outside Miletos, in view of the strong influence exerted by the highly-developed South Ionian workshops on their neighbouring territories and on the colonized regions.49 In this connection we should consider also the southern dissemination of South Ionian pottery, following here the supposition of Robert M. Cook and Pierre Dupont: ‘Future discoveries are likely to show that it [Fikellura] was popular generally throughout the southern part of the East Greek region’.50 With the finds from Emecik we have, then, for the first time, samples of Archaic South Ionian pottery from the Knidian peninsula that also include Fikellura.51 Because the sherds from Emecik were found in a filling layer at the southern temenoswall, we lack a stratigraphically-based chronology for the Archaic ceramics.52 In addition to the plate fragment cat. no. 10 (Emec 63) sherds from Fikellura-style jugs and from banded plates were found. Among them are three samples (Emec 64, 72, 115) belonging to the well-established group A (Kalabaktepe workshop) and one sample (Emec 116) belonging to group D (probably Miletos).53 From the Archaic settlement layers at Burgaz/Datça five sherds (not sampled) from Fikellura-style amphorae or from trefoil-mouthed jugs are known (MileA I-II).54 At present the number of Fikellura-style pottery in the Knidia is not very high, and the amount of Fikellura among the South Ionian material in general can be given as nearly a third. So at least the supposition of Robert M. Cook and Pierre Dupont concerning the Knidia can be confirmed. It seems, then, that the assessment of Ionian influence on the assumed Cnidian pottery production depends particularly on the localization of group EMEE. Recently, the features and chronologies of South Ionian pottery have been comprehensively reviewed by Michael Kerschner and Udo Schlotzhauer.55 In addition, as far as general observations on the different features of South Ionian and East Dorian pottery are concerned, I partly follow Elena WalterKarydi.56 First, there is her strong point that ‘the East Dorians have a much greater interest in human figures and mythological scenes than the South Ionians’, as we have seen here for example with the Potnia Theron plate, cat. no. 6 (Fig. 4). WalterKarydi’s second point, that in the East Dorian regions Wild Goatdecorated vessels are generally less dominant, is a more complex matter. Her theory of the islands’ supremacy in comparison with the mainland is to be rejected so long as there is such an evident lack of comparable ceramic material from coastal findspots and Naukratis: Greek Diversity in Egypt | 89

Attula from the Aegean islands. Now several degrees of Ionian influence can be distinguished, especially various local variations in the incorporation of Ionian styles or patterns (as, for example, elements from the Fikellura style) into the traditions of the Dorian, or Carian, workshops.57 Referring here to recent historical and epigraphic research,58 this complex question can be formulated clearly with regard to one special aspect: following the epigraphic evidence, there is a linguisticdialectical borderline between Ionia and Caria (koine Greek and Dorian dialect). To what extent does this phenomenon correspond with the evidence from the material culture – Ionian (Milesian) influence on the one hand and ‘genuine Dorian art’ on the other? Research into this important question seems more fruitful than the construction of differences between the Aegean islands and the mainland in general. If we go back to the localization of the group EMEE, we have to mention once more the Fikellura-style examples from Berezan and from elsewhere. The situation that is known from the more northern East Greek regions, from the Troad (Abydos)59 to the western Black Sea shore (as, for example, Istros),60 could be similar to the Dorian regions (for example the region of Iasos or the Knidia); at least we cannot exclude such a model for Caria at present.61 What is significant here is the fact that at the same time the production model in Ionia itself is now being reviewed differently. In fact, the equation of Fikellura production in Ionia with a genuine Milesian workshop (groups A, D) must be revised in the light of recent investigations showing us different chemical compositions for Fikellura products, some of which may belong to non-Milesian workshops.62 At present, the question of the localization of group EMEE thus has to be left partially unanswered. But because of the high frequency with which new archaeologicalarchaeometrical results now keep on changing our state of knowledge of East Greek pottery, I am convinced it will soon also address this important aspect. Outlook The aim of this contribution is to understand Archaic Knidian, and more generally, Archaic East Dorian, art as a component part of a general phenomenon. At present, Dorian Archaic sculpture,63 along with Hellenistic and Roman pottery,64 are still better understood than Archaic Dorian pottery. We don’t know much about Archaic Dorian workshops and the actual dissemination of Archaic Dorian pottery.65 We can expect further insights in future from other findspots at Dorian coastal sites (such as Halikarnassos) as well as from the islands.66 The numerous coarse and unglazed wares (jars, pithoi, mortaria [Villing Fig. 17]67 and transport containers)68 must be included in this. The final aim should be a chronological and stylistic model of the East Dorian pottery production, in which the Knidia will surely occupy an important position.69

90 | Naukratis: Greek Diversity in Egypt

Catalogue Cat. no. 1. (Fig. 6) Attula 2006, 120 cat. 204, pls V.2; 60.1-2. Emec 1, EMEB. Segment plate, fragmentary. Depiction of a ship, beneath a dolphin and two filling ornaments. From Emecik. Cat. no. 2. (Fig. 7) Attula 2006, 120-21 cat. 205, fig. 19, pl. 62.2. Emec 60, EMEB. Fragment of a segment plate. Depiction of the back part of a ship, only the steering oars and a part of the hull are preserved. From Emecik. Cat. no. 3. (Figs 8–9) British Museum GR 1893.11-13.5; Schefold 1942, 129 fig. 1; Basch 1987, 242 fig. 511; Attula 2006, 119 pl. 61.4. Not sampled. Segment plate, restored, diameter 17.5 cm. With four spool-shaped lugs and two pre-firing drilled holes. Bottom completely covered in reddishbrown glaze. Depiction of a ship with side oars, steering oars, and aphlaston. In the upper segment an animal and filling ornaments, partially preserved. Without incisions. From ‘a small temenos near Datcha’. Cat. no. 4. (Figs 9–10) British Museum GR 1893.11-13.4; Schefold 1942, 129 fig. 2; Basch 1987, 242 fig. 510; Attula 2006, 119, pl. 61.2-3. Knid 1, EMEb. Fragment of a segment plate, diameter 29cm. Depiction of the bow side of a ship with embolon and akrostolion, on deck two seamen, a dolphin. In the upper segment only the lower legs of an animal are preserved. Without incisions. From ‘a small temenos near Datcha’. Cat. no. 5 (Fig. 11) Attula 2006, cat. 203 fig. 19, pl. 61.1. Emec 59, EMEB. Fragment of a plate. Depiction of an underwater-scene with two dolphins and perhaps a seamonster. From Emecik. Cat. no. 6 (Fig. 4) Attula 2006, 120-2 cat. 206 fig. 19 pl. 62.1. Emec 71, EMEB. Fragments of a plate. Depiction of a Potnia Theron figure. Without incisions. From Emecik. Cat. no. 7 (Fig. 2) Attula 2006, 122 cat. 207, fig. 21, pl. VI.3; 62.3. Not sampled. Fragment of a segment plate. Depiction of a bird, filling ornaments. Added colour. From Emecik. Cat. no. 8. (Fig. 3) Attula 2006, 122 cat. 208, pl. 62.5, 8. Not sampled. Fragments of a segment plate. Depiction of a bull. Added colour. From Emecik. Cat. no. 9 (Fig. 12) Attula 2006, 122 cat. 210, pl. 62.6. Emec 62, single. Fragment of a plate. Ornamental decorated with bands of double-volutes and hooks around a stylized flower and small dot-rosettes. From Emecik. Cat. no. 10 (Fig. 13) Attula 2006, 142 cat. 279, pl. V.3; 73.6. Emec 63, EMEE. Fragment of a Fikellura-style stemmed plate, cf. cat. no. 11. Concentric zones decorated with meander and lotus-frieze between glazed bands. From Emecik. Cat. no. 11 (Fig. 14) British Museum GR 1924.12-1.1113; Petrie 1886b, pl. 7.6. Nauk 8, pair with Emec 31. Fragment of a Fikellura-style stemmed plate, cf. cat. no. 10. Concentric zones decorated with meander, lotus-frieze and hooks between glazed bands. From Naukratis. Cat. no. 12 (Fig. 15) Attula 2006, cat. 221, pl. 64.8. Emec 58, EMEE. Fragment of a plate of remarkable weight and different quality of glaze. Concentric zone decorated with a zigzag-band between glazed bands. From Emecik.

Archaic Greek Plates from the Apollo Sanctuary at Emecik, Knidia Illustration credits

Fig. 1 after Mommsen, Schwedt and Attula 2006, 202 fig. 36; Fig. 4 R. Attula; Figs 5, 8, 9, 10, 14 the British Museum; drawings by K. Morton; others: Johannes Kramer, after Berges 2006.

Notes *

1 2 3

4 5 6

7

8 9 10 11 12

13

14

15

16

17 18 19

20

21

22 23

This project under the directorship of N. Tuna (ODTÜ Ankara, TAÇDAM), D. Berges (University of Hamburg) and the Museum at Marmaris was supported by the Gerda-Henkel-Stiftung, Deutsche Forschungsgemeinschaft and Deutsches Archäologisches Institut Berlin. I wish to thank the organizers Alexandra Villing and Udo Schlotzhauer for the invitation to the colloquium. Berges and Tuna 2000; Berges 2002, 2006; Tuna 2004. Mommsen, Schwedt and Attula 2006. The local reference group were stamped handles from Knidian transport amphorae and debris from the pottery workshop in Reþadiye, near Emecik. In addition to Berges 2006, in this contribution are used some corrected abbreviations as EMEA, EMEB to EMEG for the former names of the new chemical groups EME-A, -B- to -G; or the now obligatory classification term MileA for the former MilA. For the history of research see Kerschner 2001, 88-9; Schlotzhauer 2001a, 115-6; Attula 2006, 113-4. For the terracottas from Emecik see Kleibl 2006. Attula 2006, 114-26 cat. nos 203-40. Here, too, only fragments are preserved; see Özer 1998, 30-5 cat. nos 42-8. I thank the author for providing his manuscript. On the Archaic East Greek sherds from Cape Tekir, at the tip of the peninsula, see Attula 2006, 114. Attula 2006, 114-5 n. 279. The formal and stylistic analogies between the various East Greek segment plates and Attic, Corinthian, Cycladic, and Laconian plates are to be examined in a separate study. See also Callipolitis-Feytmans 1974; Todd 1973; Manyas 1984. BM GR 1860.4-4.1; Walter 1968, pl. 129.623; Williams 1999, 43-4 fig. 31. BM GR 1860.4-4.2; Walter 1968, pl. 130.626. Walter-Karydi 1973, pl. 136.1121. Mandel 2005, 152 n. 123 with reservations regarding the designation of this figure as Perseus. Walter-Karydi 1998, 292. For polychromy in early East Greek vase painting, see Furtwängler 1980, 188-95 figs 8-11, pls 54-5 (hydria from the Samian Heraion, c. 600 bc); Schaus 1988; Boardman 1998b, 143, 145-6, 221-2; Lemos 2000, 384-8. Schefold 1993, 18: ‘Archaisch ist das Aneinanderfügen der Bildelemente, hocharchaisch die einfache Größe der Formen und typisch ostgriechisch, wie die Gestalten in bunte umgebende Ornamentik verwoben sind’. BM GR 1888.2-8.1. See Weber, this volume. In addition, two plate fragments from Emecik may depict a crouching sphinx (or a lion?), but as only the paws are preserved the figure cannot be identified with certainty; see Attula 2006, cat. nos 209, 228, pls 62.5, 65.6. For mythological representations in Greek pottery from Egypt, see Schlotzhauer and Weber 2005. For Eastern, especially Cypriot influence on the early iconography of Herakles, see Iacovou 1988, 19 cat. no. 33, figs 77-8; Karageorghis 1997, 221-3. Schefold 1964, 8-9, 64 fig. 75; Walter-Karydi 1998, 292. The partial similarity between the three warriors on the Euphorbos plate and the battle-frieze figures on the Corinthian Chigi olpe (LPC, c. 640 bc) is well-known, see Hurwit 2002. Jeffery 1990, 153-4, 353-4; Ahlberg-Cornell 1992, 65 no. 42. Tempesta 1998, 50-68; Coulié 2002, 121 cat nos 327-8, pl. 85. Lemos 1991, 160 figs 12-13; Villing 1998. See Barclay (forthcoming), I thank the author for information on her manuscript. For female winged figures in Archaic Klazomenian painting see Cook 1981, 1212. Rumpf 1933, 61 separated the Nisyros group and on Rhodes the Vroulia, Kameiros and Euphorbos workshops. In this field, variations in terminological use are frequent, as Koch 1996, 9 n. 27 subsumed the Euphorbos plate under ‘Ostionien’. Mommsen, Schwedt, and Attula 2006, 199-200; Attula 2006, 114-6. As a sub-group to EMEB there is the small group EMEb with six samples (which differs only in higher Cr, Ni and Co values). In addition to the fine pottery also a relief-decorated pithos fragment (Emec 129) belongs to EMEb, see Berges 2002, 139 cat. no. 32. Schefold 1942, 129 figs 1-2; Schefold 1993, 17-8 cat. no. 4. Walter-Karydi 1986, 76; Walter 1968, 89-92. For concluding remarks

see Cook and Dupont 1988, 61 n. 55. 24 Walter-Karydi 1998, 292 n. 5. 25 Thasos 2057, see Boardman 1998b, 131 cat. no. 256. 26 BM GR 1893.11-13.4 and 5; Schefold 1942, 129 figs 1-2. I thank A. Scollan and A. Villing for information and drawings. See Attula 2006, 116-7 pl. 61.2-4. For the identification of this sanctuary with the Apollo sanctuary at Emecik, see Berges 2002, 112-7. 27 Supra n. 21. The technical abbreviation Knid (so far there is Knid 1 only) refers to the origin of the sample from the Knidian peninsula outside Emecik, not the localization of the clay source. For the 137 samples from Emecik see Mommsen, Schwedt and Attula 2006. 28 The upper part of his body is more elaborate and he is definitely bearded. Athens, National Archaeological Museum 14935. ProtoAttic, attributed to the Analatos Painter (c. 700 bc); Morrison and Williams 1968, 73 pl. 8b; Sweeney 1987, 91 fig. 18; Koch 1996, 13-4; Cook 1960, 67. 29 Ahlberg 1971, 25-38. 30 Höckmann 1985, 103, 155; Basch 1987, 242-3 fig. 511. 31 Even Torr 1894, 27, after the announcement of the fragments in the British Museum in 1893, admired them for their ‘grandeur réelle’, and Basch 1987, 242-3 fig. 511, especially appreciated the depiction on no. 3 as ‘une galère différente de la précédente’. 32 Schaus 1985a, cat. no. 353, pl. 21. 33 Delos Museum B 6013; Basch 1987, 243 fig. 512. 34 From Kouklia-Skales, Cyprus Museum T.58/104 (CGIB, 11th or 10th century bc); Yon 1970, 311-7; Iacovou 1988, 19 cat. nos 33, 69-70, figs 77-8; supra n. 15. Well-preserved shallow plate with two remarkably elaborate handles and with a narrative scene underneath the base. Underneath several animals two men with bow and arrow and with two swords are shown fighting a bicephalic snake monster with a dotted body and a forked tail – maybe an early scene of Herakles and the Lernaean Hydra? 35 For parallels concerning the worship of Athena, see Wagner 2001. 36 Supra n. 28. 37 Walter-Karydi 1973, cat. nos 1102, 1100, 1090; Attula 2006, 116 n. 300. 38 Attula 2006, 119-24 nos 211 (Emec 70), 212 (Emec 26), 215 (Emec 2), 218 (Emec 56), 225 (Emec 121), 227 (Emec 43), 232 (Emec 113). 39 For the East Dorian cup (‘Knickrandschale’) with Phoenician graffito (BM GR 1888.4-1.96; sample Nauk 51, EMEBe) see Schlotzhauer 2006, 301-7, no. 2, 316 figs 4-6. I thank the author for providing his manuscript before publication. See also Mommsen et al., this volume; and Schlotzhauer and Weber (forthcoming). 40 On mortaria of Cypriot origin (samples Nauk 55, EMEA; Nauk 67, EMEa; Nauk 68, EMEa), which match most terracotta figurines as well as a sherd from a Cypriot bichrome jug from Emecik, see Villing, this volume. 41 BM GR 1924.12-1.1113. Petrie 1886b, pl. 7.6. I thank U. Schlotzhauer for information, cf. also Schlotzhauer, this volume. 42 Schlotzhauer (forthcomingb); Kerschner and Schlotzhauer 2005, 46-52, fig. 52 (SiA IIa). 43 Mommsen, Schwedt and Attula 2006, 200. Group EMEE includes a black-glazed handle from a krater (Emec 114), an unstamped amphora handle (Emec 24) and a fragment from a terracotta bull figurine (Emec 110); see Kleibl 2006, 178 cat. no. 552. 44 Depot, Inv. ST 01 I8b-10, 96. According to the inventory, this is a ‘Boden mit Standring, grob, oxydierend gebrannt’. 45 Attula 2006, 128, 145. 46 Posamentir and Solovyov 2006, Berezan B 65-40, sample no. Bere 125. I thank the authors for providing their manuscript before publication. 47 Mommsen, Hertel and Mountjoy 2001, 198-201. 48 Schlotzhauer (forthcoming b); Schlotzhauer and Villing, this volume. 49 Schlotzhauer 1999, 239; Schlotzhauer 2001a, 119-21; Schlotzhauer (forthcoming b); Attula 2006, 128 n. 319. For Ionian influence on Archaic Carian pottery production, especially in the late 7th and the first half of the 6th centuries bc, see Fazlýoðlu (forthcoming). 50 Cook and Dupont 1998, 88. 51 Attula 2006, 136-7. 52 For the situation on the lower terrace of the sanctuary, see Attula 2006, 101-2, pl. 17 figs 10-1. 53 Attula 2006, cat. nos 281, 284, 289. 54 Özer 1998, 41-3 figs 31-3. 55 Kerschner and Schlotzhauer 2005. 56 Walter-Karydi 1998, 293; Attula 2006, 114-18. 57 Supra n. 49. On an assumed local variation of stemmed plates see Attula 2006, 131 cat. nos 250, 251 (Emec 65, EMEB), 280. Naukratis: Greek Diversity in Egypt | 91

Attula 58 59 60 61 62 63

Bresson (forthcoming); see also Blümel 1993. Supra n. 46. Dupont 1983; Cook and Dupont 1998, 88-9. Attula 2006, 128. Information U. Schlotzhauer; Schlotzhauer, this volume. Jenkins and Waywell 1997; Walter-Karydi 1998; Bruns-Özgan 2004, 201-8. 64 Mandel, Hübner and Kögler 2000, 161-94; Attula 2006, 113, 146-48 cat. nos 305-31; Berg Briese 2005; Kögler 2005; Þahin 2003. 65 One also needs to consider the assessment by Walter-Karydi 1982, 9 ‘Daß Ostdorisches ... fehlt, ist weniger verwunderlich, denn außerhalb der Ostdoris scheint diese Keramik sehr wenig verbreitet zu sein’; Schlotzhauer (forthcoming b), 6; Attula 2006, 145 n. 344. 66 I would like to refer to the current research, including archaeometric

92 | Naukratis: Greek Diversity in Egypt

investigations, by M. Berg Briese at Halikarnassos and by M. d’Acunto at Rhodes Museum. 67 Mortarium sherd from Emecik, ST 99 K8c-16, 78. 68 Berges 2002, 134-53; Attula 2006, 124 cat. no. 241 pl. 67.2; Attula (forthcoming). 69 Finally, on terminology, see Schlotzhauer 2001a, 111; Kerschner and Schlotzhauer 2005, 5. The adoption of this classification system, which was applied successfully to Archaic Ionian pottery, is not only desirable but also necessary for Dorian pottery. Since the term OdA (Ostdorisch Archaisch = Archaic East Dorian) is at present the only available designation, it is not yet truly distinctive. The application of this terminology may succeed first with regard to local Fikellura wares.

The Non-figured Wares from the Anglo-Turkish Excavations at Old Smyrna Points of Contact with Naukratis

Stavros A. Paspalas Abstract The excavations at Naukratis produced a large quantity of East Greek pottery of various categories which resulted in renewed efforts in their study and classification. The present paper presents a body of material – comprised mainly of dishes and fruitstands, but also a number of plates – from the Anglo-Turkish excavations at Old Smyrna that may be compared to some of the East Greek finds from Naukratis. In the light of the results of more recent excavations most of these Old Smyrna pieces and their parallels at Naukratis may be identified as North Ionian, and similar vessels are now testified to at many sites which attracted the attention of East Greeks during the Archaic period. A contrast, though, is apparent between Old Smyrna and Naukratis – at least as far as the published record is concerned – as regards pieces probably of late 6th- and 5th-century bc date decorated with loose floral schemes, as they occur at the former site, but are not testified to at the latter. The pottery from Old Smyrna discussed in this paper forms a small part of a body of material from the Anglo-Turkish excavations collected and documented by J.M. Cook.1 I have not examined the material, and my knowledge of it is based on Cook’s notes, drawings and photographs.2 The wider corpus is largely comprised of simply painted wares which may be classified as belonging to the broad wave-line and banded categories. There are also a number of more ambitious florally decorated vessels, some of which are discussed here. Firstly, however, the point should be made that if one were to examine the published Naukratite material one would have the impression that plain wares and simply decorated pottery were not found in great numbers at the site. Although a small number of unpainted vessels – all complete or substantially so – were published, the fact remains that Naukratis, as we now have it, cannot provide us with a full range of the pottery groups present at the site during the Archaic period.3 Consequently, the points of contact between the Old Smyrna material included in my wider study and that published from Naukratis are few. They are primarily restricted to fruitstands and dishes that are commonly placed in the Late Wild Goat tradition and a number of plate fragments. All belong to John Cook’s ‘Slipped Wares’ except for the fruitstand 11 (Fig. 9) which he classified with his ‘Stripped Wares’, though it too in all likelihood bears at least a simple slip. His general description of the ‘Slipped Wares’ is: ‘The slip was normally white (sometimes chalky and fugitive), and the glaze was most commonly red… . The fabric is generally a pinkish buff and contains gold mica.’ In Cook’s papers find contexts of only a small number of pieces are noted. In this article attention is focused on presenting the Old Smyrna material and on the links which can be established between the decorative schemes of its constituent pieces and finds from Naukratis. Categories of plates, fruitstands

and dishes found at Old Smyrna but not at Naukratis are also examined in order to present the full range of these shapes from the former site as preserved in Cook’s notes. Given that the find contexts of most of the Old Smyrna pieces are not known they add little to our knowledge of the chronology of the series. The few pieces for which a datable context is recorded were either excavated in the ‘white tuff chips level’ dated by Cook to the late 7th century bc (9–10, Figs 7–8; 13, Fig. 11),4 or in the Temple Deposit dated c. 600 bc (12, Fig. 10).5 Cook and R.V. Nicholls associated both these deposits with the Alyattan destruction which they dated c. 600 bc, though E. Akurgal would lower the date by a period of at least 20 years.6 Cook’s dating, of course, reflects the conventional chronology which is also used by the excavators of most of the other relevant sites. These carefully and elaborately decorated vessels may be earlier than some of the simpler fruitstands and dishes from other sites which are dated well into the 6th century bc. The richly painted fruitstands 9 (Fig. 7) and 10 (Fig. 8) which are dated c. 600 bc find a simpler parallel, as regards decoration, in a fruitstand identified as North Ionian from a burial excavated at Pitane. The burial also contained a Middle Corinthian aryballos, datable c. 600–575 bc according to the conventional chronology. On the basis of the chronology of the Pitane grave (and caution is advised as it only supplies one limited context) 9 (Fig. 7) and 10 (Fig. 8), as well as 7 (Fig. 5) and 8 (Fig. 6), which are decorated in the same manner, could be placed immediately on either side of 600 bc if complexity of decorative design and precision of execution are taken as indicators of an earlier date.7 The possibility, however, must be kept in mind that their ornate decorative schemes may be due to the fact that they were intended to be used as display pieces within a religious milieu as those with a known context were found in the Sanctuary of Athena. Plates The fragment 1 (Fig. 1) preserves the rim of a plate, the concave upper surface of which is decorated with a running scrolled spiral; each scroll carries one pendent and one ascendant drop. The form of the rim may be readily compared with those of other East Greek plates, but it is the decorative spiral which links it to a group which, although small in number, is well defined. The best preserved example is the plate now in Kassel,8 which is reported as having been found at Klazomenai. Ornately decorated with Late Wild Goat motifs in the three zones of its floor, it is the scroll pattern on its rim which relates it to the Old Smyrna fragment. Y. Ersoy has studied two other fragments of such plates from the recent excavations at Klazomenai which are closely related to the Kassel plate, down to the detail of the dots along the band of the rim’s lower edge. Furthermore, he reports five more fragments from similar plates found during survey work in the Klazomenai area, and recent analysis of examples of Naukratis: Greek Diversity in Egypt | 93

Paspalas

Figure 1 Plate no. 1 from Old Smyrna

this type from Berezan have confirmed the localization of production at Klazomenai.9 The descriptions of the fabric and surface treatment of these three plates tally well with those of Cook’s ‘Slipped Wares’. I know of one piece from Naukratis which is of relevance here: the rim fragment of a plate published by E. Price.10 A drawing of its profile, slightly heavier than that of 1 (Fig. 1), appears here as Figure 2. It offers a very close match to the Klazomenian and Kassel examples in all its decorative details. All these pieces are closely comparable to the Old Smyrna rim, but it will be noted that the latter has a simpler spiral motif. On the other plates individual double-scrolled spirals are linked by two parallel horizontal bars. Above and below these bars there is a drop, as there is between the scrolls of each spiral. This scheme also appears on a ‘bowl’ fragment found at Naukratis.11 This motif is a variant of a similar pattern in which the position of the scrolled spirals is actually occupied by concentric circles. At least three such pieces are know from Naukratis, one from a dish, the others from a lid and a krater.12 It may be noted that 1 bears a decorative scheme simpler than that of its parallels. Here there is a running scrolled spiral, not individual scrolled spirals, consequently there is no need for linking bars with the result that the frequency of the drops has been reduced. The rim and outer floor of the Kassel plate are closely paralleled by fragments excavated at Berezan and Olbia, while the scroll pattern also appears on a similar fragment from the latter site.13 However, from what is preserved the individual scrolled spirals appear not to be linked by bars. This observation also holds for a plate fragment from Histria.14 If one was disposed to think in linear developmental terms one could say that these pieces stand between the Kassel plate and its cognates and the Old Smyrna fragment. Closely related decorative motifs are found at Naukratis on a krater handleplate fragment and on a lid fragment where a zone of scrolled spirals sits above one occupied by lotus palmettes and buds of a type we shall shortly see on Old Smyrna dishes and their parallels.15 Cook in his notes assigned the Old Smyrna plate fragment to the 6th century bc, and, indeed, it is to that century that all its parallels have normally been dated, though to wildly varying quarters. The most comprehensively presented pieces, those from Klazomenai, were found in contexts dated to the late 6th century bc,16 while the Olbia fragment is presented with sherd material dated to the second quarter. To date it would appear that open vessels with decorative schemes related to that of 1 (Fig. l) are, within East Greece, concentrated in North Ionia: Old Smyrna and Klazomenai, though it must be admitted that the numbers on which this observation is based are very few indeed. Beyond this region, they are found at Naukratis and in the Black Sea region along, it may be noted, with other North Ionian material. A small number of other examples of plates are found 94 | Naukratis: Greek Diversity in Egypt

Figure 2 Plate from Naukratis (BM GR 1965.9-30.500)

among the Old Smyrna material, again with a white slip, though a grooved rim differentiates them from the piece just examined but not from its related piece illustrated in Figure 2 excavated at Naukratis. They normally bear floral decorative motifs or crosshatching on their rims, though a more substantially preserved example, 2 (Fig. 3), bears geometrical schemes: triangles, series of dots and what may be some form of cruciform motif in its tondo. Its rim can be compared to, but not closely paralleled by, plates found at Buruncuk (‘Larissa-on-the-Hermos’) ascribed by the excavators to the second half of the 6th century bc.17 The plate fragments included under the numbers 3 and 4 (Fig. 4), too, may be generally compared to finds from Burunçuk assigned to the second half of the 6th century bc, though Cook suggests that they may date into the 5th. The zones of crosshatching are reminiscent of those on the top zones of skyphoi from that site,18 while the disarticulated palmettes and other floral elements find parallels on vessels of the ‘Pflanzenornamentik’ category.19 Closer parallels for the decorative schemes of these plates are not available among the published material, but it may be noted that a few fragments from a context dated by the excavators to the mid-6th century bc at Old Smyrna bear loosely-structured florals (though more undisciplined that those of 4 (Fig. 4)).20 A krater dated as Late Archaic from the area of Metropolis shares in the same general repertoire.21 Loosely-structured and disarticulated floral elements appear not to be a feature restricted to the pottery found at Buruncuk. They do not have any parallels among the known material from Naukratis. The fragment 5 (Fig. 4) (and 3 (Fig. 4 top right)) bear properly constituted lotus-bud chains. Cook places the two fragments numbered 32 (Fig. 4) with these plates, and their decorative details are also paralleled by material excavated at Buruncuk.22 The fragments of five different vessels are incorporated under 6 (Fig. 4). Cook describes all three rim fragments as deriving from plates with flat rims. Given the lack of any profile drawings for these pieces it may be that they are similar in form to the dishes discussed below, though shallower. Their simple decorative schemes can be paralleled elsewhere in North Ionia and Aiolis.23 Figure 3 Plate no. 2 from Old Smyrna

The Non-figured Wares from the Anglo-Turkish Excavations at Old Smyrna

Figure 4 Plates nos 3, 4, 5, 6, 15 and 32 from Old Smyrna

Dishes and fruitstands The dishes and the fruitstands are largely to be dated approximately to the first half of the 6th century bc on the basis of parallels with finds excavated elsewhere, though the earliest examples may, as Cook places them, date to the late 7th century bc. Many of the simpler examples are dated by Cook into the 5th century bc. Their exteriors regularly carry series of horizontal lines and bands. They may be divided, on the grounds of size, into two basic groups. Most are small, with a rim diameter between 11 and 15.5cm; a minority are larger with a rim diameter between 21 and 26cm. Most of the parallels from Naukratis fall within the latter range. Typically, they carry a slip on their interior, while the exterior is not slipped. The most impressive of the fruitstands or dishes excavated at Old Smyrna is the piece published by E. Akurgal which pictures a Potnia Theron on its interior. It has long been recognized as a North Ionian work, and I mention it here as a reference point for the dishes from the site to which it is related. It may be noted that its incised lotus bud and flower zones can be profitably compared to those of a fruitstand from the temenos of Aphrodite at Naukratis.24 Figure 5 Fruitstand no. 7 from Old Smyrna

The first of the Old Smyrna pieces, 7 (Fig. 5), is close to the Potnia dish, but is somewhat simplified. It is a dish with cut-out handles, a form known previously from the site in Wild Goat.25 Its horizontal rim carries a dotted running dog pattern, which may be seen as a simplification of the guilloche on the rim of the Potnia dish. Its main zone is decorated with an alternatelylinked lotus flower and bud chain, simpler – as it is not incised – in execution than the chain on the Potnia dish but still embellished with added red. The inner zone bears a series of tongues. It may be noted that the chain’s buds are rather stout. Naukratis supplies a number of examples of similar nonfigural dishes, though none are as elaborately decorated as 7, which is distinguished by the running dog pattern on its rim, whereas all other examples bear meander hooks. Furthermore, the lotus flowers of 7 consist of four elements, those on its simpler parallels only of two. It can be noted, however, that these simpler parallels from Naukratis have more complicated radial arrangements on their floors rather than a series of tongues.26 One dish fragment shares the feature of complicated lotus flowers with 7,27 but they are different in form as they possess an outlined central element. Furthermore, its buds are slender, as are those of the other parallels from Naukratis. The lotus chain on a fragment now in Alexandria stands between the one on this piece and 7, the bud is similar to that on the former, whereas the flower consists of solid elements.28 None of these parallel lotus chains, where the relevant section is preserved, are alternately-linked as is that of 7. Here note may be made of the chain on the interior of a turned-up rim fragment, probably from a fruitstand, illustrated by Akurgal from Old Smyrna.29 Again the lotus flowers have solid elements, but the buds are slender as are those on the cited Naukratite parallels, and there is a pellet beneath each flower, a feature also seen in the chain of 7 as well as on the following piece (8, Fig. 6) from Old Smyrna and – in a simpler form – on a ‘bowl’ lower body fragment from Naukratis and dish fragments excavated on Delos and at Syracuse.30 Despite the differences in specifics between 7 and its Naukratite parallels their shared features show that they belong to the same stylistic tradition. Fragments of a dish simpler than 7, though similar to those from Naukratis, has been excavated at Sybaris.31 The fruitstand 8 (Fig. 6) is even more ornately decorated, though it lacks the narrow encircling meander hook zone seen on so many simpler examples of the shape. The middle zone is occupied by a chain in which the flowers have an outlined central element and two rhomboid secondary petals, and so share features both with 7 and a fragment already looked at from Naukratis.32 The outer zone bears a metopal scheme in Figure 6 Fruitstand no. 8 from Old Smyrna

Naukratis: Greek Diversity in Egypt | 95

Paspalas

Figure 7 Fruitstand no. 9 from Old Smyrna

Figure 8 Fruitstand no. 10 from Old Smyrna

which the ‘triglyphs’ divide concentric circles outlined by solid circles alternating with dots. This is a slightly more complicated scheme than that seen on the exterior of the fruitstand fragment published by Akurgal referred to above.33 The inner zone which surrounds the now lost tondo carries a series of alternating solid triangles and diamonds with smaller diamonds occupying the area between their apices. The overall trizonal arrangement is not met among the other known finds from Old Smyrna, other than the Potnia piece. Most other known East Greek trizonal pieces differ considerably as concerns decorative motifs since they are mostly figured, as a fruitstand from Naukratis shows,34 though its intricate lotus chains provide a point of contact with their un-incised counterparts on non-figural pieces such as 7–9 (Figs 5–7). A close trizonal parallel, however, to 8 (Fig. 6) has been excavated in the vicinity of the Artemision in Thasos on which the inner two zones are decorated with floral motifs, and the outer with a close variant to the scheme found on the corresponding zone of the Old Smyrna vessel.35 Naukratis does, though, supply at least one example of a two-zone dish with a metopal outer register in which triglyphs separate concentric circles encircled either by solid circles, or just possibly stemmed circles.36 The floor of the vessel was occupied by an example of the usual radial motif. Similar dishes with an outer zone on their interior wall that is definitely comparable to 8 (Fig. 6) are known from the site. These dishes have horizontal rims decorated by meander hooks; their outer zones are metopal in arrangement and compare well with that of the trizonal piece.37 There are, however, differences. On the three pieces from Naukratis offered as parallels the painting is not as carefully executed, and the tongues are somewhat more elongated. A dish fragment from Tell Sukas is also to be placed with these pieces as is a fruitstand from Sybaris.38 The differences between these parallels reinforce the special character of 8 and the extra effort invested in its manufacture. The stout tongues of 8, which are characterized by a nearly consistent width, appear on two other fragmentary fruitstands among the Old Smyrna material. On the first of these, 9 (Fig. 7), which was found in the ‘Temenos chips’ layer dated c. 600 bc, they flank a single lotus flower which rests on the groundline of its register. The flower is composed of several elements, the central one of which is outlined – a feature seen, as already noted, in a simpler form, on a fragment from Naukratis.39 The liberal use of added red, along with white, associates this fruitstand with 8 (Fig. 6) and the Potnia vessel from Old Smyrna. The second of these fruitstands, 10 (Fig. 8), comes from the same find context. It too bears an outer zone of tongues 96 | Naukratis: Greek Diversity in Egypt

Figure 9 Dish no. 11 and krateriskos from Old Smyrna

separated, at least on the fragment preserved, by a solid triangle-like ‘leaf’ comparable to those in the inner zone of 8 (Fig. 6). The small rhomboidal elements in its inner field may be compared to those in the corresponding zone on 11 (Fig. 9), a far simpler fruit dish from Old Smyrna as well as on a two-zone dish with monochrome decoration from the Samian Heraion, the outer zone of which carries a series of lotus flowers and buds without a linking chain.40 A number of other tongue-bearing fragments from Naukratis should be mentioned. All carry stout tongues but of a slightly more elongated form than those of 8–11 (Figs 6–9); some occasionally carry added paint. One is very small and preserves little else other than its rilled horizontal rim; a leaf or bud may be seen between the tongues so the scheme can be interpreted as a simpler version of that seen on 10 (Fig. 8).41 The second is similar to the simpler vessels we have seen so far in that its tondo carries a radial scheme while the ‘triglyphs’ divide a single motif, in this instance quartered squares (note the scheme of 11, Fig. 9).42 The third is from a dish with inturned rim. The upper band carries a series of meander hooks, below which tongue ‘triglyphs’ alternate with lotus flowers (compare 9, Fig. 7);43 three other fragments carry the same decorative details, but they have horizontal rims.44 A better idea of what such a dish would have looked like is offered by the example, also from Naukratis, now in Boston, where each metope is occupied by a lotus flower, and the tondo carries a radial motif.45 The Old Smyrna metopal dishes and their parallels from Naukratis are characterized by ‘triglyphs’ comprised of stout tongues. They can be contrasted with corresponding vessels on which the triglyphs consists of rays. Scientific analyses have now confirmed older views that we should generally identify the first group as North Ionian, though with some Aiolian representation, and the second as South Ionian.46 There is, of course, little surprise in finding North Ionian material at Old Smyrna, and it is clear that it also made its way to Naukratis in some numbers, which – again – occasions no surprise given Herodotus’ (2.178) testimony that Teians, Phokaians and Klazomenians were officially involved in the Hellenion. However, the distribution of such material was not restricted to regions where a North Ionian presence is testified to by the written sources. A good parallel for the metopal decorative scheme, with a lotus flower in each metope, can be found on a fruitstand from Pitane found with a Middle Corinthian aryballos,47 and so should date no earlier than the beginning of the 6th century bc according to the conventional chronology. From the same site a

The Non-figured Wares from the Anglo-Turkish Excavations at Old Smyrna

Figure 12 Dish no. 14 from Old Smyrna Figure 10 Fruitstand no. 12 from Old Smyrna

Figure 11 Fruitstand no. 13 from Old Smyrna

dish with metopes occupied by concentric circles encircled by a series of solid circles also came to light.48 Naukratis provides evidence that this scheme common in North Ionia and Aiolis was happily used on contemporary figured wares also assigned to North Ionia, as indeed were groups of tongues with intervening lotus flowers as seen on 9 (Fig. 7).49 Beyond Old Smyrna, its neighbouring sites, Aiolis and Naukratis, the bizonal decorative schemes examined here are found on pieces, usually identified as North Ionian, both with inturned and horizontal rims, in widely distributed regions. Dishes with lotus chains in the outer zone have been excavated at Samos, Naxos, Rhodes, Cyrene, Selinous, Megara Hyblaea, Berezan, Olbia and Pantikapaion.50 A fragment of a dish with a lotus flower flanked by the tongues of a ‘triglyph’ comes from Gravisca and another from Delos, while a close parallel is now known from Amorgos.51 The rims of all these pieces carry a meander hook pattern; an inturned rim fragment from Delos bears a lotus flower between ‘triglyphs’ but lacks the meander hook zone.52 A piece with more elaborate lotus flowers was excavated at Cyrene.53 Similarly, bizonal dishes and fruitstands with concentric circles in metopes, reminiscent of those seen on the trizonal fruitstand 8 (Fig. 6) can be documented at Delos, Thera, Leukas, Kerkyra, Cyrene and Berezan (where the usual meander hook in the outer zone is replaced by a series of dots).54 A scrap of a dish with an inturned rim decorated with a meander hook zone followed by one preserving parts of tongues is also known from Ephesos, while an even smaller piece from Phokaia preserves the same tongue motif but there is only a horizontal band at its rim, not a meander hook series.55 This quick survey of findspots of pieces related to those from Old Smyrna and Naukratis concentrates on a number of sites in the Black Sea region, the Cyclades, and with somewhat less intensity on the west, though one may also note a lid from Selinous on which we see both a lotus chain zone and a concentric-circles-in-metopes zone.56 The impression received from this survey is that 7–10 (Figs 5–8) are more carefully produced products than their parallels cited here. Two of the Old Smyrna fruitstands stand apart. The first, 12 (Fig. 10), was found among the votives from the ‘Cella’ area of the Temple of Athena, and the second in the ‘Temenos chips’ layer, and so both were dated by the excavators to c. 600 bc. The first carries below the outermost zone of meander hooks a series

Figure 13 Fruitstand no. 16 from Old Smyrna

of individual suspended meander elements, and then radiallyplaced petals around a tondo with a solid centre. On the second, 13 (Fig. 11), a series of individual meander elements rise from the groundline; the tondo is comprised of a number of circles and a series of stemmed solid circles. The use of meander elements in this fashion is not commonly met on dishes.57 The closest parallel I have been able to muster is from Naukratis though the meander elements do not supply a perfect match,58 nor do those on a fragment excavated at Selinous.59 A dish with a meander from Tocra is placed slightly earlier than the majority of such examples at that site, i.e. closer to the beginning of the 6th century bc.60 The other major category of dishes and fruitstands from Old Smyrna are those that carry the one-zone decorative scheme in which the ornamental motif is restricted to the floor of the vessel, while the interior upper wall is banded. The dish fragment 14 (Fig. 12) from the City Wall NE shows the scheme at its simplest: meander hooks on the rim, bands on the upper interior wall followed by a series of petals around the tondo. Most of the known fragments from Naukratis which may belong to this class preserve little more than parts of their rim and upper body.61 Two bear sections of their floor decoration, which consists of bud and leaf schemes,62 or radial schemes,63 though more complicated than the decoration on 14. Added red and white paint in the decorative schemes of these Naukratite finds distinguish them from 14. Comparable dishes are also known from Old Smyrna, though too little remains of 15 (Fig. 4) to determine its decorative scheme.64 Many of the comparable fruitstands and dishes which have been found elsewhere, as at Perachora, Akragas and Leontini,65 are decorated with a more complicated series of encircling lines and bands than that found on 14, or alternatively, the bands are of different colours, often with a group of white-red-white horizontal lines set upon a black band or a sequence of blackred-black bands. The dishes published by F. Utili from a cemetery of Assos provide a good range of the various decorative schemes that may be met, as do those from Pitane, Delos, Tell Sukas as well as Histria, Pantikapaion, Myrmekeion and Olbia on the Black Sea coast.66 Northern Aegean sites have also produced examples of this category.67 Such open vessels, characterized by their meander-hook zones, have been associated with groups defined by Neutron Activation Analysis that have been attributed to North Ionian centres.68 The fruitstand 16 (Fig. 13) excavated in the Temenos area carries two sets of white-red-white horizontal lines in the banded upper wall of its interior, and its floor ornament is more Naukratis: Greek Diversity in Egypt | 97

Paspalas

Figure 14 Dish no. 17 from Old Smyrna

Figure 15 Dish no. 18 from Old Smyrna

complicated than that of 14 (Fig. 12). What is preserved is enough to indicate that groups of petals alternated with a triangular leaf-like element, for which I have not found an exact parallel. Nonetheless, the scheme on a dish fragment from Naukratis is close,69 and similar floor schemes occur on dishes and fruitstands excavated at Berezan, Apollonia Pontica, Histria, Corinth, Gravisca, Cyrene, Tell Sukas, and on an example in Cambridge and a dish fragment in Reading.70 Two plates from Tarsus also have a closely comparable floor scheme, although their upper interior walls are covered by a single glazed band,71 a feature which distances them from the dishes and fruitstands presented here. The dish 17 (Fig. 14), with a cut-out rim, was found in a child’s pithos grave at City Wall East.72 Its floor decoration is essentially geometrical in concept rather than floral and does not have any close parallels among the known material from Naukratis. However, we should see the scheme as within the North Ionian decorative tradition despite the fact that it is not floral as are most of its cognates; the floor decoration of a shallower dish from Klazomenai can act as a link.73 A recently excavated fruitstand from Liman Tepe, Klazomenai, offers a parallel for a floor decorated with radially arranged geometrical motifs separated by dot rosettes, while on the floor of a ‘dish’ from Naukratis a dot rosette is positioned between each of the stylized leaves and petals of the radial composition.74 The profile of 18 (Fig. 15) is characterized by a rather ornately worked rim and a sharp carination point on its exterior wall. This profile type is closely paralleled by dishes with meander hook-decorated rims from Assos.75 The dotted angular ‘s’ zone on its rim sets the Old Smyrna piece apart from the more mundane pieces with simple meander hooks, though it does link it to dishes from Klazomenai with ‘s’ zones on their rims.76 Its cut-out rim, and especially the dot rosette in added white on the preserved handle also distinguish this dish. This use of added white may be paralleled by a similarly discrete motif on a rim fragment from Naukratis that probably comes from a bizonal dish with a metopal zone with lotus flowers in its outer register.77 The interiors of the parallels from Assos to 18 are regularly decorated with bands and a radial motif on the floor. The interior of the Old Smyrna dish, though, is plain and can be paralleled by an example with a comparable profile from Assos.78 F. Utili dates all the Assos pieces c. 580–60 bc, a period considerably later than Cook’s estimation of the date of 18 ‘seems 7th century context’ (see Appendix part B infra). The simple fruitstand 19 (Fig. 16) finds its proper place 98 | Naukratis: Greek Diversity in Egypt

Figure 16 Fruitstand no. 19 from Old Smyrna

Figure 17 Fruitstand no. 20 from Old Smyrna

among the output of North Ionian dishes and fruitstands. The profile of its distinctive vertical rim is closely paralleled by finds at Klazomenai, Assos and Tocra, as well as on the far more ornately decorated fruitstand 9 (Fig. 7).79 The use of a series of dots to encircle a central decorative element is not unknown,80 and again appears in a more ambitious form on 9 (Fig. 7) and at Naukratis on a ‘plate’, two dishes and a fruitstand; the latter and the ‘plate’ also have similar floor patterns to that of 19 (Fig. 16).81 The fruitstand 20 (Fig. 17) stands apart. Its floor decoration of radial petals occasions no surprise. What is preserved of its interior’s outer zone – a group of drop-petals that consisted of at least two petals – is unusual (though see too 27, Fig. 22). Both fruitstands and dishes normally carry fully decorated interior zones; here there are clearly extensive areas of this zone which were left undecorated, and groups of drop-petals alone may have appeared at intervals in this field. Furthermore, the treatment of the exterior of the piece also differentiates it from most of its cognates as it carries zones that bear, in Cook’s words, a yellow/white slip. These zones contrast with the painted bands and the ‘natural’ zones. The lack of known parallels for this piece leaves open the possibility that it is to be seen as an example of a group with a restricted distribution within the area of northern Ionia. The remaining fruitstands and dishes are far more simply decorated and do not find parallels among the published Naukratis material. On the basis of the approximate parallels offered by the material excavated at Burunçuk they may be dated to the second half of the 6th century bc or later. The fruitstand 21 (Fig. 18) is characterized by its loose palmette, a feature which may not possibly be related to the suspended drop-petals on the 20 (Fig. 17). Dishes may be decorated, as is 24

Figure 18 Fruitstand no. 21 from Old Smyrna

Figure 19 Dish no. 22 from Old Smyrna

The Non-figured Wares from the Anglo-Turkish Excavations at Old Smyrna

Figure 20 Dish no. 23 from Old Smyrna

Figure 21 Dish no. 24 from Old Smyrna

(Fig. 21), with hatched triangles along with drop-petal formations which distantly reflect their floral origins, or primarily with loosely arranged floral motifs alone, such as 25b and 27 (Fig. 22). These are to be compared to the material dated by its excavators to the 6th century bc from Buruncuk.82 Hatched triangles also dominate the decorative scheme of 25a (Fig. 22); approximate parallels are known from Buruncuk.83 Similarly, the painted schemes on the dishes 22 and 23 (Figs 19–20) find parallels among material excavated at that site; the former may bear a hatched triangle on its outer zone.84 It may be noted that their floral decorative elements are, as are those of the plate fragments 3–6 (Fig. 4), generally more degenerate than the most disarticulated palmette flowers and buds encountered in K. Iren’s Aiolische Tierfriesstilkeramik, which date within the first half of the 6th century bc.85 It is probable that they should be seen as later in date than the more accomplished Old Smyrna dishes and fruitstands examined above, though other examples and parallels from dated contexts would be welcome so that further work on their dating could be undertaken. The interior wall of the small dish 28 (Fig. 22) carries a series of solid circles encircled by dots, a decorative scheme apparently related to the series of more delicate dot rosettes well known among simpler East Greek schemes.86 Even more parochial is the little ‘stemmed dish’ 29 (Fig. 23). It lacks a proper white slip and is covered rather with a ‘thin wash’. While its central decorative element may be reminiscent of earlier East Greek representations of trees,87 it may be best to

Figure 22 Dishes nos 25, 25a, 25b, 26, 27 and 28 from Old Smyrna

Figure 23 Dish no. 29 from Old Smyrna

Figure 24 Dish no. 30 from Old Smyrna

simply consider it a rough cruciform scheme. It would be difficult to relate this piece directly to most of the other vessels presented here, and it may well be that its manufacturer stood beyond the pottery-producing traditions represented by the other pieces. The dish 30 (Fig. 24), with one of probably four lug handles preserved, bears testimony to connections with better-known stylistic traditions. Cook does not record its find context, so there are no external grounds from which to determine its date. The generally loose nature of its painting is comparable to that of a dish excavated at Troy.88 It may be argued that the decorative motifs of 30 indicate that its painter was not unaware of Late Wild Goat filling ornaments such as dotted concentric circles and the quartered square – motifs found on a cognate piece at Naukratis and, of course, elsewhere.89 The decorative scheme – quartered squares separated by groups of tongues – of the middle field of the Naukratis parallel can be seen in a simplified form on the Old Smyrna fruitstand 11 (Fig. 9), where the squares are solidly painted. This nexus of relationships between the decorative elements of North Ionian vessels excavated at Naukratis and Old Smyrna is further reinforced by the krateriskos that is pictured next to 11 (Fig. 9) in Cook’s photograph. The painted scheme on this vessel, the shape of which can be documented also at Klazomenai, Cyrene and Tocra,90 is comprised of the stout tongues seen on many North Ionian dishes and fruitstands, and of meander hooks seen on the same vessels and on its counterpart from Tocra. Many of the simpler dishes and fruitstands examined here find their place in a pottery-producing tradition which included a wider range of shapes, as the Old Smyrna krateriskos indicates. The small dish 31 (Fig. 25) belongs to a category the examples of which are characterized by added white stripes, and occasionally other motifs such as rosettes, on the upper surface of their rims. Such dishes have been excavated at Buruncuk, Tocra, Tell Sukas and Cyrene,91 as well as in the Black Sea region at sites including Pantikapaion and Myrmekeion.92 While the category is well represented among the published finds of the first three sites, neither the profiles of the Old Smyrna dish nor its decorative scheme are exactly paralleled. The added white elements on the rim of 31 are more tongue-like than simple stripes, and the interior and exterior surfaces of these dishes

Figure 25 Dish no. 31 from Old Smyrna

Naukratis: Greek Diversity in Egypt | 99

Paspalas carry more bands than are usually found on their parallels. Despite the fact that the sites at which parallels for this vessel have been found produced North Ionian material that corresponds to finds made both at Naukratis and Old Smyrna, no vessels from Naukratis have been published which may be closely compared to 31.93 In conclusion An examination of the plates, dishes and fruitstands from Old Smyrna shows that they belong to categories characteristic of North Ionia, features of which are also met on vessels produced in Aiolis. The earlier Old Smyrna pieces find, by and large, parallels in the corpus of published pottery from Naukratis, though the most elaborate pieces stand apart. What direct links exist between the relevant pieces from these two sites are further strengthened by finds made elsewhere which are related to both bodies of material. These sites are located primarily on the Black Sea coast, and in the Cyclades, the eastern Mediterranean and Sicily, though a greater concentration of finds is noticeable in the Black Sea area and especially at Naukratis. The distribution of these end of the 7th- and 6thcentury bc vessels largely corresponds to the routes of East Greek mariners and merchants who we know, from the written and epigraphical sources, were conspicuous at Naukratis.94 It may also be noted that decorative motifs which characterize many of the Old Smyrna pieces and their parallels are also found on other North Ionian shapes, and so these open vessels are tied into a broader pottery manufacturing milieu. Equally notable from the above discussion is that the relationship between the Old Smyrna and Naukratis material does not hold for 3–6 (Fig. 4) and 20–30 (Figs 17–24). Vessels decorated with loosely-arranged or disarticulated floral motifs are rarely present beyond their areas of manufacture, and are not found among the known Naukratis finds nor among other finds of later Greek pottery made in Egypt. Of the dishes and fruitstands from Naukratis mentioned in this paper only a few have a published findspot, and all these are recorded as having been excavated in a temenos.95 A number of the pieces from Old Smyrna, and notably some of the more ornate examples, were found in the Athena Sanctuary.96 Many of the parallels from other sites cited here, for example those form Tocra and Cyrene, were also found in sanctuaries. This evidence indicates that dishes and fruitstands could frequently be encountered in East Greek sanctuaries, either as votives or items of equipment. Equally, though, material such as that from Assos and 17 (Fig. 14) from Old Smyrna show that these vessels could be placed in funerary contexts, while the parallel pieces studied by Ersoy from Klazomenai appear to come from domestic deposits, as well may 18 and 20 (Figs 15 and 17).

100 | Naukratis: Greek Diversity in Egypt

Appendix The aim of this Appendix is to provide the reader with the ‘raw material’ of this paper as presented in J.M. Cook’s notes. It is divided into three parts. Part A is a concordance which gives the numbers used in this article for the Old Smyrna material and their corresponding numbers in Cook’s notes. In Part B I give a list of most of the pieces discussed in this paper. The list is divided into three sections defined by Cook. The first number in the left-hand column is the number assigned to the piece in Cook’s notes; the following number is the number of the piece in the text of the present article. In the right-hand column I give any details Cook included by his drawing of the piece. (Note that the added purple mentioned by Cook in his notes for the drawing of ID [13] is not indicated on his drawing.) The details which accompany the drawings should be read in conjunction with Part C of the Appendix which is comprised of Cook’s notes on his categories. Cook’s no. 20 in Part C is not presented here as it has been published in Cook 1958/9, 33, pl. 6.e. I have not been able to find any details or drawings for the pieces II and IJ among Cook’s notes. It is conceivable that these pieces may be represented among the vessels referred to in Cook and Nicholls 1998, 23–6 (see n. 4 supra). In order to avoid the risk of introducing any unwarranted features into Cook’s drawings they have been inked exactly as he prepared them. Consequently, the exterior decoration of 16 is only shown summarily as is that of 21. Note that, uniquely, 19 is illustrated by an inked drawing found in Cook’s notes. Part A Concordance No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 25a-b 26 27 28 29 30 31 32

Cook’s No. 10 (drawing) 14 12 12 10 photograph 15 IK IH IE IC No number IB ID IN 9 IF IL IM IA IG 2b 16 17 18b 3 5 4 7 6 2a 19 18 13

Figure 1 3 4 4 4 4 5 6 7 8 9 10 11 12 4 13 14 15 16 17 18 19 20 21 22 22 22 22 22 23 24 25 4

The Non-figured Wares from the Anglo-Turkish Excavations at Old Smyrna Part B

Slipped Ware: dishes and plates, 6th and 5th centuries

Slipped Ware: dishes and plates, 7th century

Note: Plain text: notes on drawings. Italics: instructions on drawings IA. IB.

19 12

drawing drawing

IC.

10

drawing

ID.

IE.

IF.

13

9

16

drawing

drawing

drawing

IG.

20

drawing

IH. II.

8

drawing

IK. IL.

7 17

drawing drawing

IM.

18

drawing

IN.

14

drawing

IJ.

No details (K318) Temenos H3 Buff micaceous clay. On inside black glaze design on white slip. Black band on inside rim on natural. Presumably late 7th century. (K370) Temenos chips. Light buff clay, glaze fired red on interior over white slip; colour (possibly originally white). Stripes of red on exterior on natural. Late 7th century. Probably Smyrnaean. (K383) Temenos chips. Clay pink in biscuit and little mica; gray on surface and black paint. Interior slipped and faded paint partly retouched in purple. Black stripes on exterior on natural. Late 7th century. (K384) Temenos chips. Reddish clay and mica. Exterior, buff dark brown glaze. Interior and exterior of lip slipped; purple retouches. Dark brown stripes on exterior on natural. Late 7th century. (K371) Temenos 820-750. Rather porous light buff clay. Very little mica and white slip on inside and faint traces of wash outside (perhaps thinned clay). Purple and white stripes. 7th century, not latest. (K367) H XIV G? 1000-770. Brown clay rough and reddish in break, micaceous. Exterior: dark glaze on yellow/white slip. Interior: dark and light brown paint on white slip. Light brown paint: the two narrow bands around floor rosette and the inner circles. Should be 7th century. No details. No details, and no illustration included among drawings or photographs. No details, and no illustration included among drawings or photographs. No details (K304) City Wall E, pithos grave. Buff clay with very little mica. Dark brown to black glaze. Exterior: black on natural. (K365) H XIA 10.74-9.97. Buff clay, little mica. Brown glaze on thin white slip. Dot rosette painted on white on handle spur. Seems 7th century context. (K319) CW NE. Dark brown paint on white slip. 7th century.

2a. 2b. 3. 4. 5. 6. 7. 9. 10. 10. 12. 13. 14.

29 21 25 26 25 a & b 28 27 15 1 5 3&4 32 2

drawing drawing photograph photograph photograph photograph photograph photograph drawing photograph photograph photograph drawing

15. 16. 17. 18.

6 22 23 31

photograph drawing drawing drawing

18b.

24

drawing

19.

30

drawing

P46. No details. No details. No details. No details. No details. No details. No details. No details. No details. No details. No details. No details. P201. H West (North) 1310-1300 Plain outside. No details. 6th century or later. No details. Stripes outside and white painted lines on rim. C pre-wine shop. Red, rather rough ware, much mica. Surface badly corroded. Exterior plain. On interior traces of red painted design on white slip: double chain of orange pips on rim, traces of hatched triangle and tongues in bowl. Reddish clay with light wash; polished, micaceous. Thinned clay swastika on handle. To be restored with four lugs.

Striped Ware fruit dish with Striped Ware jar IA No no.11

‘Found in level distinctly earlier than the destruction.’

Naukratis: Greek Diversity in Egypt | 101

Paspalas Part C: John Cook’s Notes97 Dishes and plates 7th century Stemmed dishes and plates with Orientalising ornament were found in considerable numbers in late seventh-century levels. They were not systematically studied, and there are no notes of the pieces shown in the photographs. Pieces illustrated here are numbered IA-N. The clay is usually reddish or buff, with mica and the glaze red or dark brown. Applied purple is common, white is only occasionally found (apart of course from the slip), and on IM where a dot rosette seems to have been painted in white. On IC there appeared to be an orange paint (unless it were discoloured white) on some of the tongues in addition to purple and the red glaze. A high proportion of these dishes and plates were found at the Temenos, but two of those recorded came from habitation areas (IG and IM) and two (IL and IN) from the City Wall east. No. 9 in the following list may also be of the late 7th century. 6th and 5th centuries The ware is normally reddish and contains much gold mica. The slip is white, generally thin and often rather fugitive. The glaze is normally red, almost always so in the abundant material that seems to be of relatively late date. Slipped dishes and plates do not, however, seem to have continued into the later part of the 5th century. Proveniences are not well recorded, but the majority of the pieces illustrated probably date about the beginning of the 5th century: no. 3 was from a sixth-century level, nos 2a, 13, 17 and 18 were from contexts earlier than the late 5th century; no. 9 could be late seventh-century. Some of the numbers in the illustrations comprise a plurality of fragments of similar vases. A little stemmed dish in this ware is attested by a number of examples (e.g. nos 2a-b and 3). A fragment with a radiating petal pattern in the centre of the bowl was found in the midsixth-century deposit in the Temple Pylon, and no. 3 is dated sixth-century. No. 2a, of porous buff clay with a thin wash and streaky glaze, is less carefully executed and probably fifthcentury. The spiral pattern shown on no. 4 also occurs in the centre of the bowl of little stemmed dishes. More commonly, however, to judge by the fragile and comminuted material, the little dishes of this ‘fruit dish’ profile had a low ring foot in place of the high stem. These little dishes are generally less than 18cm in diameter. Common zone patterns are cross-hatched triangles and wheels, short pendants of drops and palmettes; disintegrated lotus flowers occasionally appear (cf. no. 5, on the left), and one fragment (no. 5, bottom row next to left) shows a streaky swathe that resembles marbling. Radiating petals,

102 | Naukratis: Greek Diversity in Egypt

concentric rings, or the spiral pattern of no. 4 occupy the centre of the bowl. A small number of fragments come from small dishes of thick fabric and a diameter of about 13.5cm, with a thicker white slip which is generally confined on the underside to a band below the lip (as no. 7, upper); no feet were recovered to match them. This well slipped ware might be Chian. It seems to date about the early 5th century. On one fragment there seemed to be a deliberate contrast of a darker shade of glaze along with the red. The commonest form among the little slipped dishes is a plate with out-turned rim. The shape seems to differ from that of the late 7th century in that the upper surface of the rim tends to be more convex. It is often set off by grooves; occasionally there are grooves on top and bottom of the body of the plate (as no. 13). Apart from rare fragments of rims with a lotus chain (no. 10 photograph) or spiral pattern (no. 10 drawing) and the big plate with an animal band (BSA 60 [1965] 120 no. 32), these plates were of modest size (under 24cm in diameter). Lotus chains and meander were found decorating the rims, and cross-hatched triangles the body (one example being from a mid sixth-century context). But the commonest elements in the decoration are crisscross work and the drops and swags like those on slipped ware jars or dinoi from relatively late levels (...). Numerous plates are of a smaller format and have a flat rim (examples no. 15 and no. 18). Many of them have simple glaze rings in the bowl, with at most radiating petals in the tondo, and sets of drops on the rim; occasionally the drops are not done in glaze on the slip but painted in white on a glazed band. Here again one fragment showed a deliberate contrast of red and dark tones in the glaze. No. 19 is a freak; it was not available for study in the workroom. Diam. 11.5cm. Micaceous reddish ware, polished, with a light wash; brick-red glaze. The surviving lug (one of four?) had the swastika on the handle painted in thinned clay. It is not clear whether this piece is seventh-century or late, nor whether it would be better listed among the striped wares. No. 20 is exceptional. Only the upper fragments in the photograph (a) had a good white slip; diam. c. 22.5cm; buff with some mica. The fragment (b) was found at the Temple in a late seventh-century context. The vases depicted should be Chian. The exteriors or undersides of these dishes and plates vary in treatment. Those here illustrated were noted as follows: exterior plain nos 12, 14, 16, 18B; exterior streakily glazed no. 2; partly glazed no. 6; with glazed stripes (where noted) no. 20; exterior slipped no. 3 (with glazed stripes). no.4, no.5 (some with glazed stripes), no. 7 (see above), no. 9, no. 10, no. 15 (some with glazed stripes), no. 17, no. 18 (with glazed bands). No. 13 had two perforations on the rim for suspension.

The Non-figured Wares from the Anglo-Turkish Excavations at Old Smyrna Illustration credits

Fig. 2 photo the British Museum, drawing Kate Morton; all other drawings and photos R.M. Cook; drawings inked in by Anne Thomas.

Notes 1

2 3

4 5

6

7

8 9 10 11

12

13 14 15

16

17 18 19 20 21 22 23

I thank Professor J. Boardman for suggesting to the late Professor R. Cook that I be entrusted with the publication of this material. The cost of the inking of Cook’s pencil drawings was borne by the British School at Athens. See the Appendix for Cook’s notes on the pieces examined here. For this problem see Schlotzhauer and Villing, this volume; Kerschner 2001, 75-6; Petrie 1886b, pl. 4 (mainly impressed wares and mortaria) and pls 16-7 (mainly transport amphorae); Hogarth et al. 1905, 123-6 (C.C. Edgar). See Bernand 1970, pl. 19 for coarse ware fragments that carry graffiti. Cook and Nicholls 1998, 18-9, 22. Ibid. 24-6. An ‘East Greek standed bowl with central star’ (K316) is also recorded (ibid.) from this deposit, but it does not appear in Cook’s notes and illustrations. Nor do the fruitstands ibid. 23 SF 137778 from the ‘Alabastron Deposit’ of c. 600 bc. Akurgal 1983, 72-113 (Cook and Nicholls 1998, 27 for the opposing view). James 2003, 262 would lower the destruction date by 20 years. Bowden 1991, 52-3, with particular reference to Naukratis, would generally down date the conventional pottery chronology by 40 years. Pitane fruitstand: Ýren 2003, 152-3 and 186 no. 306, pl. 65. The unpublished thesis by M. Manyas (Oryantalizan Stil Tabaklarý, Ankara University 1984) in which East Greek dishes and fruitstands, and their contexts, are discussed was unavailable to me. CVA Germany 35 Kassel 1 pl. 15.6. Ersoy 1993, 176 no.422, 242 no. 423 and pp. 386-7. For the Berezan pieces, see Posamentir and Solovyov 2006, samples Bere 139 and 142 (group E). Price 1924, pl. 7.3 (BM GR 1965.9-30.500). Venit 1988, pl. 31.120 (Alexandria 16893). See also the second zone of the floor of a plate excavated at Tocra: Boardman and Hayes 1966, 67 no. 819 (=Boardman and Hayes 1973, 17 no. 819, pl. 9). For the incorporation of the basic motif into a more complex double-spiral motif: Venit 1988, 18 no. 58, pl. 14 (Cairo 26137, possibly from Naukratis). Price 1924, 196, pl. 7.2 (BM GR 1965.9-30.527); Fairbanks 1928, 108 no. 319.8 (Boston MFA 88.1085), pl. 34 (though note the multiplication of the ascendant drops); CVA Great Britain 9 Oxford 2 IID, pl. 4 no. 24 (G119.48). A fruitstand with a similar zone found in a tomb at Kameiros (Papatislures Grave 2) has a context date of c. 600–575 bc: Jacopi 1932/3-41a, 19 no. 4 fig. 5, pl.3; Gates 1983, 6. Note, too, the dish fragments: CVA Great Britain 12 Reading 1, pl. 22 no. 4b (26.ii.34) and CVA Germany 10 Heidelberg 1, pl. 2 no. 19 (I 15) (with added red). For a parallel to the krater handle plate fragment: Ploug 1973, 65 no. 274, pl. 14 (Tell Sukas). For examples of the motif expanded to a grander scale see: Boardman and Hayes 1966, 47 no. 589, pl. 30; Skudnova 1988, 36-7 no. 4; Korpusova 1987, 40 fig. 15.1; Butyagin 2001, 190-1 fig. 8. Berezan: Kaposhina 1956, fig. 2 third row right-most sherd. Olbia: Levi 1972, 45-7 fig. 13.1; Rusiaeva 1999, 77 fig. 1 bottom row, rightmost sherd. Alexandrescu 1978, 50-1 no. 133. Fairbanks 1928, 107-8 nos 318.2 (Boston MFA 86.557 [30]) and 319.7 (Boston MFA 88.849 [72]), pl. 34. The lotus palmette and buds are simplified versions of those seen on 7. It may be further noted that the double scroll with dots pattern in the upper part of the shoulder of the Apries amphora (Bailey Fig. 1) may also be compared to the scroll patterns discussed here. Ersoy 1993, 386-7. Note the fruitstand from Kameiros (n. 12 supra) on which the place of the spirals of 1 is occupied by circles. Its context is dated c. 600–575 bc and this may indicate that this decorative scheme could be earlier than that with spirals. Boehlau and Schefold 1942, 153-4 fig. 64 c and d (‘Pflanzenornamentik’ group), though more ornate. For the group of triangles on its rim ibid. 82 pl. 35.3 (‘Stufe V’). Ibid. 155 no. 3, pl. 53.11-12. Ibid. 154, pl. 51.9 and 12. Cook 1958/9, 29, pl. 6b; Cook 1985, 26-7 a. Meriç 1982, 47 and 108 no. K47, fig. 115. Boehlau and Schefold 1942, esp. pl. 51.12. For the petal-group rim pattern: Özkan 1999, 54 no. 98 (Pitane). For

24

25 26

27 28 29 30 31 32 33 34 35 36

37 38 39 40 41 42 43 44

45 46 47 48 49 50

the interlocking S rim pattern: ibid. 42 no. 68 (Pitane); Utili 1999, 178 no. 211 fig. 12; Boehlau and Schefold 1942, pl. 31.1-2 and pl. 49.20 (Buruncuk). Groups of bars: Utili 1999, 178-9 nos 214-5, 217- 8. Groups of petals: ibid. 179 no. 222 fig. 13. For the hatched triangles see nn. 83 and 84 infra. For a plate and fruitstand with more disciplined decorative schemes similar to that on 6 bottom right from a Pitane grave with a context date of c. 580–570 bc see Greenewalt 1966, 200 f and g, pl. 13E-F. Potnia Theron: Akurgal 1950, 64, pl. 10b; Walter-Karydi 1973, 146 no. 186, pl. 122. Excavated in the ‘white tuff chip stratum’: Cook and Nicholls 1998, 22. Naukratis fruitstand: Gardner 1888, 44, pl. 9.2-4 (BM GR 1888.6-1.538b-c); Walter-Karydi 1973, 136 no. 185, pl. 120. Note, too, a fruitstand from Selinous: n. 33 infra. Hundt and Peters 1961, 19-20 no. 120, pl. 11. Petrie 1886b, pl. 7.1 (BM GR 1886.4-1.1309; for which see Gardner 1888, 44 Type F.a.1); Fairbanks 1928, 114 no. 324.7 (Boston MFA 88.973 [125]), pl. 35; Hayes 1992, 205-6 nos N9-N11 (Toronto 910x 234.44, Toronto 910x234.2, Toronto 910x234.17; CVA The Netherlands 2 Musée Scheurleer 2 IID pl.2.3 (T2911) (fruitstand). See too a fragmentary fruitstand from Sybaris: Guzzo et al. 1972, 96 no. 142, fig. 96. Petrie 1886b, pl. 7.3 (BM GR 1924.12-1.1104). Venit 1988, pl. 27.100 (Alexandria 17264). Akurgal 1983, fig. 92. Naukratis: Venit 1988, 27 no. 92, pl. 26 (Alexandria 9475). Delos: Dugas 1928, 39 no. 62, pl. 13. Syracuse: Orsi 1918, col. 528 fig. 115, top right. And the Potnia Theron dish: n. 24 supra. Guzzo et al. 1972, 95 no. 139, fig. 94. Its rim carries a series of meander hooks. Petrie 1886b, pl. 7.3 (BM GR 1924.12-1.1104). See n. 29 supra, and compare the plate with a winged figure Dehl-von Kaenel 1995, 352 no. 3454, pl. 61 (Selinous, first third of the 6th century bc). Gardner 1888, 44 Type F.a.2, pl. 9.1-4; Walter-Karydi 1973, pl. 120. 985. See too ibid. pls 120-1. 980-1 (Selinous). Daux 1966, 944-6, fig. 22 ‘coupe rhodienne’. Venit 1988, pl. 27.96 (Alexandria 17233). Note the ‘dish’, also from Naukratis, with this scheme in its outer zone, though its ‘triglyphs’ consist of elements that are better termed ‘drop-petals’ (cf. the floor motif of 16, Fig. 13): CVA Great Britain 11 Cambridge 2 pl. 18.29 (N.25). Compare the use of drop-petals on the fruitstand Ýren 2003, 186 no. 307, pl. 65 (identified as North Ionian: ibid. 152-3). Fairbanks 1928, 112 and 114 nos 323.2 (Boston MFA 89.937 [81]), 323.3 (Boston MFA 86.534 [82]) and 324.9 (Boston MFA 86.545 [126]), pl.35. Tell Sukas: Ploug 1973, pl. 15.289. Sybaris: Guzzo et al. 1972, 96 no. 140, fig. 95. Petrie 1886b, pl. 7.3 (BM GR 1924.12-1.1104). 11 (Fig. 9) is illustrated in the catalogues: Dedeoðlu 1993, 12 middle photo ‘Footed Vessel’; Özkan 1999, 54 no. 100 (first half of the 6th century bc). Samian Heraion: Isler 1978, 148 no. 511, pl. 69, Beil. 12. CVA Great Britain 9 Oxford 2 IID pl. 4.20 (G 117.11). Ibid. pl. 4.19 (1912.36.6). Venit 1988, pl. 27.101 (Alexandria 17240). Ibid. pl. 27.100 (Alexandria 17264); CVA Belgium 3 Brussels 3 IID pl. 3 nos 15 (A1776) and 17 (A2042) (the latter with a wavy line, rather than meander hooks, on its rim); and possibly the fruitstand Venit 1988 pl. 29.113 (Alexandria 9351) and Piekarski 2001a, pl. 12.3 (Bonn 697.18). Fairbanks 1928, 112 no. 323.5 (Boston MFA 88.826 [78]), pl. 35. Kerschner 2001, 85-7; Akurgal et al. 2002, 38-9, 90-2 (M. Kerschner); Ersoy 2003, 255. For the commonalities between these vessel forms from North Ionia and Aiolis see Utili 1999, 28. See n. 7 supra. Ýren 2003, 186 no. 305, pl. 65. Walter-Karydi 1973, 147 nos 1020-1 (from the Samian Heraion), pl. 124. Samos: Boehlau 1898, pl. 12.2 (fruitstand). Naxos: Bikakis 1985, pl. 5.51a-b. Rhodes: CVA Germany 33, Berlin 4, pl. 162.1 (V.I.2958) (Siana). Cyrene: Schaus 1985a, 64-5 no. 363, pl. 22. Selinous: Dehlvon Kaenel 1995, pl. 64.3511, pl. 65. 3561 and 3565. Megara Hyblaea: Vallet and Villard 1964, 80, pl. 22.2 and 4. Berezan: Kopeikina 1981, 197 fig. 4ã, and see, too, Posamentir Fig. 12 this volume. Olbia: Kaposhina 1956, fig. 8 left; Korpusova 1987, 41 fig. 16.4 (more complicated meander hooks). Pantikapaion: Sidorova 1992, 134 fig. 3 A. See, too, the fragment CVA Belgium 3 Brussels 3 IID, pl. 3.16 (A2413), which is given the provenance ‘Kertsch (?)’. Note, too, the Naukratis: Greek Diversity in Egypt | 103

Paspalas

51

52 53 54

55 56 57

58 59 60 61 62 63 64

65 66

67

68

69 70

fruitstand from Saint-Blaise: Bouloumié 1992, 196 no. 518 figs 50a-b and p. 271. Gravisca: Boldrini 1994, 97 no. 160. Delos: Dugas 1928, 40 no. 64, pl. 8. Amorgos: Marangou 1993, pl. 119a (Marangou 2002, fig. 121.4). A fragment of a dish with a similar inner zone scheme from Assos has tentatively been identified as Rhodian: Utili 1999, fig. 11.192. Daux 1963, 865 fig. 4 bottom right. Schaus 1985a, 64 no. 356, pl. 21. They can be compared with that on another fragment from Amorgos: Marangou 1996, 294 and 97 fig. 11a; Marangou 2002, fig. 121.7. Delos: Robert 1952, 37 fig. 34.3. Thera: Pfuhl 1903, 176 no. G2, Beil. 23.2. Lenkas: Fiedler 1999, 413; Fiedler 2003, 367 no. 2106 pl. 162 (no. 2105). Kerkyra: Kallipolites 1956, 160, pl. 61g. Cyrene: Schaus 1985a, 60-1 nos 331-32, pls 19-20. Berezan: Skudnova 1960, 162 no. 4, fig. 11.2, see also now the fragment illustrated at the top right of Posamentir Fig. 4 in this volume. Ephesos: Kerschner 2001, pl. 12.3. Phokaia: Özyiðit 1994, fig. 37 top right (which may be compared to the fragment Miltner and Miltner 1932, 183 fig. 92.22). Dehl-von Kaenel 1995, pl. 65.3568a. For a more elaborate piece, excavated at Kameiros, with a far more complicated meander: Jacopi 1932/3-41a, 19 Papitsilures Grave 2 no. 9 fig. 3, pl. 3 (for the context date see Gates 1983, 6 c. 600–575 bc). Note too the fruitstand Özkan 1999, 53 no. 95. For a series of more complicated meander elements on the rim of a fruitstand excavated at Pitane and Aiolic in character see Greenewalt 1966, 195 h pls 11A and 12E (p. 198 for its context date of c. 570–565 bc). Fairbanks 1928, 112 no. 323.1 (Boston MFA 86.633 [83]), pl. 35. Dehl-von Kaenel 1995, pl.65.3553. Boardman and Hayes 1966, 44 and 50 no. 632, pl. 34. Venit 1988, pl. 28.104-6 (Alexandria 17333, Cairo 26147 and Alexandria 17276); Piekarski 2001a, pl. 12.1 (Bonn 697.18). Fairbanks 1928, 112 no. 323.4 (Boston MFA 88.975 [79]), pl. 35. Note, too, the decoration of ibid. pl. 35.324.6 (Boston 88.821 [124]). CVA The Netherlands 2 Musée Scheurleer IID 2 pl. 2.1 (T2912). Old Smyrna: Özkan 1999, 55 nos 101-2 (somewhat more complicated decorative schemes than on 14-15 [Figs 4 and 12]). Note too the meander-hook rim fragments from Urla, in the wider vicinity of Old Smyrna: Meriç 1986, 303 fig. 3 nos 17-8. Perachora: Shefton 1962, 374-5 no. 4056, pl. 156. Akragas: de Waele 1971, 96 no. 77, pl. 16. Leontini: Rizza 2000, 100 no. 158 fig. 62. Assos: Utili 1999, 168-9 nos 130-4, fig. 8, nos 137, 139 and 142, fig. 9. See too an example from Troy: Blegen et al. 1958, 269 no. 38.1245 fig. 296. Pitane: Özkan 1999, 56 no. 105. Delos: Dugas 1928, 40-1 nos 689, pl. 13. Histria: Alexandrescu 1978, 48-9 nos 118, 121-2, pl. 12. Pantikapaion: Sidorova 1962, fig. 6A,1. Myrmekeion: Butyagin 2001, 192-3 fig. 10. Olbia: Knipovich 1940, 97 fig. 8. See also now fragments from Berezan in Posamentir Fig. 4 (bottom left) and Fig. 10 (centre bottom) this volume. Tell Sukas: Ploug 1973, 68-9 nos 296-8, 300-2, pl. 15. For monochrome examples like 14 [Fig. 12]: Boardman and Hayes 1966, 50 nos 646-50, pl. 36, but with alternating buds and leaves. For a monochrome fruitstand ibid. 50 no. 621, pl. 34. For a rim and upper body fragment from Phokaia: Özyiðit 1994, fig. 37 bottom left; and another from Gela: Orlandini and Adamesteanu 1962, 398 no. 3, fig. 72A. See too Calvet and Yon 1978, 45 fig. 2 for examples from Salamis, Cyprus. Akanthos: Kaltsas 1998, 165 no. E39, pl. 176e. Kavala (Neapolis): Bakalakis 1938, 116-8 fig. 8. Ainos: Baþaran 2002, 78 fig. 8. Note, too, comparable vessels from Chios: Kourouniotes 1915, 79 fig. 15 bottom right; Kourouniotes 1916, 205 fig. 24 top. As well as the examples identified as Chian from Emporio: Boardman 1967, 164 no. 795 and 165 no. 805, pl. 61 (with inturned rim). Akurgal et al. 2002, 75-6 fig. 77, associated with ‘Gruppe B/C’ (M. Kerschner); Kerschner 2001, 85-7 (Teos suggested as a possible manufacture centre). Old Smyrna has been suggested as the production centre of ‘Gruppe F,’ another North Ionian category: Akurgal et al. 2002, 83-4 (M. Kerschner). For the suggestion of local production at Kyme, just north of Ionia, of vessels with meander hook zones: Frasca 1993, 55 fig. 9 and p. 64 no. 28. CVA The Netherlands 2 Musée Scheurleer 2 IID pl.2.1 (T2912). Berezan: Skudnova 1960, fig. 11.1, see too Posamentir Fig. 2 (bottom right) in this volume. Apollonia Pontica: Nedev and Panayatova 2003, 98, Table II,3. Histria: Alexandrescu 1978, 50 no. 125, pl. 12. Corinth: Williams et al. 1974, 21 no. 23, pl. 4. Gravisca: Dehl-von Kaenel 1995, 361-2, pl. 64. Cyrene: Schaus 1985a, 67 no. 392, pl. 23. Tell Sukas: Ploug 1973, 69 no. 304, pl. 15. Cambridge: CVA Great Britain 6 Cambridge 1 IID pl. 7.6 (131). Reading: CVA Great Britain 12

104 | Naukratis: Greek Diversity in Egypt

Reading 1 pl. 21.33b (26.ii.31). 71 Hanfmann 1963, pl. 101 nos 1493-4. 72 Özkan 1999, 55 no. 103 (first half of the 6th century bc). It is possible that this burial is to be identified with the one excavated in quadrant F8.xxi (Nicholls 1958/9, 65-6), although its vessels were identified as ‘... a feeding bottle, a cup, and a diminutive “fruit-stand”.’ Nicholls compares these ceramic finds to those from what the excavators identified as the Alyattan destruction deposit of the end of the 7th century bc. For the Archaic child burials at Old Smyrna see: ibid. 446, 48, 93 and 134; Mariaud 2006. 73 Ersoy 2003, pl. 43A. 74 Liman Tepe: Erkanal et al. 2002, fig. 1. Naukratis: CVA Great Britain 11 Cambridge 2, pl. 18.29 (N.25). 75 Utili 1999, 169-70 nos 137 and esp. 142 and 151, fig. 9. 76 Ersoy 2003, 255 pl. 43A-B. 77 Price 1924, pl. 7.1 (BM GR 1924.12-1.220). 78 Utili 1999, 171 no. 153, fig. 9. 79 Klazomenai: Ersoy 1993, 82 no. 572 and p. 392, pl. 63. Assos: Utili 1999, 184 fig. 14 no. 254 (pp. 27-8 identified as an import). Tocra: Boardman and Hayes 1966, 50 no. 627 fig. 24. 80 Schaus 1985a, 65 no. 371 pl. 22 (Cyrene, c. 580–560 bc); Sidorova 1962, 116 fig. 6A.3 and 5 (Pantikapaion); Boardman and Hayes 1966, 50 no. 633, pl. 34 (Tocra); Ersoy 2003, 255, pl. 43A (Klazomenai). For a series of dots on the outer zone of a rim: Dugas 1928, 39 no. 62, pl. 13 (Delos); Miltner and Miltner 1932, 178 fig. 90.16 (Old Smyrna); Iþik 1989, 57 no. 36, pl. 7 (Klazomenai); CVA Great Britain 9 Oxford 2 IID pl. 4 no. 19 (1912.36.6) (Naukratis); Bikakis 1985, 55-6 no. 50 (Naxos). 81 CVA Great Britain 11 Cambridge 2, pl. 18 nos 30 (99, N.235) (‘dish’), 36 (94-6, N.21) (‘plate’); Fairbanks 1928, 112 no. 323.4 (Boston MFA 88.975 [79]) and 114 no. 324.6 (Boston MFA 88.821 [124]), pl. 35. 82 25b and 27: Boehlau and Schefold 1942, 82, pl. 35.5 and 9, pl. 51.10. For the shape of 26: ibid. pl. 35.9 and 13 (fruitstands). Hatched triangle: ibid. pl. 35.8. 83 Ibid. pl. 35.8 (fruitstand). 84 Ibid. 82. 23: ibid. pl. 35.3. 85 Ýren 2003, pl. 29.239, pl. 53.290. 86 For the scheme see: Boehlau and Schefold 1942, pl. 35.1; Silanteva 1959, 42 fig. 18 top. For East Greek dot rosettes see Paspalas 1999, 91-2. 87 Ýren 2003, 112 fig. 54a-b. 88 Blegen et al. 1958, 269 Sherd VIII.186 pl. 295 no. 3. See, too, the stemmed dish Akimova 2005, 34 no. 27. Note, though, the dish Ýren 2003, 182 no. 246 (and p. 91), pl. 51 (‘Aiolischer Tierfriesstil IIIb’) the underside of which is painted in a loose manner, but the scheme of the bowl was executed in a more orderly fashion. Clearly, the possibility exists that the degree of care with which a vessel was painted does not necessarily determine its position in the wider sequence. 89 E.g. CVA Great Britain 9 Oxford 2 IID pl. 4.19 (1912.36.6). Selinous: Dehl-von Kaenel 1995, pl. 65.3570. Corinth: Williams et al. 1974, 21 no. 23, pl. 4. 90 Tocra: Boardman and Hayes 1966, 49 no. 604, pl. 31. Cyrene: Schaus 1985a, 55 no. 296, pl. 17. Klazomenai: Tzannes 2004, 109, 112-3 no. 7 fig. 22.2 (with references). 91 Buruncuk: Boehlau and Schefold 1942, 82, pl. 35.2 (‘Stufe V’) and 153 pl. 50.14 (‘Streifenware’). Tocra: Boardman and Hayes 1966, 44 and 52-3 nos 681-713, pl. 37 (mostly from Deposit II c. 600/590–565 bc); Boardman and Hayes 1973, 19 nos 1998-2001, pl. 11, and note, too, the bowl ibid. 20 no. 2008, pl. 11. Tell Sukas: Ploug 1973, 69 nos 306-7, pl. 15. Cyrene: Schaus 1985a, 68 nos 407-9, pl. 24, c. 580–560 bc. 92 Sidorova 1962, 147 fig. 21.5 (the rim is also decorated with a lotus flower). Myrmekeion: Butyagin 2001, 193-4. 93 Though note the dotted cross in added white on a dish from Naukratis: n. 77 supra. 94 Hdt. 2.178. Möller 2000a, 167-74. 95 Temenos of the Dioskouroi: Fairbanks 1928, 108 no. 319.8 (Boston MFA 88.1085). Temenos of Hera: ibid. 112 and 114 nos 323.2 (Boston MFA 88.937 [81]), 323.4 (Boston MFA 88.975 [79]) and 324.7 (Boston MFA 86.621 [68]). Temenos of Apollo: ibid. 112 no. 323.1 (Boston MFA 86.633 [83]). Temenos of Aphrodite: Gardner 1888, 44 Type F.a.2. For surveys of these temene see Möller 2000a, 94-104. 96 Though plainer examples are by no means unknown in East Greek sanctuaries, for example Held 2000, 100 nos K18-K19, fig. 55. 97 I have not found among Cook’s drawings and photographs illustrations of his nos II, IJ and 2, so, by necessity, these pieces have not been discussed in the main text of this paper. His no.20 also is not considered as it has already been published: Cook 1958/9, pl.6e. Note that (…) indicates a gap in the Ms.

Chemical Provenance Determination of Pottery: The Example of the Aiolian Pottery Group G Hans Mommsen and Michael Kerschner Abstract The important facts concerning chemical provenance determinations of archaeological pottery are summarized, explaining the Neutron Activation Analysis (NAA) and the statistical data evaluation procedure applied in Bonn. As examples of a successful determination, the chemical pattern for workshops at Kyme and/or its vicinity, called G, and its subgroup called g, are presented and discussed.* One of the most successful activities of the Bonn laboratory for archaeological science concerns the chemical provenance determination of pottery: that is the determination of the place where it was produced. Two main complementary methods, which look at different information stored in pottery, are known to yield useful results: petrography and elemental analysis. In the usual petrographic method,1 thin sections of pottery are inspected and give information about the minerals in the clay paste, whether naturally present or added. Alternatively, the clay paste itself can be characterised by an analysis of the minor and trace elemental content.2 The firing procedure does not change the relevant clay paste composition, except sometimes for the volatile elements As and Br.3 Also the burial conditions in most cases leave the sherd composition unchanged, again except for some elements such as Ba, Ca and sometimes the alkali elements (Na, K, Rb and Cs) and P.4 Since clays have generally much higher concentrations of trace elements compared to the admixtures like quartz or calcite, an elemental analysis characterises mainly the clay and, therefore, the measured elemental patterns of pottery point to the location of the clay beds exploited – if the assumption is correct that raw clay has not itself been traded. Ethnoarchaeological studies of modern Mediterranean potters working in a traditional way show that most of them are using clay beds in their vicinity within a radius of only a few kilometres.5 In many cases, the workshops have been built close to the clay beds. ‘Data of a worldwide sample of resource distances have demonstrated that potters travel no more than 7km to obtain their raw material.’ 6 Only in rare cases, when a raw material with special properties is required, can clays be transported over a longer distance.7 Nowadays, in discussions especially with colleagues in the field of archaeological science, one can notice that the predominant doctrine reflects a certain reluctance to accept results if only one of these provenance methods is applied. Instead, integrated studies are favoured. The results are considered to be trustworthy only if the so-called integrated approach is applied, using both methods of provenancing: the chemical and the mineralogical.8 There seems to be a general feeling that each of the methods by itself is insufficient. But if only provenancing is the aim, and not technological questions, our experience is that provenancing by chemical analysis alone works very well, provided this method is applied correctly and

used to its full potential. The advantage of chemical analyses when compared to petrography is that the measured elemental concentrations are hard data, which can be evaluated according to certain rules. In petrography one depends on the ability of an expert to recognise and memorize specific features in thin sections. Neutron Activation Analysis method The analysis method used in Bonn is Neutron Activation Analysis (NAA). This old and well-proven method9 is very well suited to measuring elemental concentrations in pottery, since it is multi-elemental, has high sensitivity for trace elements and is able to produce precise results with measurement uncertainties of a few percent. Samples of 80mg are taken by a pointed sapphire (corundum) drill, usually at the back of the sherds or from the bottom of the whole vessels to be analysed, leaving no more damage than a shallow extraction hole of 10mm diameter and a depth of about 1mm. Alternatively, as done for most of the samples from the British Museum, a thinner drill of 3mm diameter can be chosen to take a sample from the broken edge of a sherd. The Bonn pottery standard is used, which is calibrated with the well-known Berkeley standard.10 The whole measurement procedure has already been described at length elsewhere.11 As an example of the result of a NAA measurement, the concentration data of sample Kyme 1 are given in Table 1, second column, including the measurement uncertainty for each elemental value in the third column and expressed also in percent in the fourth column. As mentioned above, general agreement prevails that this pattern characterises the clay paste the ancient potter prepared. This paste composition itself depends on the geochemical composition of the clay deposit or deposits exploited, if several clays have been mixed,12 and, secondly, on the clay refinement techniques the ancient potters applied, for example levigating the clays or adding tempering material. Now, in provenancing, the assumption is made that all wares having the same composition will have been made from the same, well-homogenized clay paste prepared according to a certain recipe and will belong to a certain production series defined in this way by its characteristic clay paste. The continued or repeated appearance of a pattern across centuries at some sites suggests that at times local clays of identical composition and ready for use without much processing could be in use for long time spans. Furthermore, the elemental pattern of such a production series is likely to be unique in the world. This assumption, especially the uniqueness, will hold well, if a) many elemental concentrations – at least 20, and the more, the better – are measured, and b) the measurement precision, including trace elements, is high. For these reasons NAA was chosen in Bonn.

Naukratis: Greek Diversity in Egypt | 105

Mommsen and Kerschner Table 1 Concentrations of elements measured by NAA in µg/g (ppm), if not indicated otherwise a) sample Kyme 1, concentrations C and measuring errors d, also in % of C, with fit (dilution) factor 1.19 with respect to group g; b) averages M and spreads s, also in % of M, of the groups G and g. For the groups, each individual sample was dilution corrected with respect to the average values M of the group.

Kyme 1 1 samples factor 1.19 C +/- d % As Ba Ca% Ce Co Cr Cs Eu Fe% Ga Hf K% La Lu Na% Nd Ni Rb Sb Sc Sm Ta Tb Th Ti% U W Yb Zn Zr

53.5 757. 7.42 95.5 28.8 265. 27.6 1.60 5.37 -5.24 3.15 46.3 0.51 0.98 37.0 362. 196. 5.52 19.1 6.96 1.10 1.15 18.5 0.44 3.54 2.63 3.53 105. 192.

Group G 60 samples factor 1.00 M +/- s %

± ± ± ± ± ± ± ± ±

0.24 40.9 0.38 0.92 0.23 1.97 0.28 0.042 0.029

0.5 5.4 5.2 1.0 0.8 0.7 1.0 2.6 0.5

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.11 0.036 0.13 0.024 0.006 2.04 70.3 4.93 0.21 0.038 0.018 0.050 0.11 0.13 0.13 0.19 0.24 0.11 2.65 43.9

2.1 1.1 0.3 4.6 0.6 5.5 19. 2.5 3.9 0.2 0.3 4.6 9.4 0.7 30. 5.5 9.2 3.2 2.5 23.

44.7 810. 4.82 121. 27.6 188. 23.8 1.99 6.04 29.1 5.79 3.12 56.1 0.60 0.96 51.0 173. 178. 4.79 21.8 9.93 1.18 1.29 21.1 0.50 4.00 2.70 4.23 119. 173.

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

Pattern comparison To find samples of similar composition the elemental patterns, each consisting of about 30 concentration values, have to be compared. Comparison by hand is cumbersome and computeraided methods are used for this task. In such work, usually each sample is visualized as a point in concentration space. This space has one dimension for each measured concentration value. Samples of similar composition will fill the same region in this multidimensional space and form clusters of points at close quarters: the distance between two points in this space, therefore, can be used as a similarity measure. The usual methods like PCA (Principle Component Analysis) or different methods of CA (Cluster Analyses resulting in dendrograms) calculate these distances between all the data points, neglecting measuring errors. But since each concentration value has a different experimental measuring error (compare Table 1), distances should be calculated taking account of these errors. To give a simple example, two points in a one-dimensional space having the same distance (difference) may be considered as being statistically similar or dissimilar depending on the errors: for example 4 ± 0.1 and 5 ± 0.1 are not similar, but 4 ± 1 and 5 ± 1 are similar! Therefore, as the first improvement, a method was developed,13 which takes errors into account by normalising the distances to the error (distance 5 – 4 = 1; first case: 1/0.1 = 10 = not similar; second case: distance 1/1 = 1 = similar). In addition, a second effect during the comparison of patterns should be considered, since pottery is man-made. If potters diluted the clay by varying amounts, for example, of 106 | Naukratis: Greek Diversity in Egypt

Group g 28 samples factor 1.00 M +/- s % 20.3 98.6 0.83 3.54 1.06 12.1 2.80 0.049 0.21 3.75 0.56 0.12 1.11 0.023 0.12 2.41 45.4 6.48 0.58 0.38 0.51 0.061 0.083 0.56 0.11 0.59 0.29 0.12 6.69 56.9

45. 12. 17. 2.9 3.8 6.4 12. 2.5 3.5 13. 9.7 3.9 2.0 3.9 13. 4.7 26. 3.6 12. 1.8 5.1 5.2 6.5 2.6 22. 15. 11. 2.7 5.6 33.

45.2 739. 5.53 103. 26.8 211. 27.1 1.68 5.55 24.5 5.40 2.92 48.1 0.53 1.05 42.2 210. 172. 4.67 20.0 8.18 1.13 1.08 18.9 0.50 3.44 2.69 3.69 107. 154.

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

29.3 87.4 0.84 5.17 1.71 23.7 2.98 0.093 0.24 6.11 0.51 0.14 1.75 0.024 0.35 2.82 45.1 10.3 0.83 1.01 0.67 0.067 0.068 1.09 0.14 0.32 0.26 0.19 9.89 59.8

65. 12. 15. 5.0 6.4 11. 11. 5.5 4.3 25. 9.5 4.9 3.6 4.6 33. 6.7 22. 6.0 18. 5.1 8.1 5.9 6.3 5.8 28. 9.3 9.7 5.1 9.3 39.

sand, all concentration values will be lowered by a constant ‘dilution factor’. To correct for dilutions a best relative fit is done with regard to the centre value of two points or of a point and an already formed group of samples with similar composition. This mainly reduces the spreads (root mean square deviations) of the sum pattern formed. It often reduces positive correlations due to such dilutions. This is demonstrated in Figure 1 for group G of 53 samples made in the region of Kyme, where the concentration values of the rare earth elements Eu and Ce are shown before and after the dilution correction. To summarize, it is considered to be important during group formation to include the measurement errors and dilution effects to find ‘sharp’ concentration patterns of groups of samples, which were in fact produced with a certain clay paste at a pottery workshop, and thus avoid overlapping group patterns. Also, elements that are often part of the diluent like Ca or Na should be handled with care and at first not be considered during the search for groups. In publications, as the final result, the patterns formed should be given together with the spread values. Single, very large spread values point to either inhomogeneities of the clay paste or a wrong grouping. Example: reference patterns assigned to the region of Kyme The Bonn data bank now holds more than 1,100 samples from the East Aegean, including 30 pottery samples found at Kyme in Aiolis. This sample set was selected by M. Frasca (Catania) and M. Kerschner (Vienna) and is archaeologically discussed in a separate contribution in this volume.14 The chemical

Chemical Provenance Determination of Pottery: the Example of the Aiolian Pottery Group G

Figure 1 Plot of Eu and Ce concentrations of pottery group G, most probably made in Kyme, before and after the correction of dilutions by a best relative fit of the single samples with respect to the average concentrations of the group.The positive correlation coefficient is reduced from 0.95 to 0.60

classification according to the Bonn statistical procedure revealed that 21 samples from Kyme had a composition not very different from the already known group G, which at that time was formed of only seven samples, comprising sherds found at Ephesos, Klazomenai and Smyrna.15 There are three reasons why we now assign the provenance group G with high probability to production workshops at Kyme itself or in its vicinity. There are, firstly, the conspicuous prevalence of this pattern within the whole set of samples from this site; secondly, the longevity of this group at the site, ranging from the Archaic to the Roman Imperial period; and thirdly, the great diversity of ceramic classes showing this element pattern, comprising painted fine-ware as well as Grey and banded wares and even a water pipe. The prevalence and longevity of group G also at the neighbouring small town of Larisa (12 samples) may indicate a collateral production there exploiting the same clay beds. Exports from these workshops have been excavated also at Klazomenai (1 sample) and Smyrna (11 samples), Phokaia (15 samples), at Sardeis (2 samples) and even overseas at Naukratis (4 samples) and at Berezan on the northern Black Sea shore (12 samples).16 The now much larger number of members of this group G – altogether 88 samples (+ one repetition measurement) – permits a better definition of the average grouping values and their spreads. A small displacement of the group in concentration space with respect to the small old group G now results in a statistical separability of the two samples from Ephesos, a bird kotyle (Ephe 007) and a krater (Ephe 015), both from the Late Geometric period and both previously assigned to G.17 These two samples now form, with eight other sherds from Ephesos, a new provenance group named X.18 X can be localized with certainty at Ephesos because it comprises a miniature vessel (Ephe 029) that was part of the original fill of a Late Hellenistic potters’ kiln at this site.19 Exports of the Ephesian provenance group X have been detected at Larisa in Aiolis (Laris 12, only chemically associated to X), Thebes in Boeotia20 and at Tell Kazel in Syria.21 This demonstrates the importance of precise measurements and exact group forming procedures. After a closer and more elaborate inspection of the NAA data of group G the presence of a not very different, but statistically

separable subgroup in G, named ‘g’ and including 28 samples, was detected. The average concentration values M of the two groups G with the remaining 60 samples and g with 28 samples and their spreads (1s) are listed in Table 1. As is usually encountered, spread values of As, Ca and Na, although measured with small errors, are quite large and point to an inhomogeneity of these elements in pottery. Other elements with large spread values (Ga, Ni, Ti, Zr) are measured with large experimental errors. In Fig. 2 the normalized differences G – g are plotted as a bar diagram after a best relative fit of subgroup g with respect to G. Group g has a dilution of about 10% compared

Figure 2 Graphical comparison of chemical compositions of the two very similar groups G and g given in Table 1. Plotted are the differences of the average concentration values G – g normalised by the averaged standard deviations (sG2 + sg2)1/2. The values of group g are multiplied first by the best relative fit factor with respect to group G of 1.10 (10% enlarged).

Naukratis: Greek Diversity in Egypt | 107

Mommsen and Kerschner to G and has higher Cr and lower rare earth element concentrations. Except for the Cr value, the concentrations of all other elements vary less than 2s. This close agreement in composition of G and g might be due to a slightly changed paste recipe at these workshops. There is also a strong archaeological argument in favour of the localization of both provenance groups G and g at the same site: both comprise an almost identical range of ceramic wares and classes.22 Furthermore, even a well-defined stylistic group like the ‘London Dinos group’ shows both element patterns G and g.23 Such an extremely homogeneous group must have been produced by a single workshop at a certain site.24 Both groups can be separated very well from all other groups in our data bank. This is demonstrated in Kerschner Figure 34, which shows the result of a discriminant analysis of the groups assignable to the different East Aegean production sites represented up to now in our data bank. We hope that the assignment of groups G and g to workshops at Kyme and/or its vicinity (with a possible branch at neighbouring Larisa), which is most likely for the archaeological reasons asserted above, can be proven in the future by additional reference material from the site. Notes * 1 2 3 4 5 6 7

The authors wish to thank the staff of the research reactor in Geesthacht for their technical support. E.g. Whitbread 1995. Perlman and Asaro 1969; Jones 1986; Mommsen 2001, 2004. Cogswell et al. 1996; Schwedt and Mommsen (forthcoming) and references therein. Schwedt et al. 2004 and references therein. Hampe and Winter 1962, 4, 26, 49; Hampe and Winter 1965, 4, 27, 33, 38, 44, 51, 62, 87, 103, 127, 131, 133, 137, 139, 143, 147, 149-52, 177; Psaropoulou 1996, 81, 97, 236, 264. Arnold 1985, 38–42; Arnold et al. 1991, 85. A famous example are potters of Siphnos who carried with them raw clay from their island, cf. Dugas 1912b, 103. But, as Dupont 1983, 38, has pointed out, this transport of raw clay is mainly confined to the manufacture of cooking pots for which special properties are required. (cf. e.g. Tite and Kilikoglou 2002; for a discussion of the

108 | Naukratis: Greek Diversity in Egypt

8 9 10 11 12

13 14 15 16 17 18 19 20 21 22 23

24

special properties of mortaria fabrics see Villing, this volume). J. Boardman in Jones 1986, 663 and Boardman 1986, 252-3, 258 n. 14, expressed the hypothesis that potters’ clay from the island of Chios might have been exported to Naukratis (for a sceptical view, see Williams, this volume). Such a hypothesis, however, cannot be scrutinized by archaeometric methods. Karageorghis 2000/1, abstract, 92 proposed transport of raw clay from the Argolid to Cyprus to explain archaeometric chemical results obtained for Mycenaean pictorial pottery. On the use of different raw clays in order to achieve different colours of slip: Psaropoulou 1996, 144. On transport of raw clay on economic grounds, which has become easier due to modern motor vehicles, cf. Hampe and Winter 1965, 57, 117, 169, 177. Tite 1999. E.g. supra n. 2. Perlman and Asaro 1969. Mommsen et al. 1991. Archaeometric evidence for clay mixing has been provided by Schwedt et al. 2004. Ethnoarchaeological field studies have shown that clay mixing is a frequent procedure among traditional potters in the Mediterranean: Hampe and Winter 1962, 88–9; Hampe and Winter 1965, 44, 106, 137–8, 143, 147, 150, 161, 169; Psaropoulou 1996, 26, 97, 147, 178, 236. Mommsen et al. 1988; Beier and Mommsen 1994a, b. Kerschner, this volume. Akurgal et al. 2002, 84–92, nos 18, 51, 73, 79, 80, 84, figs 40, 48, 50, 55, pls. 1, 3, 6, 8. Mommsen, Kerschner and Posamentir 2006; Posamentir and Solovyov 2006 ; Kerschner 2006. Generally on the pottery finds from Berezan: Posamentir, this volume. Akurgal et al. 2002, 85–6, 106 no. 57, pl. 4 (Ephe 015); 85, 98 no. 18, pl. 1 (Ephe 007). Badre et al. 2006, 17, 19, 36-7. S. Ladstätter, in Akurgal et al. 2002, 117-19, 115 no. 102, pl. 8. At that time, group X was unknown and therefore Ephe 029 was still a chemical single. Schwedt et al. 2005. Badre et al. 2006. Cf. Kerschner, this volume. Provenance group G: Smyr 06, 45, Akurgal et al. 2002, 109-10, no. 73, fig. 40 pl. 6; cf. also Posamentir and Solovyov 2006: Bere 106, 109, 110; Mommsen, Kerschner and Posamentir 2006. Provenance group g: Smyr 22, Akurgal et al. 2002, 112 no. 84, fig. 55, pl. 8; Bere 178, Kerschner 2006; Mommsen, Kerschner and Posamentir 2006. On this point, all archaeologists have agreed up to now, cf. for further references the contribution of Kerschner in this volume.

On the Provenance of Aiolian Pottery Michael Kerschner

Abstract Naukratis is a major finding place of Archaic Aiolian pottery outside the East Aegean. The exact provenance of the various wares and styles of pottery in Aiolis is still largely unknown and controversial. This paper uses an archaeometric approach to attempt to localise the production places and to establish their repertoire. Based on NA Analysis of samples mainly from Kyme and Larisa, a chemical provenance group G, including a subgroup g, has been detected, comprising different styles of painted pottery as well as Grey and banded wares. This provenance group G/g represents an important and prolific pottery centre that dominates the production and even more the export of Archaic Aiolian pottery. It was most likely situated at Kyme. Neighbouring Larisa, however, may possibly have played a role, too.* The Aiolians and Naukratis Aiolis is the most northerly of the East Aegean regions (Fig. 1) that had been settled by immigrants from mainland Greece at the beginning of the Iron Age.1 The newcomers took possession of the island of Lesbos and of the opposite coast around the gulf of Elaia. Some of them penetrated into the mountains behind the fertile, but narrow coastal plain. To the south, the Aiolians soon lost the territory of Phokaia and Smyrna to the Ionians. To the north, however, they conquered the coast of the Troad and the off shore island Tenedos, presumably in the 8th century bc.2 The Aiolians spoke their own dialect. Their arts and crafts have much in common with the products of their Ionian neighbours, but at the same time they show peculiarities, as can be demonstrated in the case of the pottery. The Aiolians founded only few apoikiai during the great colonisation movement of the late 8th to 6th century bc. Yet they took part in the emporion of Naukratis in Egypt, which was mainly an enterprise by East Greeks. Herodotus provides some information on the organisation of the emporion during the reign of Pharaoh Amasis. He says that Ai0ole/wn de\ h9 Mutilhnai/wn mou/nh ‘among the Aiolians only the Mytileneaens’ (2.178.2) held a share in the sacred precinct of the Hellenion and also in the administration of the harbour. This, however, does not mean that other Aiolian poleis were not involved in the trade with Egypt via Naukratis, either directly with their own merchants acting or indirectly with their products. With regard to the latter, pottery offers a good possibility to trace the exact provenance of the exported goods. The history of research on Aiolian pottery The British excavations at Naukratis carried out by W.F. Petrie and E. Gardner in 1884–86 and by D.G. Hogarth some 15 years later stimulated the beginnings of research on East Greek pottery.3 Naukratis was, together with Rhodes, the first site where Aiolian pottery of the Archaic period was found (Figs 2–11). Publishing a selection of East Greek wares from the site in

1924, E.R. Price was the first to recognize a stylistically coherent group of Orientalising pottery that was later christened the ‘London Dinos group’.4 She argued that ‘they are all, if not the work of one hand, at least the output of one workshop’.5 In the following decades, many of the studies on Aiolian pottery focused on this single Wild Goat style group, which is superior to the other Aiolian painted wares in quality of execution and more widely distributed. E. Homann-Wedeking accepted Price’s grouping, emphasising the homogenous style of the filling ornaments.6 W. Schiering and N. Sidorova added some fragments to the group, which they supposed to be the work of one single painter, whom Schiering called ‘Dinosmaler der Vlastosgruppe’,7 but Sidorova the “master Londonskogo dinosa’, thus creating the present name of the group.8 Ch. Kardara compiled a comprehensive list of this group, considering the vessels as products of a workshop on Rhodes, which she called the ‘ergasterion dinou’.9 K. Schefold was the first to realize the Aiolian origin, comparing the dinos in Basle with the Orientalising pottery from Larisa.10 E. Walter-Karydi followed his localisation of this group, which she named after the dinos in Basle, and compared its style with other examples of Aiolian vase-painting.11 L.V. Kopeikina published a number of fragments of the ‘London Dinos group’ from Berezan pointing out that this Milesian colony on the northern Black Sea shore was, together with Naukratis, the main finding place outside the East Aegean.12 Publishing an early example from Pyrrha on Lesbos, W. Schiering assented finally to the Aiolian provenance of the group,13 and R.M. Cook, by using the name ‘London Dinos group’ in his handbook on East Greek pottery, firmly established it as the canonical name for the group.14 Recently, K. Ýren enlarged the list of pertinent vessels and fragments considerably, mostly with finds from Phokaia and Pitane, and went more fully into details of style.15 In addition to stylistic studies dealing exclusively with the Wild Goat style, excavations have provided an insight into a broader range of ceramic classes produced by Aiolian potters. In Aiolis, however, digging started later and the explored sites have remained fewer than in other parts of the Aegean, including adjacent Ionia. At most of the sites, the excavations of early Greek levels have been limited to small areas and therefore the stratigraphic evidence was meagre. There are only two exceptions, where excavations were carried out on a large scale: Larisa and Pitane. The extensive excavations carried out by L. Kjellberg and J. Boehlau 1902 and again 1932/4 at the site of Buruncuk, conventionally identified with ancient Larisa,16 proved particularly prolific in Archaic pottery (Figs 12–14; 16–19; 21–24), yielding a great variety of painted and Grey wares. In 1942, K. Schefold published the ceramic finds in a comprehensive analytical monograph, which is still the cornerstone of Aiolian pottery studies.17 The Larisa material comprises a much larger range of typological and stylistic groups Naukratis: Greek Diversity in Egypt | 109

Kerschner

Figure 1 Map of Aiolis

than those known before from Naukratis or Rhodes. Evidently, a fair number of ceramic classes were never exported outside Aiolis and its vicinity. At Pitane a large necropolis of the Archaic period was excavated by E. Akurgal from 1959 to 1965.18 The graves contained a considerable number of painted Aiolian vessels that were only recently published by K. Ýren.19 Some grave contexts with Corinthian imports provided useful indications for absolute chronology. Ýren devoted a meticulous study to the stylistic development not only of the Wild Goat style but also of a simpler variety of Aiolian vase-painting christened ‘Punktstil’ (‘dot style’) by him (Figs 13–14).20 Less rich in painted vessels, but important for the chronology of Aiolian pottery, is the necropolis of Assos, excavated by R. Stupperich 1989-94.21 A smaller number of painted vessels were found in graves at Myrina and Gryneion.22 Until recently, archaeological evidence for the early history of 110 | Naukratis: Greek Diversity in Egypt

Kyme was very scanty,23 even though this polis enjoyed, according to Strabo, a reputation as ‘the largest and best of the Aiolian cities’.24 From 1988 onwards, M. Frasca succeeded in revealing a stratigraphic sequence going back to the Late Geometric period in a trench on the southern of the two hills of the city.25 On the island of Lesbos, Geometric and Archaic pottery was unearthed at the ancient towns of Mytilene, Methymna, Antissa and Pyrrha.26 None of these sites has so far proved rich in painted vessels. ‘The ceramic repertoire in the island throughout the Archaic period continues to be dominated by Aiolic Grey wares.’27 To sum up: great progress has been made in studies on the stylistic development and, to a lesser extent, on the chronology of Aiolian pottery. The question of the exact provenance of the different categories of Aiolian pottery, however, has remained unresolved.

On the Provenance of Aiolian Pottery The unsolved question of the exact provenance Interpreting quantitative concentrations and distribution patterns of finds was for a long time the only feasible method to identify the origin of a certain typological or stylistic group of pottery.28 Both methods are important, but neither is definitive. Some ceramic classes were produced mainly for export and therefore may occur in extremely high quantities at other places. An indicative example is bird bowls of the standard fabric: they were made in the ‘bird bowl workshops’ of provenance group B on the North Ionian mainland, and thousands of them were unearthed at Ephesos and Miletos to where they had been exported.29 The reconstruction of distribution patterns depends greatly on the state of research in the region in question. If only a few sites have been investigated while major historical centres remain unexplored, the risk of error is considerable.30 That is the case in Aiolis, where only two minor sites have been excavated extensively – Larisa and Pitane – while we have only scarce and patchy information on the important poleis of Kyme, Myrina or Mytilene during the Geometric and Archaic periods. This want of appropriate archaeological evidence is the reason why general historical considerations have played a decisive role in the discussion about the location of Aiolian pottery centres. Vague conclusions drawn from notes by ancient authors on the economic and political roles of different cities encouraged the idea that the North Ionian polis of Phokaia could have been the leading centre of Aiolian art.31 R.M. Cook commented critically on this view which still prevails: ‘In what place or places this Aeolian pottery was made is not yet known; ... and claims for Phocaea are based mainly on its having been Ionian and therefore progressive’.32 P. Dupont was the first to raise a substantial objection to the Phokaian claim, basing himself on scientific analyses of clays and sherds.33 He introduced, at the same time as J. Boardman,34 archaeometric methods into East Greek pottery studies. An archaeometric approach to detect the pottery centres of Aiolis Following in their wake, H. Mommsen and I started an archaeometric programme in 1991. Our aim was to locate the pottery centres of the East Aegean and to investigate their repertoire of ceramic classes. In the course of time, a number of colleagues joined this project, contributing East Greek pottery from all over the Aegean and from colonial sites from Naukratis in the South to Berezan in the North and Sicily in the West.35 Each of the participants is working on one or more production sites or finding places. At the same time, all are networked through the Bonn data bank, thus ensuring the comparability of the chemical analyses and providing a platform for the discussion of the results and their interpretation. From 1997 onwards H. Mommsen and I have focused our research on Aiolis, in close co-operation with M. Akurgal (Ýzmir) and M. Frasca (Catania) as well as with the kind support of the curators of the university museums at Bonn (W. Geominy) and Göttingen (D. Graepler). Up to now, we have analysed 30 finds from Kyme and 27 from Larisa in Aiolis as well as 119 from the neighbouring North Ionian sites of Phokaia and Smyrna. The chemical provenance group G and its subgroup g The NAA of pottery finds from the Aiolian sites of Kyme and Larisa shows a striking predominance of the chemical

provenance group G, including its subgroup g (cf. Appendix 1). The differences in the concentrations of the single elements between G and g are only small, as H. Mommsen detected.36 This close agreement in composition points to a common origin rather than to two distant production sites. The slight variations in the element pattern might be due to minor inhomogeneities within the clay bed or / and to different recipes used by the potters in preparing the paste. In addition, there are archaeological arguments that suggest a common origin for G and g. Both element patterns cover a long span of time (cf. Appendix 1): G can be traced at Kyme and Larisa from the Subgeometric to the Hellenistic period; g even from Late Bronze Age until Roman Imperial times. Furthermore, both G and g comprise essentially the same range of wares and even stylistic groups during the Subgeometric and Archaic periods (cf. Appendix 1). There are a few more categories represented in G, a fact which may be explained by the circumstance that our data bank includes at the moment 59 samples of provenance group G, but only 25 of its subgroup g. The parallel occurrence of both element patterns at Kyme and Larisa over a long duration and, even more, their almost identical repertoire support the interpretation of the chemical data as a main provenance group (G) and a subgroup (g), both originating from the same pottery centre, rather than two distant production places. Ceramic wares and stylistic groups of provenance group G/g It is a communis opinio that ‘unlike all other Greeks, the eastern Aiolians ... did not make any painted Geometric pottery’ during the Geometric period, but started with a Subgeometric style early in the 7th century bc.37 Before that date, they appear to have been satisfied with their traditional Grey ware. The few finds of Late Geometric vessels in Aiolis were generally regarded as imports.38 Our NAA so far comprises three examples of Late Geometric pottery found at Aiolian sites, none of them belonging to the provenance group G/g. The element pattern of a krater found at Kyme (Kyme 08) with crosshatched triangles, a hatched meander and a chequerboard painted on a greyish brown surface matches no provenance group hitherto known from the East Aegean.39 The same applies to a krater or dinos from Larisa in Göttingen (Lari 12, Fig. 12) with vertical zigzags and a crosshatched lozenge chain framed by multiple lines.40 The third sample, a bird kotyle of standard fabric, shows the element pattern B of the North Ionian ‘bird bowl workshops’.41 The subsequent development of Aiolian vase painting in the Archaic period shows two very different styles practised simultaneously, as Ýren found out from the grave contexts at Pitane and Gryneion.42 On the one hand, there is an Aiolian variant of the East Greek Wild Goat style and on the other hand a simpler, schematic style, which Schefold called ‘Subgeometric’,43 whereas Ýren created the name ‘dot style’ (‘Punktstil’), as a result of the predilection for dots as framing lines and for the filling of ornaments.44 All three analysed examples of the latter style belong to the provenance group G: a dinos from Larisa (Lari 15, Fig. 13)45 with a metope frieze containing the typical water birds of East Greek Geometric tradition alternating with herringbone pattern and crosshatched fields, a pyxis from Larisa (Lari 16, Fig. 14),46 and a stand from Phokaia (Phok 27).47 The artistic quality of the ‘dot style’ lies in its spontaneous and dynamic brushwork as well as its gay and Naukratis: Greek Diversity in Egypt | 111

Kerschner colourful overall effect,48 which compensate for the evident carelessness in the finish. Ýren underlines the ‘provincial character’49 of these vases. This is the reason why he assumes that they were ‘produced in the small Aiolian towns of the hinterland’ in the poorer mountainous part of Aiolis.50 Our archaeometric analyses proved, however, that the ‘dot style’ was at home at the same place as the most elaborate version of the Aiolian Wild Goat style, the above-mentioned London Dinos group (Figs 2–8, 31). These dinoi can be grouped by stylistic criteria around the name piece from Kameiros in the British Museum (Fig. 2). Their decoration is characterized by competent drawing of the figures, a vivid, bright colouring and a peculiar choice of angular and voluminous filling ornaments. Trademarks of the London Dinos group are the cross with inserted chevrons and the doubly outlined band of tongues with dotted peaks in between. There are different ways to decorate the rim: either with a broad cable, also adorned with dotted peaks, with a lozenge net, with a single line meander, or with a meander hatched at right angles and framing broad hooks in purple like on a fragment from Naukratis (Nauk 13, Fig. 8).51 The extensive use of purple is typical for the London Dinos group. The finest examples show narrative scenes including human figures, as on the fragment from Phokaia that has been interpreted by E. Akurgal as the judgment of Paris.52 The homogeneity in style indicates that the London Dinos group was made in one single workshop. This observation is confirmed by our NAA. All examples analysed so far turn out to be members of the provenance group G/g (Smyr 06,53 Smyr 22,54 Phok 29,55 Nauk 13 – Fig. 8, Bere 178 – Fig. 1556 – as well as three further examples from Berezan in St. Petersburg analysed by R. Posamentir, Bere 106 – cf. Posamentir Fig. 13).57 Beside the London Dinos group, there are other varieties of Aiolian Wild Goat style, and again the majority belongs to the provenance group G/g, including the skyphos kraters Lari 18 58 (Fig. 16), Lari 1959 (Fig. 17), Lari 20 60 (Fig. 18) and the oinochoe Lari 21 (Fig. 19),61 all found at Larisa. The animals are stylised, displaying the typical Aiolian tendency towards the abstract. The filling ornaments tend to be even more enlarged and sometimes the figures seem nearly to be swallowed by the tapestry of ornaments; even the belly of a wild goat can be filled with a meander (Fig. 18). The fragment of a skyphos krater (Fig. 17) from Larisa demonstrates how broad the stylistic range of figure drawing is within the provenance group G/g. An unnaturalistic idea of figures and a clumsy execution, however, are not a matter of the place of production, or at least not alone. The provenance group G/g also comprises a large number of Aiolian Orientalising vessels that are decorated only with ornaments. The motives are borrowed from the repertoire of the Wild Goat style, but enlarged and emphasized by putting them into places normally reserved for figures, like the metopes on the kotyle from Smyrna (Smyr 46, Fig. 20).62 In the 6th century bc, vase-painters of the provenance group G/g adopted the black-figure technique without renouncing the traditional yellowish slip. The birds on the skyphos-krater Lari 23 (Fig. 21) are executed in pure black-figure technique, admitting even incised blobs as filling ornaments.63 The shape of the vessel has a long tradition at Larisa. The black-figure frieze of ducks on the krater Lari 22 (Fig. 22) is painted in a bichrome technique, using a brilliant orange-red shade as the main colour and a dull dark brown for details in addition to those indicated by incisions.64 112 | Naukratis: Greek Diversity in Egypt

The predilection for the vivid contrast of strong colours is typical for the painters of the provenance group G/g. Another method to create a colourful effect was painting with white and purple directly over the dark glazed surface of the vessels, sometimes also using incision.65 The main group decorated in this technique are black-polychrome oinochoai with incised tongues on the shoulder and purple and white bands on the body (Fig. 9).66 These Aiolian oinochoai were exported to Naukratis (Nauk 12, Fig. 9)67 and Berezan,68 as has been shown by NAA. An exceptional example from Pitane combines the decorative scheme of these oinochoai with an animal frieze on the belly that is stylistically related to the London Dinos group (cf. Appendix 2).69 A group of oinochoai from Vroulia, possibly made locally on Rhodes, is akin to these Aiolian vessels.70 The main difference is the decoration of the neck, which is reserved on the Rhodian pieces, sometimes with a broken cable. The decorative scheme of the black-polychrome oinochoai (Fig. 9) seems to have been inspired by Corinthian examples of the Late Protocorinthian and Transitional periods.71 Beside this class of oinochoai, there is a broad range of other shapes decorated in similar way by Aiolian vase-painters, like the dinos Lari 27 (Fig. 23).72 The hastily drawn meander on the rim and the dot-rosette on the shoulder are painted in added white colour on a dark reddish brown glaze. The light-on-dark technique in general is widespread, although not frequent in East Greek pottery.73 It has a long tradition, reaching back at least to the early 7th century bc.74 Famous examples are the Chian chalices with floral and occasionally even figurative decoration on the interior (see e.g. Williams Fig. 14)75 and the Vroulian class, principally found on the island of Rhodes and presumably made there.76 A less ambitious style of Aiolian vase-painting in the late Archaic period is purely ornamental without incision. Since the motives are formed from drop-shaped elements, I propose to call this variety ‘drop style’. The ornaments are simplified versions of vegetable pedigree like the single palmettes on a kotyle from Larisa (Laris 24, Fig. 24)77 and the wreath on another from Kyme (Kyme 14, Fig. 25).78 The potters of provenance group G/g also played a part in the production of the widespread classes of rosette bowls (Smyr 32, Fig. 26)79 and dishes with meander on the rim (Phok 49, Fig. 27).80 There is little surprise that provenance group G comprises also the ubiquitous ware decorated with simple bands and wavy lines (e.g. a banded dish of the Archaic period, Kyme 05, Fig. 29).81 Apart from painted wares, the pottery centre G/g was an important producer of Grey ware, which is so characteristic for Aiolis, especially in the Geometric and Archaic periods. Our analyses comprise different kinds of Grey ware: with grey surface or with grey slip, polished or unpolished (e.g. a carinated bowl from Kyme, Kyme 23, Fig. 30,82 and a fenestrated stand from Naukratis, Nauk 64, Fig. 10).83 Apart from ceramic vessels, the pottery centre G/g produced also architectural terracottas84 and terracotta figurines. The male head (Nauk 77, Fig. 11) found at Naukratis and attributed by NAA to the provenance group G/g is one of the few known Aiolian terracottas of the Archaic period.85 The shape of the beard with its rolling outline and the pointed moustache is typical for the middle and the third quarter of the 6th century bc.86 A strange feature is, however, the omission of the forehead. The face ends abruptly with the eyebrows, turning at right angle to the flat calvaria, on which strands of hair are painted. This

On the Provenance of Aiolian Pottery has to be regarded as a regional characteristic rather than a chronological one. A plastic vase in the shape of a mythological character with horse’s ears from Sardis has many features in common with the Aiolian head (Fig. 11): the broad face, the large almond-shaped eyes, the trimmed beard, and the painting of the details on a yellowish slip.87 Characterisation of the provenance group G/g Aiolian vase-painting of the Archaic period shows a broad artistic variety. Most scholars, like K. Schefold, H. Walter, E. Walter-Karydi, E. Akurgal, R.M. Cook and K. Ýren, explain this phenomenon by assuming several production places.88 They pursue, however, two different ideas. K. Schefold and R.M. Cook generally supposed that most of the pottery was made at its finding place. They regarded the vessels found at Larisa and Pitane as works of local potters and painters.89 This means that different stylistic groups originate from the same site. In contrast, K. Ýren, following ideas of H. Walter and E. WalterKarydi,90 divides the Archaic Aiolian vases according to their quality of design and execution. Consequently, he tries to attribute the elaborate groups to big and famous cities, the simpler varieties like the ‘dot style’ (‘Punktstil’) to smaller towns, preferably in the mountainous areas: Es ist anzunehmen, daß die Punktstilgefäße in den kleinen aiolischen Städten im Hinterland der Aiolis, wie Aigai, Neonteichos, Killa, Temnos, hergestellt wurden, während die aiolischen Städte an der Küste Tierfriesstilgefäße produzierten.91

Within the Aiolian Wild Goat style Ýren proposes a distribution according to the quality of vase-painting. On the one hand, he assumes that die aTs [= aiolischer Tierfriesstil] IIIb-Gefäße und die meisten der in Larisa gefundenen schwarzfigurigen Gefäße auch tatsächlich in Larisa hergestellt worden sind’, on the other hand: ‘Da die Gefäße des aTs II und IIIa sorgfältigere Stile als aTs IIIb haben, könnte man sich die großen und wichtigen aiolischen Städte wie Kyme und Myrina für sie als Herstellungsort vorstellen.

Furthermore, he considers the most ambitious variant of Aiolian Wild Goat style, the London Dinos group, as a product of a ‘phokäischen Keramikschule’,92 being ‘pseudo-Aiolian’ rather than ‘Aiolian’.93 However, K. Ýren cautiously emphasises that the proposed localisations are hypothetical and have yet to be verified by archaeometric analyses.94 Such analyses have now been carried out. They evince a different concept of the pottery centres in Aiolis. Our NAA show that provenance group G/g covers a diversity of stylistic groups originating from only one prolific pottery centre and probably also its vicinity. This means that potters and vase-painters working in different traditions were living together in the same town or nearby without absorbing the characteristic style of the others. Their products seem to have been offered for sale to the same customers, for vases painted in diverse Aiolian styles were found together in the same graves in the necropolis of Pitane.95 Social differentiation within the grave goods was evidently not displayed by means of the style or technical quality of a vessel. There was, however, a clear differentiation with respect to the export. Only a small part of the output of the pottery centre G/g was distributed outside Aiolis. The most widespread class of figured Aiolian vase-painting is the London Dinos group: it is also the most elaborate (Figs 2–8, 15, 31, Appendix 2). These splendid dinoi must have been highly appreciated by customers

who esteemed a certain accuracy of drawing, but at the same time preferred a more vivid colouring than that generally offered by the Ionian Wild Goat style. Apart from these luxury vessels, the pottery centre G/g had a stake in the mass-produced rosette bowls (Fig. 26)96 and dishes with meander on the rim (Fig. 27).97 Both were exported widely around the Mediterranean and Black Sea. These two classes of pottery were produced in several centres, mainly in North Ionia (provenance groups B and E) and, as far as meander-rim dishes are concerned, also at Ephesos (X)98 and presumably at Smyrna (F).99 Moreover, there is a class of late, standardised Wild Goat vases with a similar range of production places: the so-called ‘Borysthenes amphorae’ (cf. Posamentir Fig. 10).100 Those were produced mainly in North Ionia (provenance groups B and E), but also in the Aiolian pottery centre G/g, like a fragment from Berezan in Halle (Bere 174, Fig. 28).101 The ‘Borysthenes amphorae’ were exported far and wide, to the Black Sea (Berezan, Istros, Pantikapaion), the Levant (Tell Sukas) and Kyrenaika (Taucheira/Tocra).102 Such close interconnections between neighbouring pottery centres in North Ionia and Aiolis are best explained by the hypothesis of migrating potters.103 Such potters left, as individuals or in small groups, their home and settled permanently or for a certain time in a nearby polis. A number of possible reasons may have caused their movement, among them the search for economic advantages, for further education by other masters, or simply private reasons. Migration within the same region is a natural phenomenon in all periods. The immigrants brought along their own style, which they had learned at home, and thus spread it among their new colleagues. Flourishing in a new cultural environment, this implanted style was susceptible to new elements.104 Location of the provenance group G/g Where was the prolific pottery centre G/g situated that dominated the production of Archaic Aiolian fine ware? The distribution pattern of the finds analysed up until now shows a clear concentration on the mainland coast to the north and to the south of the Hermos estuary in the border region between Aiolis and North Ionia (Fig. 1). Nearly all East Aegean samples come from Kyme, Larisa, Phokaia and Smyrna.105 Thus, it is very likely that the home of provenance group G/g was in this very area. In order to assess the exact location, misfired vessels and undisturbed fills of pottery kilns are the best reference material to prove a local production. Unfortunately, they are rare in this region. Phokaia is the only site where a pottery workshop has been excavated. It dates from the Roman Imperial period (1st/2nd century ad).106 The NAA of 16 wasters from the dump of this workshop107 and two other misfired vessels discovered by E. Langlotz (Fig. 32)108 resulted in a number of different element patterns (T and Y as well as several chemical singles), which are all distinctly different from provenance group G/g (Fig. 33).109 This is a strong argument against a localisation at Phokaia,110 which had been proposed by E. Walter-Karydi and K. Ýren for the London Dinos group and related pottery belonging to provenance group G/g.111 The discriminant analysis of the 56 samples from Phokaia (Fig. 33) is revealing for the character of the site. The amount of imports is comparatively high. A number of different provenance groups are represented among the Archaic fine Naukratis: Greek Diversity in Egypt | 113

Figure 31 Finds of the London Dinos group in the Mediterranean and the Black Sea (cf.Appendix 2)

Kerschner

114 | Naukratis: Greek Diversity in Egypt

On the Provenance of Aiolian Pottery wares: Attic (KROP), the North Ionian provenance groups B and E, the Aiolian provenance group G/g as well as, at a smaller scale, Miletos (D) and Z (probably Lydian). It is important to underline that G/g is only one provenance group among several others represented at Phokaia. Among the remaining samples, there is no dominant chemical group which could account for an important local production of Archaic fine ware. We detected two small chemical groups (T, Y) among the Roman wasters and a number of singles, none of them going with any of the Archaic pieces. The chemical disparity of the wasters that are surely local products points to a complicated geology of the region around Phokaia with diverse clay beds and probably also clay mixtures. Ö. Özyiðit recently reported the discovery of an Archaic pottery workshop at Phokaia.112 The publication of the ceramic finds and their contexts, ideally supplemented by archaeometric investigations, will show, which classes of pottery were made on site in the Archaic period.113 The predominance of imports with the element pattern G/g among the Aiolian Orientalising pottery finds from Phokaia suggests, however, that a possible local share in it was not plentiful. With regard to the exports, all Aiolian pieces analysed so far from colonial sites (cf. Appendix 1) belong to the provenance group G/g that clearly differs in its element pattern from the local Phokaian wasters. Thus, a possible Phokaian share in the export of ceramic fine wares can have been only marginal. In any case, the most widespread class of Aiolian figured vase-painting, the London Dinos group (Appendix 2; Fig. 31), shows the element pattern G/g and, therefore, most likely was not made at Phokaia. At Kyme and Larisa, however, the situation is significantly different. At both sites the provenance group G/g is almost totally dominant in all wares and over a long span of time. Since no kiln wasters have been found at either site up until now, we have tried to analyse pots of different types and periods – from Subgeometric painted ware to a Roman water pipe (Appendix 1). Among the 30 samples from Kyme, five pieces that could typologically be identified as imports proved to be of Euboean, Corinthian and North Ionian origin (Fig. 34).114 Twenty-one out of the remaining 25 samples show one consistent element pattern – that of provenance group G/g. The remaining four samples (Kyme 08, 17, 26, 27) are chemical singles, the origin of which cannot yet be determined. This predominance of one single element pattern at the same site over a long period of time is a weighty argument in favour of a localisation of provenance group G/g at Kyme itself or in its vicinity.115 Otherwise, we would have to assume that the Kymeans imported all kinds of ceramic wares from elsewhere over several hundred years. At Larisa, the NAA show the same predominance of the provenance group G/g as at Kyme, comprising painted and Grey wares from the Late Bronze Age to the Late Archaic period. Larisa is, therefore, another possible candidate for the location of group G. At the moment we have only historical and geographical arguments on which to make a decision. Most of them speak in favour of Kyme, which was an important harbour city of considerable size and economic power. Strabo called it ‘the

largest and best of the Aiolian cities’.116 Larisa, on the other hand, was a small town and flourished only for a short period. Since both sites are situated only 12km apart (Fig. 1), it is also conceivable that they shared the same clay beds, or that the same geological layer extends into the territories of both poleis. If this is the case, the pottery of both Kyme and Larisa will show an identical element pattern, unless the potters prepared the raw material according to divergent recipes. Such a result contradicts the opinions of E. Walter-Karydi and K. Ýren on the place of origin of the London Dinos group, which they assume was Phokaia.117 Only K. Schefold considered Kyme as the most likely home of the workshop.118 The arguments that have been brought forward against the localisation of an important pottery workshop at Kyme have been mainly based on the alleged insignificance of the city in long distance trade119 and her reluctance to found colonies overseas. Nor was Kyme involved in the organisation of Naukratis, as Herodotus (2.178) indicates. In this kind of overseas venture, Phokaia was much more active, and that is the reason why some scholars presumed that Phokaia was also the main centre of arts and crafts in Aiolis.120 But is it inescapable to conclude that potters and merchants lived in the same city? There exist counter-examples demonstrating that neighbouring cities can create an economically efficient symbiosis, one focusing on the production, the other on the sale. There are strong indications that merchants of Aigina traded painted pottery from Athens and Corinth, as J. Boardman pointed out.121 Whereas Aiginetan potters did not produce painted pottery during the Archaic period, the seafarers of the island turned out to be most adventurous and successful merchants in overseas trade.122 In a similar way, although to a much lesser extent, Phokaians versed in long-distance trade may have dealt in the ceramic products of neighbouring Kyme, especially within their sphere of economic interest in the Western Mediterranean (cf. Fig. 31). Summary In the discussion on the provenance of Aiolian pottery, archaeometric analyses have proved to be an appropriate way out of the aporia caused by the lack of archaeological and literary evidence. This series of NAA should be seen as a beginning, which has to be extended by further investigations. So far the following results have been obtained: in Aiolis, one important and prolific pottery centre prevailed in the production and even more in the export of Archaic painted wares. Its ceramic products are defined by the chemical provenance group G and its subgroup g. The repertoire of the pottery centre G/g is impressive, comprising different styles and techniques of painted pottery as well as Grey and banded wares. The pottery workshops of provenance group G/g were situated most likely at Kyme. Neighbouring Larisa may possibly have had a share in G/g, too. Thus, the place of production of the splendid dinos in the British Museum (Fig. 2), the name-vase of the most famous group of Aiolian painted vases, can be considered revealed.

Naukratis: Greek Diversity in Egypt | 115

Kerschner

Appendix 1 Differentiation of the chemical provenance group G and its subgroup g According to ceramic wares Provenance group G

painted pottery of Aiolian Subgeometric style (Lari 15, associated with G, Fig. 13) painted pottery of Aiolian Orientalising style (Wild Goat style of the London Dinos group: Phok 29, 31; Smyr 06, Smyr 45; Nauk 13, Fig. 8; other groups of Wild Goat style: Lari 18, Fig. 16; Lari 20, Fig. 18; Lari 21, Fig. 19; Phok 28, 53, 56; with human figures: Lari 19, Fig. 17; ornamental: Kyme 19; Lari 17; Smy 46, Fig. 20) painted pottery of Aiolian ‘dot style’ (‘Punktstil’) (Lari 16, associated with G, Fig. 14; Phok 27) painted pottery of Aiolian bichrome style (black-figure technique: Lari 22, Fig. 22; ornamental: Smyr 60) painted pottery of Aiolian black-figure style (Lari 23, Fig. 21) painted pottery of Archaic Aiolian ‘drop style’ (Lari 24, Fig. 24; Lari 25, 26; Smyr 59, 62) painted pottery of Archaic Aiolian light-on-dark style (Lari 27, Fig. 23) dishes with meander on the rim: (Phok 49, Fig. 27; Klaz 01)123 rosette bowls (Phok 15 ?; Smyr 32, Fig. 26) black glazed oinochoai with added red and white (Bere 107, 108, Posamentir Fig. 1 top left; Nauk 12, Fig. 9)124 banded ware of the Archaic period (Kyme 05, Fig. 29; Kyme 18, 28; Phok 07, 50, 51; Bere 138, Posamentir Fig. 14) Grey ware of the Archaic period (with grey surface: Kyme 21, 30 and dark grey slip: Kyme 22, 23, Fig. 30; indeterminate: Phok 10) undecorated ware of Hellenistic period (Kyme 03)

Sub-group g

Red Wash ware of the Late Bronze Age (Lari 10) painted pottery of Aiolian Orientalising style (Wild Goat style of the London Dinos group: Smyr 22; Bere 178,125 Fig. 15; other groups of Wild Goat style: ornamental Kyme 12, Phok 54) painted pottery of Aiolian Orientalising style (Wild Goat style of the ‘Borysthenes amphorae’: Bere 174, Fig. 28)126 painted pottery of Aiolian Orientalising style (Wild Goat style combining reservation and Corinthianising black-figure techniques (Bere 001)127 painted pottery of the Archaic Aiolian ‘drop style’ (Kyme 14, Fig. 25; Smyr 47) black glazed oinochoai with added red and white (Bere 105, Posamentir Fig. 1 top right)128 banded ware of the Archaic period (Kyme 02, 06, 16, 20, 25, 29; Phok 21) Grey ware of the Archaic period, with grey surface (Kyme 13; Nauk 64, Fig. 10). painted pottery of Archaic Lydian marbled ware (Smyr 58) architectural terracottas of the Archaic period (Phok 32) terracotta figurine of the Archaic period (Nauk 77, Fig. 11) trade amphorae of the Hellenistic period (Kyme 01) lamps of the Hellenistic period (Kyme 24) water-pipes of the Roman period (Kyme 15)

According to find spots Finds from Aiolis

At Kyme, both subgroups are represented with a broad range of ceramic classes.129

Provenance group G

painted pottery of Aiolian Orientalising style (Kyme 19) banded ware of the Archaic period (Kyme 05, Fig. 29; Kyme 18, 28) Grey ware of the Archaic period (both with grey surface: Kyme 21, 30 and dark grey slip: Kyme 22, 23) undecorated ware of Hellenistic period (Kyme 03)

Subgroup g

painted pottery of Aiolian Orientalising style (Kyme 12) painted pottery of the Archaic Aiolian ‘drop style’ (Kyme 14, Fig. 25) banded ware of the Archaic period (Kyme 02, 06, 16, 20, 25, 29) Grey ware of the Archaic period (with grey surface: Kyme 13) trade amphora of the Hellenistic period (Kyme 01) lamp of the Hellenistic period (Kyme 24) water-pipe of the Roman period (Kyme 15) At Larisa, group G is much better represented than subgroup g. 116 | Naukratis: Greek Diversity in Egypt

Provenance group G

painted pottery of Aiolian Subgeometric style (Lari 15, Fig. 13) painted pottery of Aiolian Orientalising style (Wild Goat style: Lari 18, Fig. 16; Lari 20, Fig. 18; Lari 21, Fig. 19; with human figures: Lari 19, Fig. 17; ornamental: Lari 17) painted pottery of Aiolian ‘dot style’ (‘Punktstil’) (Lari 16, Fig. 14) painted pottery of Aiolian bichrome style in black-figure technique (Lari 22, Fig. 22) painted pottery of Aiolian black-figure style (Lari 23, Fig. 21) painted pottery of Archaic Aiolian ‘drop style’ (Lari 24, Fig. 24; Lari 25, 26) painted pottery of Archaic Aiolian light-on-dark style (Lari 27, Fig. 23)

Subgroup g

red wash ware of the Late Bronze Age (Lari 10) Nine further examples, mostly of Grey ware, were analysed by D. Hertel and H. Mommsen (Lari 01-09), six of them belonging to provenance group G, two to subroup g.

Finds from Phokaia130 Provenance group G

painted pottery of Aiolian Orientalising style (Wild Goat style of the London Dinos group: Phok 29, 31; other groups of Wild Goat style: Phok 28, 53, 56 painted pottery of Aiolian ‘dot style’ (‘Punktstil’) (Phok 27) dish with meander on the rim (Phok 49, Fig. 27) rosette bowl ? (Phok 15) banded ware of the Archaic period (Phok 07, 50, 51) Grey ware of the Archaic period (polished: Phok 10)

Subgroup g

painted pottery of Aiolian Orientalising style (ornamental: Phok 54) banded ware of the Archaic period (Phok 21) architectural terracottas of the Archaic period (Phok 32)

Finds from sites outside Aiolis Provenance group G

painted pottery of Aiolian Orientalising style (Wild Goat style of the London Dinos group: Smyr 06, Smyr 45; Nauk 13, Fig. 8; ornamental: Smyr 46 (Fig. 20)). painted pottery of Archaic Aiolian ‘drop style’ (Smyr 59, 62) painted pottery of Aiolian bichrome style (ornamental: Smyr 60) black glazed oinochoe with added red and white (Nauk 12, Fig. 9) dish with meander on the rim (Klaz 01) rosette bowl (Smyr 32, Fig. 26) Six further examples were found at Berezan (Bere 106, 107, 108, 109, 110, 138, Posamentir Figs 1 top left, 13, 14 top).131

Subgroup g

painted pottery of Aiolian Orientalising style (Wild Goat style of the London Dinos group: Smyr 22, Bere 178,132 Fig. 15) painted pottery of Aiolian Orientalising style (Wild Goat style combining reservation and Corinthianising black-figure techniques (Bere 001)133 painted pottery of Aiolian Orientalising style (Wild Goat style of the ‘Borysthenes amphorae’: Bere 174, Fig. 28)134 painted pottery of Archaic Aiolian ‘drop style’ (Smyr 47) painted pottery of Archaic Lydian style (marbled ware: Smyr 58) Grey ware of the Archaic period (Nauk 64, Fig. 10) terracotta figurine of the Archaic period (Nauk 77, Fig. 11) One further example found at Sardis (Sard 41) and three from Berezan (Bere 101, 105 [Posamentir Fig.1 top right], 126).135

On the Provenance of Aiolian Pottery

Appendix 2 Dinoi of the London Dinos group and stylistically related vessels This list is based on those compiled by Ch. Kardara, E. Walter-Karydi and K. Ýren,136 augmenting their compilations with further, partly unpublished pieces and discussing questionable attributions. Amathus 1. Dinos and stand, Nicosia, inv. 1966/X-29/1: Karageorghis 1961, 312, fig. 65; Dikaios 1961/62, 38, fig. 11; Nicolau 1967, 400, pl. 115.11; Walter-Karydi 1970, 3 no. 4; Gjerstad 1977, 34-5 no. 160, pl. 18.1; Thalmann 1977, 74, 77 no. 111 (dinos), no. 112 (stand), pl. 17.1-3; Stampolidis et al. 1998, 137 no. 93. Stylistically related vessels of different shape 1. Amphora: Thalmann 1977, 77, pl. 17.4; Ýren 2002, 198 no. 1. 2. Amphora: Thalmann 1977, 77, pl. 17.10; Ýren 2002, 198 no. 2. 3. Oinochoe or small amphora: Thalmann 1977, 77 no. 113, pl. 17, 5-8; Ýren 2002, 199 no. 7. (Ýren detached these two fragments from the shoulder and belly fragments, which had been attributed to the same vessel by Thalmann, and classified them as dinoi without having seen them. Thalmann, however, did not mention glaze on the inside of these two pieces, whereas he stated that the dinos ibid. 77 no. 111 was glazed on the interior.) Ashkelon Stylistically related vessels of uncertain shape 1. Wall fragments: Stager 1996, 67, 69, fig. 10 (the 2 fragments below at the right); Ýren 2002, 185-6, 199 no. 8, fig. 10a (Ýren classified the 2 fragments as parts of a dinos, but this is uncertain considering the smallness of the sherds). Assos Stylistically related vessels of uncertain shape 1. Wall fragment of a closed vessel:Utili 1999, 196, 307 no. 314, fig. 22 (the incline of the piece on the drawing seems too steep, since the tongues suggest that it belongs to the upper end of the body); Ýren 2002, 199 no. 9. 2. Wall fragment of closed vessel: Utili 1999, 199 no. 334, fig. 23 (the fragmentary ornament is not the ‘Rest einer Lotosblüte’, but a row of tongues; Ýren 2002, 199 no. 10. 3. Wall fragments of a closed vessel: Utili 1999, 198 no. 333a+b, fig. 23; Ýren 2002, 199 no. 11. Ýren classified the three fragments Assos nos 1-3 as parts of dinoi, although Utili considered them explicitly parts of closed vessels. Berezan 1. Rim and wall fragment of a dinos, Odessa, Arkheologicheskii Muzei inv. 34396, 36924: Kopeikina 1970b, 562, 565, pl. 2.1; Kopeikina 1982, 27, fig. 22; Ýren 2002, 182, 184, 199 no. 15, fig. 14. (These fragments have subsequently been joined). 2. Three wall sherds of a dinos: St. Petersburg, Hermitage inv. Á.68-27: Kopeikina 1970b, 563, 565, pl. 2.2; Kopeikina 1982, 27, fig. 23a (on this figure, only two fragments are shown); Ýren 2002, 182, 184, 199 no. 16. 3. Two large rim and wall fragments of a dinos: St. Petersburg, Hermitage Inv. Áåð. 75-7: Kopeikina 1981, 196, fig. 4a; Kopeikina 1982, 27, fig. 23 á; Ýren 2002, 199 no. 17; Posamentir and Solovyov 2006, fig. 2, sample no. 106, provenance group G. 4. Large wall fragment of a dinos, Odessa, Arkheologicheskii Muzei inv. A-34903+39640: Kopeikina 1982, 27 fig. 23 â; Ýren 2002, 200 no. 19. 5. Reconstructed dinos, Hermitage Inv. Á. 66-7: Kopeikina 1970b, 563, 565, pl. 1.4; Kopeikina 1982, 27, fig. 24; Ýren 2002, 185, 199 no. 14. 6. Wall fragment of a dinos, St. Petersburg, Hermitage Inv. B.91.233: Solovyov 1999, 48-9, fig. 29.2; Solovyov 2001, 126, fig. 6; Ýren 2002, 200 no. 22. 7. Wall fragment of a dinos, St. Petersburg, Hermitage Inv. B.82.8: Solovyov 1999, 48-9, fig. 29.1; Ýren 2002, 200 no. 20. 8. Shoulder fragment of a dinos, St. Petersburg, Hermitage Inv. B.88.20: Solovyov 1999, 49-50, fig. 32 (bottom right); Solovyov 2001, 126, fig. 8 (bottom right); Ýren 2002, 200 no. 21. 9. Shoulder fragment of a dinos, Halle, inv. 480 (Fig. 15), sample no. Bere 178 (provenance group g): Kerschner 2006 . 10. Wall fragment of a dinos: Kopeikina 1970b, 563, 565, pl. 1.3; Ýren 2002, 199 no. 18. 11.Wall fragment of a dinos, Kiev, Institut Arkheologii inv. Á.63-1003: Kopeikina 1970b, 563, 565, pl. 1.5; Ýren 2002, 185, 199 no. 13. 12. Reconstructed dinos in Odessa, Arkheologicheskii Muzei. 13. Wall fragment of a dinos, unpublished, excavation K. Marchenko 1999.137 14. Rim fragment of a dinos, St. Petersburg, Hermitage: Posamentir and

Solovyov 2006, fig. 3, sample no. 109, provenance group G. 15. Wall fragments of a dinos, St. Petersburg, Hermitage: Posamentir and Solovyov 2006, fig. 4, sample no. 110, provenance group G. 16. Several further fragments of dinoi were announced by Kopeikina 1970b, 565: ‘Auf Berezan wurden viele Dinos-Fragmente gefunden, die man der sogenannten Gruppe des ‘Londoner Dinos’ zuschreiben kann.’; cf. Kopeikina 1982, 27. Yet R. Posamentir (in this volume; Posamentir and Solovyov 2006) has shown that the many fragments stored in the Hermitage do not represent that many individual vessels, but belong to at least 5 dinoi. Stylistically related 1. Wall fragment of a dinos, Odessa, Arkheologicheskii Muzei inv. A36082: Kopeikina 1982, 25, 29, fig. 19b; Kopeikina 1986, 28-9, pl. 1.1a (top); Ýren 2002, 172-4, 199 no. 12. Gravisca 1. Rim and wall fragments of a dinos and a stand: Boitani Visentini 1978, 216-7 pl. 90.1; Boldrini 1994, 90-3 no. 157; Ýren 2002, 200 no. 23. Gryneion Stylistically related vessels of uncertain shape 1. Lekythos: Ýren 2002, 174-7, 205 no. 86, figs 7-8. Ikaria, Sanctuary of Artemis Tauropolis 1. Wall fragment of a dinos: Politis 1939, 132, fig. 9; Kardara 1963, 275; Walter-Karydi 1970, 3 no. 12; Ýren 2002, 200 no. 25. Istros 1. Wall fragment of a dinos: Lambrino 1938, 256-7 no. 13, fig. 222; Kardara 1963, 274; Walter-Karydi 1970, 3 no. 11; Ploug 1973, 52 n. 337; Alexandrescu 1978, 41 no. 53, pl. 5; Dupont 1983, 30, fig. 2; Ýren 2002, 200 no. 24. Kameiros (?) 1. Dinos. London, British Museum GR 1848.6-19.1 (Fig. 2): Kinch 1914, 192-3, 234, figs 73, 118e; Price 1924, 193-4, fig. 22 (‘Cameiros’); Schiering 1957, 14, pl. 9.1; Kardara 1963, 274; Walter-Karydi 1970, 3 no. 2 (‘unbekannter Herkunft’); Boardman 1970, 92-4, pl. 44.5; Kopeikina 1982, 27, fig. 21; Schiering 1981-3, 202, 205, 207-8, pl. A; Akurgal 1993, pl. 114; Ýren 2002, 166, 182, 184-5, 203-4 no. 68, figs 1, 18. Katane 1. Rim and wall fragments of a dinos (meander on the rim, meander hooks and tongues on the shoulder): Catania, Soprintendenza inv. KC 6300, unpublished. 2. Shoulder fragment (with tongues) of a dinos: Catania, Soprintendenza Inv. KC 6301, unpublished. Both Katane nos 1-2 were excavated by G. Rizza (cf. the preliminary report Rizza 1960) and are prepared for publication by A. Pautasso (Catania), whom I thank for the permission to mention these pieces. Both have been analysed in the meantime and yielded the element pattern G. Larisa on Hermos 1. Wall fragment of a dinos, Göttingen, Archäologische Sammlung inv. 7/91: Ýren 2002, 177, 200 no. 27, fig. 9b. Stylistically related vessels of uncertain shape 1. Benoit 1965, 228-9, 294-5, pls 4.2, 37.3; Ýren 2002, 200 no. 29. Stylistically related vessels of different shape 1. Fragment of an olpe: Boehlau and Schefold 1942, 69, fig. 18; Ýren 2002, 177, 206. 2. Fragment of an oinochoe: Boehlau and Schefold 1942, 69, pl. 21.17; Walter 1968, 77-8 (considered to be a krater); Walter-Karydi 1973, 4 (considered to be a krater); Schiering 1981-3, 210 n. 19; Kerschner 1997a, 23 n. 94; Ýren 2002, 200 no. 26 (Ýren listed this fragment among the dinoi, although Schefold described it as an oinochoe). 3. Fragment of a small oinochoe: Boehlau and Schefold 1942, 68, pl. 20.8; Ýren 2002, 200 no. 28 (Ýren classified this fragment as dinos, although Schefold published it as a small oinochoe). Málaga 1. Rim fragment of a dinos, Málaga Inv. 10073: Gran-Aymerich 1988, 209, fig. 9.1; Olmos 1989, 500, 502, 521, fig. 7; Domínguez and Sánchez 2001, 27, fig. 24.1; Ýren 2002, 200 no. 30. Massalia 1. Wall fragment of a dinos: Vasseur 1914, 28-9, pls 5.6-7; 6.1; Villard 1960, 39; Benoit 1965, 139-40, pl. 6.2; Walter-Karydi 1970, 4 no. 14; Ýren 2002, 200 no. 31. 2. Wall fragment of a dinos: M. Derain in Hesnard et al. 1999, 24 (top left). Stylistically related vessels of different shape 1. Rim and wall fragment of a cup with concave walls, excavation Rue de la Cathédrale, inv. 549: Musée d’Histoire de Marseille 1990, 16 (‘calice de Chios’); Gantès 1999, 369, 378, fig. 2.3 (‘canthare ou ... tasse à anse’). Megara Hyblaia 1. Shoulder fragment (with tongues) and wall fragments (not reproduced), inv. 4/184 – 4/186: Vallet and Villard 1964, 79, pl. 65.3 Naukratis: Greek Diversity in Egypt | 117

Kerschner (It seems doubtful if the rim fragments ibid. 79, pl. 65.1-2 belong to the same vessel, as suggested by the authors). 2. Wall fragment (with grazing wild goat), inv. 6/3: Vallet and Villard 1964, 85, pl. 72.4. 3. Shoulder fragment (with meander hooks and tongues), inv. 6/5, possibly from the same vessel as no. 2: Vallet and Villard 1964, 85, pl. 72.7. Mytilene 1. Wall fragment of a dinos, inv. T86/17 L12 P25: Schaus 1992, 359, 361 no. 6, pl. 80; Ýren 2002, 201 no. 33. 2. Wall fragment of a Dinos (?): Schefold 1933, 154, fig. 11; Kardara 1963, 275; Ýren 2002, 182, 184, 200 no. 32. Naukratis 1. Rim fragment of a dinos: Boston, Museum of Fine Arts Inv. 86.538 (77): Fairbanks 1928, 102 no. 307.2, pl. 30; Walter-Karydi 1973, 138 no. 696, pl. 98 (considered to be Chiot). 2. Wall fragment of a dinos, London, British Museum GR 1886.4-1.1270 (Fig. 3): Price 1924, 193, fig. 20; Homann-Wedeking 1938, 16 no. P1; Kardara 1963, 274; Walter-Karydi 1970, 3 no. 1; Ýren 2002, 201 no. 39. 3. Large wall fragment of a dinos, London, British Museum GR 1886.41.1288 (Fig. 4): Price 1924, 193, fig. 21; Homann-Wedeking 1938, 16 no. P3; Walter-Karydi 1970, 3 no. 3, pl. 8.2; Ýren 2002, 182, 185, 201 no. 40. 4. Wall fragment of an open vessel, presumably a dinos, Boston, Museum of Fine Arts inv. 86.527 (54): Fairbanks 1928, 110 no. 321.1, pl. 34; Kardara 1963, 275; Walter-Karydi 1970, 3 no. 5; Ýren 2002, 201 no. 35. 5. Wall fragment of a dinos (?), formerly Den Haag, Museum Scheurleer: Prins de Jong 1925, 46; Scheurleer 1931, II D, pl. 2.8; Walter-Karydi 1970, 3 no. 6; Ýren 2002, 201 no. 43. 6. Wall fragment of an open vessel, presumably a dinos, Oxford, Ashmolean Museum inv. C 119.110: E.R. Price in Beazley et al. 1931, II D, pl. 4.3; Kardara 1963, 275; Walter-Karydi 1970, 3 no. 7; Ýren 2002, 201 no. 44. 7. Wall fragment of an open vessel, presumably a dinos, Brussels, Musée du Cinquantenaire inv. A 1761: Mayence and Verhoogen 1949, II D, pl. 3.14; Kardara 1963, 274; Walter-Karydi 1970, 3 no. 8; Ýren 2002, 201 no. 41. 8. Small wall fragment, presumably of a dinos (?), Heidelberg, Universität inv. 39: Schauenburg 1954, 11 no. 26, pl. 2.26; Kardara 1963, 276; Walter-Karydi 1970, 3 no. 10; Ýren 2002, 201 no. 42. 9. Wall fragment of an open vessel, presumably a dinos, Boston, Museum of Fine Arts inv. 88.949 (55): Fairbanks 1928, 110 no. 321.3, pl. 34; Ýren 2002, 201 no. 37. 10. Wall fragment of an open vessel, presumably a dinos, Boston, Museum of Fine Arts inv. 88.949 (55): Fairbanks 1928, 111 no. 321.11, pl. 34; Kardara 1963, 275; Ýren 2002, 201 no. 38. 11. Wall fragment of a dinos?, London, British Museum GR 1888.6-1.470 (Fig. 5): Kardara 1963, 276. 12. Wall fragment of a dinos (?), Cambridge, Museum of Classical Archaeology NA 33: Kardara 1963, 276. 13. Wall fragment of a dinos (?), London, British Museum GR 1924.12-1.11 (Fig. 6): Kardara 1963, 276. 14. Rim fragment of a dinos, London, British Museum GR 1886.4-1.1294 (Fig. 8), sample no. Nauk 13 (provenance group G). Stylistically related vessels of different shape 1. Askos. London, British Museum GR 1888.6-1.462 (Fig. 7): Gardner 1888, 40, pl. 5.1; Price 1924, 193; Schiering 1957, 14, 27, pl. 13.4; Kardara 1963, 275 no. 1; Walter-Karydi 1970, 4 no. 18, pl. 3.6. 2. Wall fragment and handle of a krater (?), Boston, Museum of Fine Arts Inv. 88.949 (55): Fairbanks 1928, 110 no. 321.2, pl. 34; Ýren 2002, 201 no. 37 (Ýren classified the fragment as dinos, although it has a horizontal handle). Pantikapaion 1. Wall fragment of a dinos, inv. M45 CM IV/4, no. 2791 (or 4791 according to Tsvetaeva): Tsvetaeva 1957, 183-4, 186, fig. 2a.2 (‘delossko-melosskoi keramiki’); Sidorova 1962, 107-8, fig. 1.1 (‘master Londonskogo dinosa’). 2. Wall fragment of a dinos, inv. M52 BM XIV/14, No. 369: Sidorova 1962, 107-8, fig. 1.2 (‘dinos or krater’). Phokaia 1. Wall fragment, presumably of a dinos: Jacobsthal and Neuffer 1933, 14, fig. 6a; Schiering 1957, 14, 116 with n. 96; Walter 1968, 79, 128 no. 628, pl. 130 (‘Amphora ?’); Walter-Karydi 1970, 3, 6 no. 13, pl. 4.5 (‘Dinos’); Ýren 2002, 202 no. 58; Kerschner 2004, 138. 2. Wall fragment of a dinos (with remnants of an animal frieze and a narrative scene with human figures, interpreted as judgment of Paris by Akurgal): Akurgal 1961, 180, fig. 128; Walter-Karydi 1970, 6-7, 12-3, 18, pl. 8.3 (erroneously considered as a chalice, as the first published photo suggested a horizontal upper edge); Akurgal 1987, 25, pl. 3b; Akurgal 1993, pl. 103d; idem in Musée d’Histoire de Marseille 1995, 38; Ýren 2002, 186, 203 no. 59; Kerschner 2004, 138. 3. Wall fragments of a dinos (?) (with a frieze of dancing girls below a 118 | Naukratis: Greek Diversity in Egypt

meander and a running spiral): Akurgal 1961, 180, figs 129-30; Langlotz 1966, 27, figs 25, 27; Langlotz, 1969, 381; Walter-Karydi 1970, 7, pl. 8.5; Langlotz 1975, 197, pl. 63.3; Akurgal 1993, fig. 103a-c; idem in Musée d’Histoire de Marseille 1995, 38; Kerschner 2004, 138-9. E. Langlotz and E. Walter-Karydi erroneously considered these fragments to be parts of a chalice, judging from the first published photos suggesting a horizontal upper edge. 4. Wall fragment of a dinos, inv. Foça 1956 H çukuru Güney çýkma 650550: unpublished; sample no. Phok 28, provenance group G. 5. Rim fragment of a dinos, Ýzmir Arkeoloji Müzesi inv. Foça 1955: Ýren 2002, 201 no. 45, fig. 13a. 6. Rim fragment of a dinos, Ýzmir Arkeoloji Müzesi inv. Foça 1955: Ýren 2002, 201 no. 46, fig. 13b. 7. Rim fragment of a dinos, Ýzmir Arkeoloji Müzesi inv. Foça 1955: Ýren 2002, 202 no. 47, fig. 13c. 8. Rim fragments of a dinos, Ýzmir Arkeoloji Müzesi inv. Foça 1955: Ýren 2002, 202 no. 48, fig. 13d. 9. Shoulder fragments of a dinos, Ýzmir Arkeoloji Müzesi Inv. Foça 1955: Ýren 2002, 202 no. 49, fig. 13e. 10. Shoulder fragment of a dinos, Ýzmir Arkeoloji Müzesi inv. Foça 1955: Ýren 2002, 202 no. 50, fig. 13f. 11. Shoulder fragment of a dinos, Ýzmir Arkeoloji Müzesi inv. Foça 1955: Ýren 2002, 202 no. 51, fig. 13g. 12. Shoulder fragment of a dinos, Ýzmir Arkeoloji Müzesi inv. Foça 1955: Ýren 2002, 202 no. 52, fig. 13h. 13. Wall fragment of a dinos, Ýzmir Arkeoloji Müzesi inv. Foça 1955: Ýren 2002, 202 no. 53, fig. 13i. 14. Wall fragment of a dinos, Ýzmir Arkeoloji Müzesi inv. Foça 1955: Ýren 2002, 202 no. 54, fig. 13j. 15. Wall fragment of a dinos, Ýzmir Arkeoloji Müzesi inv. Foça 1955: Ýren 2002, 202 no. 55, fig. 10b. 16. Wall fragment of a dinos, Ýzmir Arkeoloji Müzesi inv. Foça 1955: Ýren 2002, 202 no. 56, fig. 15a. 17. Wall fragment of a dinos, Ýzmir Arkeoloji Müzesi inv. Foça 1955: Ýren 2002, 202 no. 57, fig. 13k. 18. Wall fragment of a dinos, Ýzmir Arkeoloji Müzesi inv. Foça 1955: Ýren 2002, 203 no. 60, fig. 13l. Stylistically related vessels of uncertain shape 1. Wall fragment: Ö. Özyiðit 1993, 5, fig. 13 (2nd row, at the right edge). Stylistically related vessels of different shape 2. Shoulder fragment of a closed vessel (with a sphinx), inv. Foça 1955 O çukuru Kuyu I: unpublished, sample no. Phok 29, provenance group G. 3. Handle fragment of a closed vessel, inv. Foça 1956 D çukuru D Odasý 290-255: unpublished, sample no. Phok 31, provenance group G. Pitane 1. Dinos, Ýzmir, Arkeoloji müzesi inv. 5018: Dedeoðlu 1993, 21; Cook and Dupont 1998, 60-1, fig. 8.23; Ýren 2002, 167, 169, 178, 182, 184, 203 no. 62, fig. 3. 2. Dinos, Ýzmir Arkeoloji Müzesi inv. 5794: Akurgal 1987, 24, pl. 4a, 10d; Akurgal 1993, pl. 112; Ýren 2002, 185, 203 no. 61, fig. 2. Stylistically related vessels of different shape 1. Oinochoe: Ýren 2002, 178-9, 204 no. 79, fig. 11. Pyrrha on Lesbos 1. Several fragments of a dinos, Göttingen, Archäologisches Institut der Universität: Schiering 1967, 432-3, fig. 28; Walter 1968, 78, 128 no. 631, pl. 131; Walter-Karydi 1970, 4, pl. 4.1-2; Schiering 1981-3; Kerschner 1997a, 15, 24, 27 no. G, fig. 13; Ýren 2002, 184, 203 no. 63, figs 4-5. 2. Rim fragment of a dinos, Göttingen, Archäologisches Institut der Universität: Ýren 2002, 203 no. 64, fig. 9a. Saint-Blaise 1. Wall fragments of a dinos: Bouloumié 1992, 212, 214 no. 574, fig. 55. 2. Wall fragment of a dinos: Bouloumié 1992, 212, 214 no. 575, fig. 55. This small fragment may belong to the same vessel as no. 1, but this cannot be verified, since the exact finding spot of no. 2 is not known. Stylistically related vessels of different shape 1. Wall fragment of an oinochoe: Rolland 1964, 571-2, fig. 43; Bouloumié 1992, 224, 227 no. 613, fig. 59. Selinus 1. Wall fragment, probably of a dinos: Rallo 1976/7, 730, pl. 164.2. Wall fragment, probably of a dinos: Dehl-von Kaenel 1995, 345-6, 359 no. 3489, pl. 63; Ýren 2002, 203 no. 66. 2. Wall fragment, probably of a dinos: Dehl-von Kaenel 1995, 345-6, 359 no. 3490 (possibly belonging to the same vessel as no. 3); Ýren 2002, 203 no. 67. Stylistically related vessels of uncertain shape 1. Salinas 1884, 330, pl. 5.41; Walter-Karydi 1970, 4 no. 15; Ýren 2002, 203 no. 65.

On the Provenance of Aiolian Pottery Stylistically related vessels of different shape 1. Lower part of an oinochoe: Gàbrici 1927, 315, pl. 82, 2; Kardara 1963, 274; Walter-Karydi 1970, 4 no. 17, pl. 5.2; Dehl-von Kaenel 1995, 345-6 no. 3483a-d, pls 63, 84; Ýren 2002, 181, 205 no. 81. Ýren, making out North Ionian features in the rendering of the wild goat, assumes that ‘das Gefäß aus einer nordionischen Werkstatt stammen [muss]’. There are, however, also Aiolian characteristics, especially the abstract rectangular shape of the purple spots on the bodies of the animals and the large, dense filling ornaments with a predilection for the hook square and the cross with inserted chevrons. This mixing of elements of different styles suggests that the painter migrated from one region to the other, thus being familiar with both the North Ionian and the Aiolian style. A similar case was revealed by the NAA series on the group of the ‘Borysthenes amphorae’, cf. above and Kerschner 2006; Posamentir and Solovyov 2006. 2. Two rim fragments, presumably from the same dish: Gàbrici 1927, 314 nos 2-2a, pl. 81; Kardara 1963, 276 nos 1-2; Walter-Karydi 1970, 4 nos 1920, pl. 3.4, 7; Dehl-von Kaenel 1995, 358-9 no. 3488; Gàbrici, who alone saw the original pieces, thought they were parts of the same dish, whereas Kardara and Walter-Karydi assumed the contrary; Dehl-von Kaenel leaves the question undecided. Smyrna 1. Akurgal et al. 2002, 87-8, 109-10 no. 73, fig. 40, pl. 6. Akurgal et al. 2002, 87-8, 112 no. 84, fig. 55, pl. 8. Stylistically related vessels of uncertain shape 1. Wall fragment, possibly from the upper part of a dinos stand, inv. BYR 75 B3 *10.42 Env K 8 20: unpublished, sample no. Smyr 45, provenance group G. Unknown provenance 1. Antikenmuseum Basel inv. BS 452: Schefold 1966, 57; Walter-Karydi 1970, 3-4, pls 1-2, 3.1, 3.3; Akurgal 1987, 24-5, pl. 4b; Akurgal 1993, pl. 113; Ýren 2002, 166, 204 no. 69; Ýren 2002, 204 no. 69. The following pieces should be eliminated from the list of WalterKarydi 1970, 3-4: No. 9: Stevenson 1890/1, 100-2, fig. 36. Neither the filling ornaments nor the square pattern on the dividing band are consistent with the London Dinos group (cf. Ýren 2002, figs 19-21). No. 16: Fairbanks 1928, 110 no. 321.4, pl. 34. The fragment belongs to the late phase of the North Ionian Wild Goat style (NiA) using the blackfigure technique. No. 21 (two fragments of a dish found in the sanctuary of Parthenos in Neapolis / Kavalla): Bakalakis 1937, 61, fig. 3; Bakalakis 1938, 114-5, figs 6, 7.1. The fragments show no clear characteristics of Aiolian vase-painting, where the shape is rare. This dish might have been made in the region where it was found, on Thasos or in its peraea.

Illustration credits

Fig. 1: I.E. Kowalleck (Vienna) after a sketch by the author; Figs 2-11: British Museum; Figs 13-14, 16-19, 21-24: Göttingen, Archäologisches Institut der Universität, photo: S. Eckardt; Figs 15, 30: Robertinum der Martin-Luther-Universität Halle-Wittenberg, photo: H. Löhr (Halle); Figs 20, 27: M. Akurgal (Ýzmir), photo: author; Fig.26: author; Figs 25, 29-30: M. Frasca (Catania); Fig. 31: I. Benda-Weber (ÖAI, Vienna) after a sketch by the author; Fig. 32: author, photo: U. Gericks (Münster); Figs 33-34: H. Mommsen (Bonn).

Notes *

First of all I want to thank the organisers of the 28th British Museum Classical Colloquium, U. Schlotzhauer (Berlin), A. Villing and D. Williams (London) for their kind invitation to participate and for creating the unparalleled possibility of discussing ceramic questions holding the original pots in one’s hands. Furthermore I thank H. Mommsen (Bonn) for hundreds of analyses and innumerable explanations during 15 years of collaboration in our archaeometric project on the pottery centres of the East Aegean. M. Akurgal (Ýzmir) and M. Frasca (Catania) kindly allowed me to publish samples from Kyme, Phokaia and Smyrna, A. Villing and D. Williams four samples from Naukratis in the British Museum (Nauk 12, 13, 64, 77). An exhaustive publication of these pieces together with them is in preparation. W. Geominy (Akademisches Kunstmuseum der Universität Bonn) and D. Graepler (Archäologisches Institut der Universität Göttingen) kindly gave me the permission to take samples of the fragments from Phokaia and Larisa in their collections, providing me also practical support. I thank M. Akurgal, M. Frasca, D. Graepler, H. Löhr (Halle), A. Villing and D. Williams for

1 2 3 4 5 6 7 8 9 10 11 12

13

14 15 16

17

18

19 20 21 22

23

24 25 26

27 28

photographs. D. Hertel (Bochum) and A. Pautasso (Catania) kindly gave me information on pieces they are preparing for publication. Concerning the spelling of Greek toponyms, I have tried to retain the Greek version, avoiding Latinisation, as Hall 1997, xv proposed. Huxley 1966, 36-9; Cook 1975, 776-80; Coldstream 1977, 262; Gschnitzer and Schwertheim 1996, 337-9. Cook 1973, 360-3; Cook 1975, 781-2; Gschnitzer and Schwertheim 1996, 339; Arslan and Sevinç 2003, 232-5, 248-9. Petrie 1886b; Gardner 1888; Hogarth 1898/9. On the history of research on East Greek pottery: Cook 1997, 295-300; Cook and Dupont 1998, 5-7; Akurgal et al. 2002, 28-36. Price 1924, 193-4, figs 20-2. Price 1924, 193. Homann-Wedeking 1938, 16: ‘Gruppe P’; he added an amphora from Saqqara (no. P 5), which, however, does not go with this group, as Schiering 1957, 116 n. 95 already observed. Schiering 1957, 14, pl. 9.1. Sidorova 1962, 107-8, fig. 1.1-2. Kadara 1963, 271-6, figs 258-65. Schefold 1966, 57. Walter-Karydi 1970. Kopeikina 1970b, 562-5, pls 1.3-5; 2.1-2. She thinks that these dinoi were produced in several workshops of the ‘North Ionian school of the Rhodo-Ionian pottery’. On the pottery finds from Berezan cf. Posamentir and Solovyov 2006 and the contribution of R. Posamentir in this volume. Schiering 1981-3; yet he stills argues in favour of ‘migrating workshops’ (ibid. 209). On the dinos from Pyrrha cf. Ýren 2002, 1703, 197, 203 no. 63, figs 4-5 (‘Gruppe A oder die Gruppe der Steinböcke mit den ausgesparten Hörnern’). On South Ionian models: Kerschner 1997a, 23-5 (‘group of the volute dinoi’). Cook and Dupont 1998, 60-1, fig. 8.23. Ýren 2002. The identification goes back to Ramsay 1881, 279-83. Yet Cook 1958/9, 20-1 n. 47, voiced doubts: ‘If not Larisa, this site should be Cyllene.’ Cf. Cook and Dupont 1998, 5, who refer to the site only as ‘Larisa’. Boehlau and Schefold 1942. Schefold used for his publication the notes of the late J. Boehlau (ibid. IX-X). His detailed analysis was based mostly on stylistic criteria for lack of stratigraphy, cf. ibid. 1-2, 58-9. The stratigraphic method was applied only in the excavation of 1932, cf. Schefold 1933, 141 (‘Dalmans sorgfältige Schichtengrabung, die sich nach seinem Auftrag auf eine Nachuntersuchung von 1902 nur teilweise freigelegten Teilen der Akropolis beschränkte...’). Akurgal 1960; Cook and Blackman 1964/5, 35-6, fig. 5; Metzger 1969, 107-8, pls 61-2; Akurgal 1987, 24-5 pl. 4.10-1, 18, 103-5; Akurgal 1993, pl. 12.111-2, 115-9. On earlier excavations at the site: Pottier et al. 1887, 504-5 figs 57-8. Ýren 2003. Ýren 2003, 9-56. On the ceramic finds: Utili 1999, 6-95, 145-267, figs 1-28, 30-44 (including a bibliography of articles in preliminary reports). Myrina: Rayet 1884; Pottier et al. 1887, 232-3, 499-504, figs 36, 55-6, pl. 51. Gryneion: Ýren 2002, 174-6, 205 nos 85-6, figs 6-8; Ýren 2003, 4, 9-10, 12, 14, 16-8, 24, 50-1, 65, 76, 83, 85, 92, 94, 97-8, 155-6, 164-5, 167, 175-9, 182, 184-5, 188, 190-1, nos 12, 15, 46, 93, 123, 125, 129, 131, 139, 140, 176, 179, 203-4, 206, 244-6, 288, 296, 322, 333, 354, 369-70, fig. 17-8, 24, pl. B, 1, 3, 24, 31, 33-4, 43, 47, 50-1, 59-61, 67-8, 71. On Geometric and Archaic pottery finds from early excavations: Dümmler 1888; Reinach 1889; Cook 1954, 27-8, pl. 14; Akurgal 1956a, 11-4; Akurgal 1956b, 23-4; Bouzek 1974, 77, pls 17-8. On the history of the excavations at Kyme: Ýdil 1989, 526-8; Frasca 1998, 273-5. Strabo 13.3.6 (translated by H.L. Jones). Frasca 1993; Frasca 1998; Frasca 2000. On Geometric and Archaic pottery from Lesbos: Mytilene: Schefold 1933, 151-2, 154, 157, figs 11-2; Schaus 1992. Methymna: Lamb 1932, 49, fig. 1-3; Buchholz 1975, 90-105, figs 25-9, pls 16-17. Antissa: Lamb 1931/2, 51-60, figs 6-9, pls 20-4. Pyrrha: Schiering 1981-3; Schiering 1989; Utili 2002. For an overview on Lesbos during the Geometric and Archaic periods: Spencer 1995. Spencer 1995, 301. Cf. Lamb 1932; Schaus 1992, 356; Bayne 2000, 21117, 307. E.g. Dugas 1912a, 519: ‘en raison des lieux de trouvaille’; HomannWedeking 1940, 28: ‘Es versteht sich, daß für die landschaftliche Bestimmung von Denkmälern der Fundort das erste Argument ist.’; Walter 1968, 9: ‘Um die Gefäße nach Landschaften zu scheiden, darf man nicht ausgehen von ostgriechischen Gefäßen in europäischen Naukratis: Greek Diversity in Egypt | 119

Kerschner

29 30 31

32 33 34 35

36 37

38

39 40 41 42 43 44 45 46 47 48 49 50 51

52 53 54 55

Museen,... sondern allein von den Gefäßen und Scherben der Grabungsorte.’; Ýren 2003, 50, 130: ‘Bevor keine Tonanalysen der aiolischen Keramik durchgeführt worden sind und konkrete Ergebnisse vorliegen, kann man Vermutungen über die Herkunft des aiolischen Tierfriesstils nur nach den Fundstücken äußern.’ Kerschner et al. 1993; Akurgal et al., 63-72; Kerschner (forthcoming). See already Cook 1959, 118, 123. Schefold 1942, 132: ‘Larisa selbst war gewiß nicht der Hauptort der Aiolis. In der geschichtlichen Überlieferung ist es Kyme, man darf aber damit rechnen, daß die noch bedeutendere ionische Stadt Phokaia zum gleichen Kunstkreis gehörte, vielleicht sogar führend in ihm war.’; Walter-Karydi 1970, 10: ‘Fragt man nach den Zentren äolischer Kunst, so scheint Phokäa ... an der kleinasiatisch-äolischen Küste führend gewesen zu sein.’; Ýren 2002, 165, 194, 197 considers Phokaia as home of the London Dinos group, which he calls ‘pseudoAiolian’, see also Ýren 2003, 157. Cook and Dupont 1998, 56-7. Dupont 1983, 22-3. Boardman 1978a; J. Boardman in Jones 1986. Cf. the contributions by R. Attula, R. Posamentir, U. Schlotzhauer, A. Villing, S. Weber and D. Williams in this volume. Participants of the network on archaeometric provenance studies of East Greek and Western Anatolian pottery are M. Akurgal (Ýzmir), R. Attula (Greifswald), T. Bakýr (Ýzmir), M. Berg Briese (Odense), J. Boardman (Oxford), N. Cahill (Madison), M. Frasca (Catania), C.H. Greenewalt, Jr. (Berkeley), G. Gürtekin-Demir (Ýzmir), M.-C. Lentini (Giardini Naxos), H. Mommsen (Bonn), W.-D. Niemeier (Athen), A. Pautasso (Catania), R. Posamentir (Istanbul), A. Ramage (Ithaca), M. Rautman (Columbia), U. Schlotzhauer (Berlin), M. Vakhtina (St. Petersburg), A. Villing (London), S. Weber (Mainz) and D. Williams (London). Cf. the contribution by H. Mommsen and the present author in this volume. Coldstream 1977, 262. Cf. Boehlau and Schefold 1942, 59, 170; Coldstream 1968, 297-8 (‘it must therefore remain an open question whether any Aeolian city ... produced any painted pottery earlier than the Subgeometric from Buruncuk’); Cook and Dupont 1998, 23; Ýren 2003, 8 (considers the possibility that ‘einige Werkstätten ... vielleicht schon am Ende des 8. Jhs. ihre Tätigkeit aufnahmen’). Only Frasca 2000, 394, fig. 280 thinks that at Kyme ‘nella produzione locale accanto al bucchero sembra presente sin dagli inizi anche la ceramica dipinta.’ He considered the krater Kyme 08 as a local product. The element pattern of this sample is still a chemical loner. Dugas 1912a, 508-9; Coldstream 1968, 298; Bouzek 1974, 77; Coldstream 1977, 263-4; Cook 1998, 23: ‘... in Aeolis, there is so far no evidence for the making of Geometric painted pottery...’; Frasca 1998, 276, fig. 8 (‘inducono a pensare ad una provenienza smirnea’); Ýren 2003, 8 (‘Die anderen spätgeometrischen Gefäße aus Kyme und Myrina erwecken zweifellos auf den ersten Blick den Eindruck der Importstücke.’). Frasca 2000, 394, fig. 280. He thought that the piece might be local because of its greyish fabric. Against: Ýren 2003, 8. Unpublished. Inv. 88.IV.26.1, sample no. Kyme 07; Frasca 1993, 60-1, fig. 25; Frasca 1998, 275-6, fig. 7; Frasca 2000, 394-5, fig. 281. Ýren 2003, 156-8, table 1. Boehlau and Schefold 1942, 59-61, pls 13-15. Ýren 2003, 9-56. Boehlau and Schefold 1942, 60, pl. 14.7 (‘subgeometrisch’); Ýren 2003, 7, 163 no. 4 (‘aiolisch-geometrische Keramik’; ‘kann subgeometrisch sein’). Boehlau and Schefold 1942, 60, pl. 14.6 (‘subgeometrisch’); Ýren 2003, 16, 18, 24, 167 no. 50 (‘Punktstil II’). Inv. Foça 1955 O çuk. B od. Unpublished. Cf. Boehlau and Schefold 1942, 193. Ýren 2003, 155. Ýren 2003, 131, 157. British Museum GR 1886.4-1.1294, unpublished; cf. the dinoi in Basle: Walter-Karydi 1970, pl. 3.1 and Berezan: Posamentir and Solovyov 2006, fig. 2 left. For a compilation of the rim ornaments see Ýren 2002, 187, fig. 19a-d. Akurgal 1961, 180, fig. 128; Walter-Karydi 1970, 6-7, 12-3, 18, pl. 8.3 (erroneously considered as a chalice, as the first published photo suggested a horizontal upper edge); Akurgal 1987, 25, pl. 3b. Akurgal et al. 2002, 109-10 no. 73, fig. 40, pl. 6. Akurgal et al. 2002, 112 no. 84, fig. 55, pl. 8. Inv. Foça 1955 O çukuru Kuyu I. Unpublished.

120 | Naukratis: Greek Diversity in Egypt

56 The small shoulder fragment was found at Berezan by E. v. Stern, who donated it to the Robertinum at Halle (inv. 480), cf. Kerschner 2006. 57 Sample nos Bere 106, 109, 110; Posamentir and Solovyov 2006. 58 Boehlau and Schefold 1942, 77, pl. 29.4. 59 Boehlau and Schefold 1942, 78, pl. 30.10-11. 60 Boehlau and Schefold 1942, 78, pl. 30.2. 61 Boehlau and Schefold 1942, 66-7, pl. 19.1. 62 Inv. BYR 74 ‘M’ Döküntü. Unpublished. For the decoration cf. the dinoi Ýren 2003, 80, 175 nos 125-6, fig. 36, pl. 32. 63 Boehlau and Schefold 1942, 97, pl. 42.1. 64 Boehlau and Schefold 1942, 97, pl. 42.2. 65 Boehlau and Schefold 1942, 91-3, pls 39-40 (‘einheimische dunkelgrundige Keramik’); Ýren 2003, 80-1, 175 no. 128, pl. 32; 178-9 no. 203, pl. 59. 66 E.g. Boehlau and Schefold 1942, 91-2, pl. 39.3-4. 67 British Museum GR 1888.6-1.573b,c; unpublished. 68 Posamentir and Solovyov 2006, fig. 3. 69 Ýren 2002, 178-9, 204 no. 79, fig. 11. 70 Kinch 1914, 190-2, pl. 11.1-3. 71 Payne 1931, 19-20, 277; Villard 1966, 49-50, pls 47.1-2, 48.3, 49; Amyx 1988, 39-40. 72 Boehlau and Schefold 1942, 92, pl. 40.12. 73 E.g. Graeve 1973/4, 88, 103-4 nos 90-2, pl. 27; Walter-Karydi 1973, 1920, 81, fig. 23, pls 36.277, 36.287, 81.941, 81.943; Kerschner 1997b, 1257, 186 no. 38, fig. 21, pl. 5. 74 Boardman 1967, 119 nos 199-204, pl. 32; 123, 125 nos 298-301, pl. 37 (Emporio); Kerschner 1999, 20-1, 41 no. 32, fig. 10 (Miletos). 75 Lemos 1991, 118-24, pls 59-60, 64, 66-7, 69-73, 75-7, 79, 83-4, 87. For kantharoi and phialai in light on dark techique see Lemos 1991, 119, 121, pl. 90. 76 Kinch 1914, 168-88, figs 58-72, pls 10, 12; Cook and Dupont 1998, 1145. 77 Boehlau and Schefold 1942, 157, pl. 52.6. 78 Inv. K 95.VIII.1.R. Unpublished. 79 Akurgal et al. 2002, 85, 104 no. 51, pl. 3. 80 Unpublished fragment from the excavations of E. Akurgal, inv. 1956 Foça D a odasý. 81 Inv. 88.IV.29.5. Unpublished, for a similar piece: Frasca 1993, 65 no. 33, fig. 11a. 82 Unpublished, for comparable pieces: Boehlau and Schefold 1942, 103-6, fig. 31, pl. 45 (‘Graue Becken’); Frasca 1993, 63 no. 12, fig. 3b. Further examples of Aiolian Grey ware, found at Larisa, were analysed by D. Hertel and H. Mommsen. 83 British Museum GR 1888.6-1.637; unpublished. The irregular shape of the bars may point to a figured decoration (A. Villing). Small impressed circles occur on Grey ware from Larisa: Boehlau and Schefold 1942, 121, 127, pl. 44.8,10-11,19, where fenestrated stands are represented, too: Boehlau and Schefold 1942, 128, pl. 48, 19-22. 84 Sample no. Phok 32, inv. Foça D 235-245, an unpublished fragment from E. Akurgal’s excavations at Phokaia, decorated with a lotuspalmette-frieze of Late Archaic style. 85 British Museum Inv. GR 1888.6-1.658; Gardner 1888, 58, pl. 14.5. The preserved height is 6.4 cm. The body of the figure, the back of the head and the ears are lost. The details of the head were moulded with the fingers in small flat pieces of clay that were added to a hollow core. Cf. the few early terracottas from Larisa: Boehlau and Schefold 1942, 25-32, pl. 4-6. 86 Cf. the faces of the plastic kantharoi and mugs: Himmelmann 1973, 28, colour pl. 2; Walter-Karydi 1973, 30, 131 no. 485, pl. 57; Schlotzhauer 1999, 236 no. 19, fig. 24; Schlotzhauer (forthcoming a). Cf. also the bearded heads of the triple-bodied monster from the Athenian Acropolis: Boardman 1978b, 154, fig. 193; Rolley 1994, 194, fig. 6. 87 G.M.A. Hanfmann in Hanfmann 1983, 80, figs 142-3. 88 Walter 1968, 77-9 (‘Larisäisch’); Walter-Karydi 1970, 10, 14; Akurgal 1987, 25; Cook and Dupont 1998, 57-61; Ýren 2003, 50, 131, 157. K. Schefold, although dealing intensively with Aiolian pottery, did not comment explicitly on the question of the production centres. Yet he considered most of the ceramic finds from Larisa to be local and distinguished them from other East Greek wares: Boehlau and Schefold 1942, 58-169. Schefold 1942, 132 seemed to regard the ceramic finds from Myrina and Pitane as local. Furthermore he conjectured ‘daß die noch bedeutendere ionische Stadt Phokaia zum gleichen Kunstkreis gehörte, vielleicht sogar führend in ihm war.’ 89 Boehlau and Schefold 1942, 169 (‘Larisa war reich nur an Geschenken des eigenen Bodens, an Werken einheimischer Hände...

On the Provenance of Aiolian Pottery Sobald das einheimische Handwerk leistungsfähig war, brauchte man nichts mehr einzuführen – solange man mit der heimischen Kunst zufrieden war.’); Schefold 1942, 132; Cook and Dupont 1998, 57 (Cook treats the London Dinos group as a third ‘assemblage’ of Aiolian Wild Goat style without proposing a precise localisation). 90 Walter 1968, 78-9 (‘Der larisäische Stil ist ein Stil der Hinterwäldler... Larisa stand im Ausstrahlungsbereich von Phokäa und Kume.’); Walter-Karydi 1970, 10 (‘Fragt man nach den Zentren äolischer Kunst, so scheint Phokäa ... an der kleinasiatisch-äolischen Küste führend gewesen zu sein... Kyme hatte anscheinend einen rein bäuerlichen Charakter.’), 14 (‘Und doch müsste Lesbos, der Stellung von Samos und Chios entsprechend, die reinste Möglichkeit äolischer Art vertreten.’); cf. also Akurgal 1987, 25 (‘Ferner kommen in den Zentren Pitane, Myrina und Larisa provinzielle, aber reizvolle Schöpfungen einer naiven Volkskunst auf.’). 91 Ýren 2003, 157. 92 Ýren 2002, 197. 93 Ýren 2002, 165; Ýren 2003, 157. 94 Ýren 2002, 197; Ýren 2003, 50, 130 (‘Bevor keine Tonanalysen der aiolischen Keramik durchgeführt worden sind und konkrete Ergebnisse vorliegen, kann man Vermutungen über die Herkunft des aiolischen Tierfriesstils nur nach den Fundstücken äußern.’). 95 Ýren 2003, table 1. 96 Akurgal et al. 2002, 76, 85, 104 nos 50-2, pl. 3; Kerschner 2006; Posamentir and Solovyov 2006. 97 Akurgal et al. 2002, 90-1 nos 79, 83, figs 48, 54, pls 6, 8; Kerschner 2006; Posamentir and Solovyov 2006. 98 On the provenance group X: M. Kerschner in Badre et al. 2006, 36-7. 99 The as yet unpublished sample Smyr 40 belongs to provenance group F that is presumably located at Smyrna. Cf. also Paspalas, this volume, esp. n. 68. 100 On the class: Kerschner 2006. Kardara 1963, 209-10 fig. 180 was the first who recognized this group and called it ‘sxolh\ oi0no/xohj 0Ocfo/rdhj’. Alexandrescu 1978, 23 n. 23, 37-8, proposed a subdivision into a ‘classe de Lévitsky’ and a ‘classe de Tocra cat. 580’, but both classes differ only slightly in the shape, whereas the style of painting is homogenous. Therefore I have proposed to reunify both classes in accordance with Ch. Kardara and J. Hayes (in Boardman and Hayes 1966, 41-2). 101 Kerschner 2006; Posamentir and Solovyov 2006. Furthermore, our NAA detected a production in a colonial workshop at the Hellespont, the Propontis or the Black Sea (sample no. Bere 007, provenance group BERa). 102 E.g. Kopeikina 1968, 44-7, figs 1-3; Kopeikina 1981, 196-7, fig. 4c; Kerschner 2006; Posamentir and Solovyov 2006 (Berezan); Sidorova 1962, 108, fig. 1.4 (Pantikapaion); Lambrino 1938, 244-9, figs 208-14; Alexandrescu 1978, 37-8 nos 2-10, 12, 16, pls 1-2 (Istros); Dugas 1935, 58-60, pls 39-40, 41.12-4 (Delos); Ploug 1973, 50-2, 59-64, pls 9-13 (Tell Sukas); Boardman and Hayes 1966, 41, 46 nos 580-1, pl. 28 (Taucheira). For a more comprehensive list see Kerschner 2006. 103 Kerschner et al. 2002, 203-5; Kerschner 2006. See also the discussion of a Wild Goat style oinochoe found at Selinus in Appendix 2. The vessel exhibits both Aiolian and North Ionian stylistic features. 104 This concept differs fundamentally from the hypothesis of the ‘wanderende Werkstätten’ formulated by Schiering 1957, 1, 8-14. In contrast to me, Schiering postulates a systematic migration of whole workshops and the foundation of branches at several places, which consistently stick to their stylistic tradition even in a new environment and over long distances. 105 In order to assess this evidence it is important to mention that we have not yet analysed any finds from the Aiolian island of Lesbos. 106 Özyiðit 1991, 137-9, figs 1-2, 7-10; Özyiðit 1992, 102-4, figs 3-16. 107 Akurgal et al. 2002, 89, 116; Kerschner and Mommsen (forthcoming). The samples were taken within the context of an unpublished project of U. Outschar (Istanbul) and R. Sauer (Vienna). 108 Langlotz 1969, 379, 381, figs 4-6. For the shape: Hayes 1972, 333, 337, figs 67-8 (type C). The analyses of this waster by P. Dupont and M. Picon (cf. Mayet and Picon 1986) prompted Hayes 1980, 525 to rename the ‘Late Roman C Ware’ ‘Phocaean Red Slip Ware’. 109 In this point, our NAA corroborate the result of Dupont 1983, 22-3. 110 It cannot be totally excluded that the Phokaians exploited other clay beds showing different element patterns during Roman Imperial

111 112 113 114

115

116 117 118

119

120

121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137

times than they did in the Geometric to Archaic periods. There is, however, an argument against this possibility: The provenance group G/g comprises also Hellenistic and Roman pottery differing from the element patterns of the wasters from Phokaia. Walter-Karydi 1970, 10; Ýren 2002, 165, 194, 197; Ýren 2003, 157, cf. already Schefold 1942, 132. A short preliminary notice: Özyiðit 2004, 443-4. The excavator announced finds of ‘Orientalising pottery produced in the region of North Ionia and the Aiolis’ (Özyiðit 2004, 444). Euboian (or Boiotian): Kyme 04, inv. 90.IV.29.99, wall fragment of a krater with bands and a wavy line in added white, Frasca 1993, 66 no. 49, fig. 16; Kyme 10, inv. 89.IV.29.17, late Geometric skyphos with chevrons, Frasca 1993, 67 no. 63, fig. 20b; Frasca 1998, 276-7, fig. 10; Kyme 11, inv. 88.IV.33.2, rim fragment of a krater with concentric tangential circles, Frasca 1993, 67 no. 61, fig. 19b; Frasca 1998, 277-8, fig. 15. Corinthian: Kyme 09, inv. 95.IV.US.2, late Geometric skyphos, unpublished, cf. Frasca 2000, 395-6, fig. 282. North Ionian (‘bird bowl workshops’ of provenance group B): Kyme 07, inv. 88.IV.26.1, bird kotyle, Frasca 1993, 60-1, fig. 25; Frasca 1998, 275-6, fig. 7; Frasca 2000, 394-5, fig. 281. Ýren 2003, 131, 157 argues in favour of Kyme or Myrina as production place of the Aiolian Wild Goat style pottery of superior quality (‘Da die Gefäße des aTs [= aiolischer Tierfriesstil] und IIIa sorgfältigere Stile als aTs IIIb haben, könnte man sich die großen und wichtigen aiolischen Städte wie Kyme und Myrina für sie als Herstellungsort vorstellen.’). Although he considers the London Dinos group as Phokaian (Ýren 2002, 194, 197), he admits: ‘... dennoch darf man sie [= Kyme] als einen möglichen Kandidaten für die Lokalisation der Werkstatt des Londoner Dinos betrachten.’ (Ýren 2002, 194). Strabo 13.3.6 (translated by H.L. Jones); cf. Cook 1975, 780. Walter-Karydi 1970, 6; Ýren 2002, 190-7. Schefold 1966, 57, on the dinos in Basle: ‘Unser Dinos gehört aber zu einer Variante [des ostgriechischen Tierfriesstils], die am häufigsten in Larisa am Hermos gefunden worden ist und in der Hauptstadt der Äolis, in Kyme, ihre Heimat gehabt haben dürfte.’ Walter-Karydi 1970, 10; Ýren 2002, 193. This opinion is mainly based on an anecdote bequeathed by Strabo 13.3.6: ‘Cymê is ridiculed for its stupidity, owing to the repute, as some say, that not until 300 years after the founding of the city did they sell the tolls of the harbour, and that before this time the people did not reap this revenue. They got the reputation, therefore, of being a people who learned late that they were living in a city by the sea’ (translation H.L. Jones). However, the phrasing of Strabo reveals that he already had doubts about this anecdote. Cf. Ýren 2002, 194 (‘... die Phokäer waren ein Seefahrervolk ... sie fuhren vom Kongo bis in die Nordsee’). For critical views on the concept of a Phokaian thalassocracy see: Niemeyer 1988/90, 269306; Gassner 2003, 261-75; Kerschner 2004. Boardman 1999a, 125. On the commercial activities of Aigina: Johnston 1972; Johnston 1979, 51-2; Hiller 2000 (with further bibliography). Akurgal et al. 2002, 111 no. 79, fig. 48, pl. 6. Posamentir and Solovyov 2006. Kerschner 2006. Kerschner 2006. Kerschner 2006. Posamentir and Solovyov 2006. A detailed publication of the NAA from Kyme is being prepared by M. Frasca, M. Kerschner and H. Mommsen. A detailed publication of the NAA from Phokaia is being prepared by M. Akurgal, M. Kerschner and H. Mommsen. Posamentir and Solovyov 2006. Kerschner 2006. Kerschner 2006. Kerschner 2006. Posamentir and Solovyov 2006. Kardara 1963, 271-6, figs 258-65 (‘ergasterion dinou’); Walter-Karydi 1970, 3-4; Ýren 2002, 198-206; cf. M. Kerschner in Akurgal et al. 2002, 87 with n. 549. The piece was kindly shown to me by the excavator at a visit on Berezan with an excursion of the University of Vienna in summer 1999.

Naukratis: Greek Diversity in Egypt | 121

Kerschner

Figure 2 Dinos,Aiolian Wild Goat style, eponymous piece of the London Dinos group, presumably from Kameiros. London, British Museum GR 1848.6-19.1

Figure 3Wall fragment of a dinos,Aiolian Wild Goat style, London Dinos group, from Naukratis. London, British Museum GR 1886.4-1.1270

Figure 5Wall fragment of a dinos (?), Aiolian Wild Goat style, London Dinos group, from Naukratis. London, British Museum GR 1888.6-1.470

Figure 4Wall fragment of a dinos,Aiolian Wild Goat style, London Dinos group, from Naukratis. London, British Museum GR 1886.4-1.1288

Figure 6Wall fragment of a dinos (?),Aiolian Wild Goat style, London Dinos group, from Naukratis. London, British Museum GR 1924.12-1.11

122 | Naukratis: Greek Diversity in Egypt

Figure 8 Rim fragment of a dinos, London Dinos group, from Naukratis. London, British Museum GR 1886.4-1.1294, sample no. Nauk 13. Provenance group G

On the Provenance of Aiolian Pottery

Figure 7 Askos,Aiolian Wild Goat style, stylistically related to the London Dinos group, from Naukratis. London, British Museum GR 1888.6-1.462

Figure 11 Head of a painted terracotta figurine, from Naukratis. London, British Museum GR 1888.6-1.658, sample no. Nauk 77. Provenance group g

Figure 9 Shoulder fragment of a black-polychrome oinochoe, from Naukratis. London, British Museum GR 1888.6-1.573b,c, sample no. Nauk 12. Provenance group G

Figure 10 Fragment of fenestrated stand (?),Aiolian Grey ware, from Naukratis. London, British Museum GR 1888.6-1.637, sample no. Nauk 64. Provenance group g

Figure 12 Wall fragment of a Late Geometric krater or dinos from Larisa. Göttingen, sample no. Lari 12. Chemical single

Figure 13 Dinos,Aiolian Subgeometric or ‘dot style’, from Larisa. Göttingen, inv. 22, sample no. Lari 15. Provenance group G

Naukratis: Greek Diversity in Egypt | 123

Kerschner

Figure 14 Pyxis,Aiolian Subgeometric or ‘dot style’, from Larisa. Göttingen, inv. 23a-e, sample no. Lari 16. Provenance group G

Figure 15 Shoulder fragment of a dinos,Aiolian Wild Goat style of the London Dinos group, from Berezan. Halle, Robertinum inv. 480, sample no. Bere 178. Provenance group g

Figure 16 Skyphos krater,Aiolian Wild Goat style, from Larisa. Göttingen, inv. 38a-c, sample no. Lari 18. Provenance group G

Figure 17 Skyphos krater,Aiolian Wild Goat style, from Larisa. Göttingen, inv. 46, sample no. Lari 19. Provenance group G

Figure 18 Skyphos krater,Aiolian Wild Goat style, from Larisa. Göttingen, inv. 44a-d, sample no. Lari 20. Provenance group G

Figure 19 Oinochoe,Aiolian Wild Goat style, from Larisa. Göttingen, inv. 447, sample no. Lari 21. Provenance group G

Figure 20 Kotyle,Aiolian Orientalizing style, from Smyrna, inv. BYR 74 ‘M’ Döküntü, sample no. Smyr 46. Provenance group G

124 | Naukratis: Greek Diversity in Egypt

Figure 21 Skyphos krater,Aiolian black-figure style, from Larisa. Göttingen, inv. 358a-b, sample no. Lari 23. Provenance group G

Figure 22 Krater,Aiolian bichrome ware in black-figure technique, from Larisa. Göttingen, inv. 343, sample no. Lari 22. Provenance group G

On the Provenance of Aiolian Pottery

Figure 24 Kotyle,Aiolian ‘drop style’, from Larisa. Göttingen, inv. 294, sample no. Laris 24. Provenance group G

Figure 25 Kotyle,Aiolian ‘drop style’, from Kyme, inv. K 95.VIII.1.R, sample no. Kyme 14. Provenance group g

Figure 23 Dinos,Aiolian light-on-dark ware, from Larisa. Göttingen, inv. 250, sample no. Lari 27. Provenance group G

Figure 26 Rosette bowl from Smyrna, sample no. Smyr 32. Provenance group G

Figure 27 Dish with meander on the rim, from Phokaia (excavations E.Akurgal). Inv. 1956 Foça D a odasý, sample no. Phok 49. Provenance group G

Figure 28 Neckamphora of the ‘Borysthenes group’,Aiolian Wild Goat style, from Berezan. Halle, Robertinum Inv. 426, sample no. Bere 174. Provenance group g

Figure 29 Dish, banded ware, from Kyme, Inv. 88.IV.29, ssample no. Kyme 05. Provenance group G

Figure 30 Carinated bowl,Aiolian Grey ware, from Kyme. Izmir, Inv. 03.IV.117.7, sample no. Kyme 23. Provenance group G

Figure 32 Kiln waster of a fused stack of 6 bowls of Hayes Form 3C from Phokaia. Bonn (E. Langlotz bequest). Sample no. Phok 05

Naukratis: Greek Diversity in Egypt | 125

Kerschner Figure 33 Result of a discriminant analysis of the grouped samples from Phokaia, shown as filled-in symbols, and assuming 10 groups.The numerous chemically single samples in the set from Phokaia have also been included and are shown as stars.The different groups are described in the text. Plotted are the discriminant functions W1 and W2 which cover 73 % and 14 %, respectively, of the between group variance.The ellipses drawn are the 2s boundaries of the groups.

Figure 34 Result of a discriminant analysis of all the grouped samples (exclusive of 4 singles) from Kyme and Larisa, shown as filled-in symbols, together with some reference samples of other patterns and assuming 5 groups. Besides the predominant local group G and its subgroup g only 6 sherds have been identified as imports to Kyme: one from the ‘bird bowl workshops’ (group B), one from the north-eastern Peloponnese (Corinth), and 4 from Boeotia or Euboea (Boe/Eub). Plotted are the discriminant functions W1 and W2 which cover 95 % and 2.8 %, respectively, of the between group variance.The ellipses drawn are the 2s boundaries of the groups.

126 | Naukratis: Greek Diversity in Egypt

The Chian Pottery from Naukratis Dyfri Williams

Abstract This paper briefly examines the range of Chian pottery found at Naukratis, noting in particular some unusual shapes unique to the site, especially the phallus cup. It also offers a general review of the development of Chian decorated pottery and the workshops that produced it. This review ends with a more detailed examination of a small group of pieces with Laconian connections, the work of the Sirens Painter, who may even have been a migrant from Laconia. Finally, it summarises the debate on the places of manufacture of Chian pottery, arguing against any production at Naukratis. Chian pottery was first called ‘Naukratian’ by Flinders Petrie when the first finds of pottery were excavated at Naukratis, as Cecil Smith notes in his essay for the 1886 volume of Petrie’s publication.1 By the time of the second volume, in 1888, E.A. Gardner was calling it ‘Naukratite’.2 Kourouniotis, however, excavated a good deal of ‘Naukratite’ pottery – decorated, dedicated and plain – in his excavations on Chios in 1914–15, and suggested that it was made on that island not at Naukratis.3 The fabric was more fully studied by Elinor Price in 1924: she still called it ‘Naukratite’, but did comment that its place of origin might have been Naukratis or Chios.4 Price’s classification has been modernised over the years, especially by John Boardman and Robert Cook, and the fabric is now confidently christened Chian by all. The subject has been most fully studied and revised by Anna Lemos in her very important monograph of 1991.5 From Lemos we have a sequence of styles – the Wild Goat Style, the Animal Chalice Style, the Grand Style, the Chalice Style, the Sphinx and Lion Style, the Black-figure Grand Style, and the Black-figure Chalice Style.6 Boardman refers, even more recently, in his handbook of 1998 to Lemos’ Animal Chalice Style as ‘Animal Chalices’ and her Chalice Style as ‘Simple Animal Chalices’.7 The Orientalizing Wild Goat Style of Chios is distinct. There seems no strong Early Wild Goat phase, what Kerschner and Schlotzhauer would perhaps call Chian Archaic Ib (ChiA Ib): possible examples come from Phana on Chios, from Aigina port and from the Samian Heraion, but none were found at Naukratis.8 A more advanced group that Lemos categorises as Middle Wild Goat I (presumably still within Chian Archaic Ib – ChiA Ib) includes pieces from Chios, Salamis on Cyprus, Al Mina, Aigina and Bulgaria, although she notes a degree of hesitation over the fabric of the Al Mina fragments.9 With Chian Middle Wild Goat II (Chian Archaic Ic – ChiA Ic), however, we leave behind the realm of uncertainty and the quantity of material from Naukratis is of particular and immediate importance. Lemos has associated a number of Naukratis fragments with the pair of chalices from Vulci in Etruria, now in Würzburg, under the sobriquet the Painter of the Würzburg Chalices.10 She charts this painter and his followers over a number of years: indeed, a chalice fragment already has added

decoration inside and so takes us down to about 610 bc.11 The other major artist of the time that Lemos isolates is the painter of the large bowl dedicated by Sostratos to Aphrodite at Naukratis (Fig. 5).12 Her list of this Painter of the Aphrodite Bowl, and the group that she associates with it, seems to follow a similar course as that of the Painter of the Würzburg Chalices, although beginning perhaps slightly later. Again, one of the pieces that she attributes to the painter himself already has interior decoration.13 As regards the Middle Wild Goat II pottery from Naukratis, we might perhaps begin with a remarkable shape that is not found anywhere else: the phallus vessel. Robert Cook listed three examples in his article of 1949, the two London fragments from Naukratis in the British Museum and a fragment on the Athenian Acropolis.14 The latter was drawn to his attention by the late Martin Robertson, but the illustration and description of it in van Buren’s book on Athenian architectural terracottas reveals that it is a very different thing, and clearly neither Chian nor any sort of vessel attachment or protome. Instead, however, we might now mention, even though not Chian, a fragment from the excavation of the cathedral in Marseilles which preserves part of the erect shaft and the testicles below; there are fingers attached to the shaft.15 The piece looks thick-walled and it is described as large; the shaft is undecorated, the glans missing, but the testicles have glazed dots, recalling Ionian plastic vases, with which it is perhaps connected. Indeed, it may have come from an object, or rather pouring vessel, like the extraordinary terracotta seated man with erect penis from Sardis or the smaller, earlier and cruder version from Samos.16 The better preserved of the two Chian phalloi from Naukratis has now been augmented by a small fragment with more of one of the goats (Fig. 1).17 The phallus is carefully modelled, not pierced, and the female pudenda strangely added at the beginning of the shaft, perhaps by way of a reminder of the intended target. The less well-preserved example is slightly larger, and damaged just where the female sex was similarly most probably added.18 Both are finely made and thin-walled and both glazed inside. Lemos has attributed the painting on the better preserved piece to the neighbourhood of the Painter of Figure 1 Phallus from phallus cup, outside and inside view, BM GR 1888.6-1.496a-c, with 1924.12-1.178

Naukratis: Greek Diversity in Egypt | 127

Williams

Figure 2 Phallus cup wall (?), outside and inside view, BM GR 1924.12-1.186

Figure 3 Phallus cup wall (?), outside and inside view, BM GR 1924.12-1.203

the Aphrodite Bowl.19 The addition of the small fragment makes it clear that it is from the same time as a fragment that she sees as a fully mature work (the Aphrodite Bowl itself being an early work).20 We should perhaps ask ourselves what the rest of these exceptional objects looked like. We know male genitalia used by Athenian potters for the feet of their cups in the late 6th century bc and early 5th – the black-figured example in Oxford is the best preserved, but there is also a ruined fragment of a redfigured one in Herbert Cahn’s collection.21 Corinthian, East Greek, and Attic potters also produced perfume pots in the form of male genitals, an interesting concept in itself. They are also attached to a variety of later vases which were offered, it seems, to children as feeding bottles. The fact that both Naukratis pieces are glazed inside suggests that they served as special drinking vessels, but what form did their upper parts take? Here one might consider two fragments that could have come from the upper part of such vessels, although the connection cannot as yet be demonstrated. The first was listed by Lemos as coming from a chalice, but the profile is clearly different from all chalices and indicates that there was a much narrower form below the frieze of animals and, moreover, one that seems to have been set at something of an angle to the upper cup wall (Fig. 2).22 The second fragment preserves slightly less of the form below, but both are clearly from a similar shape (Fig. 3).23 There is a third fragment of the class in Brussels.24 Such special Chian drinking vessels were surely dedications in the sanctuary of Aphrodite, perhaps used

at cultic celebrations. Another special shape to Naukratis, it would seem, is the class of the Aphrodite Bowl itself. Here I note a small fragment that joins the name-piece, giving the face of the sphinx on the interior (Fig. 4) – joined to the Aphrodite bowl in a photomontage (Fig. 5).25 In ceramic terms these large bowls with their rim mounted vertical handles would seem in some ways to be gentrifications of the humble lekane. The Aphrodite Bowl was perhaps a creation for ritual use, perhaps even a ceremonial washing of the fingers at the symposium, as if with cologne or limon. Lemos listed seven examples of the form, all from Naukratis.26 Another seemingly unique shape, though surely not a special product, is represented by a fragment that Lemos identified as a chalice (Fig. 6).27 It is, in fact, rather from a large, thick-walled closed vessel, undecorated on the inside, most probably a onepiece amphora, and is decorated with a goose. The painter’s hand can be seen on fragments of contemporary chalices and is closely related to the Painter of the Aphrodite Bowl.28 From Naukratis, of course, also come quite a few fragments of large dinoi. Here we should note that such vessels are in fact unglazed inside, a fact that can lead to the misidentification of fragments as being from oinochoai. The Wild Goat examples, which are the most numerous, had rotelle handles placed on the shoulder, below the rim, like metal cauldrons (Fig. 7).29 We might ask ourselves here the question as to why such Chian dinoi, and indeed dinoi in most other East Greek fabrics, were unslipped inside.

Figure 4 Fragment joining Aphrodite bowl, inside and outside view, BM GR 1924.12-1.418

Figure 6 Amphora fragment, BM GR 1888.6-1.475g

Figure 5 Photomontage of Aphrodite bowl (BM GR 1888.6-1.456) and joining fragment (BM GR 1924.12-1.418)

128 | Naukratis: Greek Diversity in Egypt

The Chian Pottery from Naukratis

Figure 7 Dinos fragment, BM GR 1888. 6-1.499g and b, 1924.12-1.21 and 1888.6-1.443

Figure 8 Lid fragment, Cambridge, Mus.Arch. and Anth. NA 103

With the last decade of the 7th century bc, the beginning of Chian Archaic Id, there came substantial change.30 Indeed, it seems to me that we can now talk about two parallel workshops in this phase, differentiated as they are by shape production, technique and style. The first uses the black-figure technique to enliven miniature animal friezes that concentrate on lions and sphinxes, but occasionally admit bulls, birds and sirens, as well as, very rarely, a human figure. We will do best not to confuse terminology too much and I propose to call this workshop the Sphinx and Lion Workshop, following Boardman’s naming of the ‘Style’. This workshop does not seem to have decorated chalices, whether large or small, but instead produced a series of stemmed skyphos-like vases with lids. As with the chalices, its smaller scions were presumably used as drinking vessels, the larger, such as the example from Pitane, as kraters.31 On a fragment of a large lid from Naukratis, now in Cambridge, we find a combination of incised filling ornaments and an outline goat (Fig. 8).32 This suggests that the piece belongs early in the series and, indeed, the painting of the goat suggests that at least one of the painters from the followers of the Painter of the Aphrodite Bowl became part of our Sphinx and Lion Workshop. Another Naukratis fragment is, exceptionally, decorated both inside and outside (Fig. 9). It is not from the bowl but rather from the lid, although most lid fragments are simply slipped inside with white.33 The very large fragment of a bowl, decorated inside and out and found by Kourouniotis on Chios, presumably came from the bowl of one of these extravagantly decorated lidded skyphos-kraters.34 The workshop also produced other lesser shapes and from Naukratis we have fragments of plates and small dishes – indeed, to a fragment of a plate in Cambridge we may join a piece in London (photomontage Fig. 10).35 There is the occasional surprise too, such as the fragment of a fine ring vase with a centaur depicted on it.36 The second workshop continued the old Wild Goat technique of the 7th century tradition: mixed outline and silhouette, usually abjuring incision. This I propose to call the Chalice Workshop (simplifying Lemos’ and Boardman’s Animal Chalice Styles), as a result of its preferred shape. It did, however, also produce kantharoi, phialai and plates, but no stemmed skyphoi, large or small. The shape of the chalice has changed and the metopal arrangement of decoration given way to free-field design. In addition, the interior is also now treated as a field for gloriously colourful decoration in red and white on the black ground. The iconography remains in the animal world, at least for a while, with lions, boars, bulls, sphinxes and sirens; but goats and geese have gone. What is new is a greater sense of monumentality, a concentration on only a few animals on each vase, painted on a larger scale.

Figure 9 Lid fragment, outside and inside view, BM GR 1886.4-1.998 Figure 10 Plate fragment, BM GR 1924.12-1.16 and Cambridge, Mus. Arch. and Anth. NA 98

A particularly splendid example from this workshop is a large, heavy-walled chalice-krater divided between the British Museum, Boston, and University College London.37 On one side there was a sphinx and a bull, all in outline, but with much added colour – second white for the sphinx’s face (Fig. 11), red on the bull’s hindquarters and a vivid interior floral band. A group of fragments with a lion attacking a boar should come from the other side of the vessel.38 On a lighter walled piece is found the beginnings of the introduction of the human figure – a woman in black and red garments.39 A liking for human figures, and for polychromy, seems to have grown with time. Indeed, some painters began to depict only human figures, on both small and large chalices: this perhaps already from soon after 580 bc. With the appearance of figures, the inclusion of filling ornaments is soon abandoned40 – they were a feature of the wider Greek animal frieze style, but had no real place in a mature 6th century figured style. Some of these figured chalices have more polychromy than others. In his key Chian article of 1956 Boardman introduced the term ‘Grand Style’ for these large-scale chalices with polychrome decoration and human figures,41 but this perhaps tends to isolate one part of the production of the workshop and so prevents us from seeing the whole. We also need to note that incision begins to make an appearance on some pieces: for example two fragments with horses or horsemen, one from Naukratis (Fig. 12), the other from Aigina.42 These are not from large heavy-walled chalices, but a fragment from such a polychrome chalice in Bonn reveals incision too.43 The subjects represented on these figured chalices range from gods and heroes to mortals in ritual processions and komos dances.44 The fragments are often too small to enable us to

Figure 11 Chalice-krater fragment, BM GR 1888.6-1.465a

Naukratis: Greek Diversity in Egypt | 129

Williams Figure 12 Chalice fragment: horseman, outside and inside view, BM GR 1924.121.343

Figure 14 Cup fragment interior: siren, Oxford G 133.2 and 6

Figure 15 Cup fragment, interior: two heads, BM GR 1888.6.4-1.1283

Figure 13 Chalice interior: lion, BM GR 1888.6-1.790

understand the full scenes with any confidence. We may also isolate particular painters – Lemos notes, in addition to the Aphaia Painter, the Naukratis Painter, and this is clearly an area where more work could be done.45 The pair of chalices from Pitane46 is particularly important and good published illustrations would be of great help: one might note too the lions inside in added colour – a fragment from Naukratis has the same motif (Fig. 13).47 Gradually, perhaps from around 570 bc, the simpler chalices began to minimise their decoration further and further until a vessel might just have a pair of animals, or even just one, with no filling ornaments, on one side, and on the other a simple rosette ornament, or even nothing at all.48 Instead of the animal we sometimes find a single figure, as on a chalice from Berezan.49 At some point, perhaps near 560 bc, it would seem that these two workshops merged again, for we suddenly find a continuation of the general style of the Sphinx and Lion Workshop on a series of chalices decorated with black-figure komasts, and even the reappearance of very debased filling ornaments. On one remarkable fragment from Berezan, the painter has used added white instead of incisions for the interior details of the komast.50 There are also a couple of fragments in black-figure komast style that might be from oinochoai or dinoi – the usual problem.51 Connected with this phase is a group of chalices (and a kantharos) that depict a variety of animals, specifically cocks, hens, geese and dolphins – Lemos calls it the Poultry Group.52 Some are slight works with debased filling ornaments and clearly go with the komast chalices.53 Indeed, on a fragmentary chalice from Berezan we find black-figure komasts on one side and a cock on the other.54 Others, however, have something grander and more restrained about them and are perhaps earlier, or at least retain more of the tradition of the Chalice Workshop.55 Indeed, the group of four chalices found at Tocra, two from the Chalice Workshop and two from Lemos’

Poultry Group would seem perhaps to be roughly contemporary, despite the differing techniques.56 Finally, the last gasp of decorated Archaic Chian is to be found on a group of unslipped, Atticising kantharoi and chalices that takes us down into the third quarter of the 6th century bc.57 This leaves us with what Lemos calls the Black-figure Grand Style.58 She lists fragments of two cups,59 five chalices, and a bowl, all from Naukratis, and an indeterminate vessel from Berezan. The bowl fragment should be omitted – it really finds a place in the Sphinx and Lion workshop.60 The Berezan fragment is also best omitted, as the fabric is surely not Chian.61 This leaves us with fragments of two cups, one in Oxford, one in London: both clearly by the same painter (Figs 14–15).62 The shape is different from that of the regular, stemmed Ionian cup, a shape which Chian potters also produced, decorating the rim with myrtle or laurel wreaths, 63 and with their exterior decoration and interior border of pomegranates point strongly towards the Laconian class of cup, as has been noted many times before.64 A rim fragment from a chalice in London is also clearly by the same hand (Fig. 16).65 Two of the other chalice fragments must come from one and the same vessel – indeed, we can add a third fragment to this group, which was not listed by Lemos (Figs 17–19).66 These fragments may even be from the same vessel as the rim fragment with the heads. The wall fragment from a chalice in Cambridge may be augmented by joining the last of the chalice fragments listed by Lemos, and a second London fragment that she did not include (photomontage Fig. 20).67 We thus have two cups and two or three chalices in a fine black-figure technique all by the same, rather accomplished painter – let us call him the Sirens Painter after the Cambridge and London chalice and the Oxford cup. He decorated two cups that show decided connections with Lakonian cups, both in shape and decorative scheme, and a couple of chalices that have figures that recall some of the standing and seated figures on

Figure 16 Chalice fragment: two heads, BM GR 1888.6-1.550a

Figure 18 Chalice fragment: drapery, BM GR 1924.12-1.204

Figure 17 Chalice fragment: seated figure, BM GR 1924.12-1.206

130 | Naukratis: Greek Diversity in Egypt

Figure 19 Chalice fragment: drapery and tail, BM GR 1924.12-1.342

The Chian Pottery from Naukratis Figure 20 Chalice fragment: siren and winged demon. Photomontage of, left: BM GR 1888.61.550b; right: Cambridge GR 97.1894; top: BM GR 1924.12-1.352

Lakonian cups, as well as a wing-footed figure flying over a siren that similarly recalls some of the Lakonian cups with winged daemons and sirens flying round symposium scenes.68 The closest connection seems to be with the work of the so-called Boreads Painter, usually dated 575–565 bc.69 Our Chian Sirens Painter was presumably allied to the Chalice Workshop and influenced one or more of the later painters who worked in the recombined workshop. His own contribution to the development of the earlier scions of Lemos’ Poultry Group is perhaps seen on a fine fragment of a thick-walled chalice with a cock raising one leg (Fig. 21).70 How do we explain this phenomenon? Boardman has noted that no Laconian pottery has yet been found on Chios and has gone on to conclude from this that the influence cannot have occurred there (incidentally using this Laconian connection to bolster the theory of production at Naukratis, where Laconian has been found). The alternative solution, however, is that a Laconian vase-painter actually moved to Chios. Since we have a Laconian shape with its distinctive decorative scheme and Laconian iconography and style, this is not perhaps such an unlikely scenario. The Sirens Painter’s impact was immediate and discernible, but sadly the quality of his painting was not maintained by his local pupils. Finally, we return to the issue that was already mentioned at the beginning of this paper – where was Chian pottery made? We are beginning to see the strength of the possibility that migrant Chians went to the region of Thrace, perhaps to the Chian colony at Maroneia, and from there served the markets of Ainos and Thasos, influencing local potters and adopting local shapes.71 In counterpoint to this, there has been a reluctance to the complete abandonment of the idea of Chian potters working at Naukratis in Egypt, even following the confirmation by scientific analysis that the pottery was all made with Chian clay. The local clay was poor by Greek standards, but some potters do seem to have tried to work it in the Greek manner and produce rough imitations of North and South Ionian wares, as is laid out in this volume by Schlotzhauer and Villing. They did not attempt anything like Chian. In extremis, then, one has to postulate the import of raw Chian clay. How might we hope to detect such a situation? Three considerations come to mind: the adoption of native shapes; the use of native languages and scripts; and the reflection of native customs and people through the iconography. For native shapes, one thinks of the situlae studied by Sabine Weber, only one fragment of which seems to have been found at Naukratis.72 Clay analysis points to Rhodes as the source of clay, but should we think of them as having been made at Naukratis with imported clay? For native scripts, one thinks of Herbert Cahn’s extraordinary amphora with the cartouche of Apries, a new

Figure 21 Chalice: cock, BM GR 1888.6-1.549

fragment of which was found by Don Bailey in the Petrie Museum with the reported provenience of Thebes (Bailey Figs 1–5).73 Clay analysis points to Northern Ionia, but was it made at Naukratis with imported clay?74 The idea that iconography might indicate local knowledge is more complex, because it could be reported and created (however closely or loosely we would never know) at home. In any case, there is nothing inexplicable or even really vaguely Egyptian about the iconography of Chian vases found at Naukratis, unlike the pottery produced by other East Greeks, especially some of the material from Tel Defenneh,75 Karnak,76 or even perhaps the Naukratis fragment showing an Ethiopian with curly hair and African features (Cover illustration).77 To conclude, there seems nothing in Chian painting and potting to suggest that any of the finds from Naukratis (or elsewhere) were actually made in Egypt. The existence of so many painted dedications on Chian vases from Naukratis has been commented on often (e.g. Johnston Fig. 9). Such pieces, however, do not point to local production but rather to the sophistication of the trading mechanisms of Chian potters that enabled them to take commissions from customers, not only at home on Chios, but also abroad.78 On Aigina they catered for prosperous traders, it would seem (e.g. Johnston Fig. 7); at Naukratis they served not only traders but also perhaps one of the famous ladies of Aphrodite, Aigyptis.79 Furthermore, Chiot pottery is not the only fabric found at Naukratis with pre-firing dedications, for we should not forget the fragments of a large North Ionian rimless bowl with a long painted dedication in added white inside that names Aphrodite in Naukratis.80 In the end, therefore, although we cannot prove absolutely that Chian potters did not work with their own clay at Naukratis, there no longer seems a single persuasive argument to support such a hypothesis. Illustration credits

Figs 8, 10(right fragment) Mus. Arch. and Anth. Cambridge; Fig. 14 Ashmolean Museum, Oxford; Fig. 20 (right fragment) Fitzwilliam Museum, Cambridge; all others the British Museum.

Notes 1 2 3

4 5 6

Smith 1886, 51-3. Gardner 1888, 38-9. Kourouniotis 1915, 64-93; Kourouniotis 1916, 190-215. Kourouniotis’ unfinished excavations were continued by Winifred Lamb for one season in 1933: Lamb 1934/5, 138-64. Price 1924, 205. Lemos 1991, with earlier bibliography. See also the following articles: Lemos 1986, 233-49; Lemos 1999/2000, 11-50; Lemos 2000, 380-1 and 384-5. In addition, note Schauss 1996, 30-42. Lemos 1991, 163-75: she does not in fact use the word ‘Style’ here, but simply calls them Black-figure Chalices. Naukratis: Greek Diversity in Egypt | 131

Williams 7 8

9 10

11 12 13 14 15 16 17 18 19 20 21

22 23 24 25 26 27 28 29

30

31 32

33 34 35 36 37

38 39 40 41

Boardman 1998b, 145. Two pieces were isolated by R.M. Cook, Cook 1949, 154-5, see Lemos 1991, nos 264, 273; see also Boardman 1967, 149 with n. 5. Lemos adds a chalice from the Samian Heraion, Lemos 1991, no. 247. For the new classification system see Kerschner and Schlotzhauer 2005, 1-56. Lemos 1991, 67-70. Lemos 1991, 71-3. Lemos 1991, no. 176 (BM GR 1924.12-1.421) is attributed to the painter himself. Note that the dinos fragment, Lemos 1991, no. 281 (BM GR 1886.4-1.1122), which she places in the larger group, is not Chian. A second non-Chian dinos is Lemos 1991, no. 280 (BM GR 1886.4-1.1003), to which belongs Brussels inv. A 1769 (CVA Brussels pl. 3.10, Belgium 105). Lemos 1991, no. 140 (BM GR 1888.6-1.473a), interior decoration not noted. Lemos 1991, 73-5. Price 1924, 216-7, had thought of the painter of the Würzburg chalices and the Aphrodite bowl as one and the same. Lemos 1991, no. 211 (BM GR 1886.4-1.1078), interior decoration not noted. Cook 1949, 158 n. 12, with pl. 41 b. Acropolis fragment: van Buren 1926, 16 and 184 no. 7, figs 49-50. See Lemos 1991, 24; the Naukratis fragments listed as nos. 365-7. Hermary et al. 1999, 62, fig. on p. 61. Sardis vessel: Greenewalt 1971, 29-46. Samos vessel: Buschor 1951, 32-41, pl. 8. Lemos 1991, no. 365 (BM GR 1888.6-1.496a-c). The joining fragment is BM GR 1924.12-1.178 (Lemos 1991, no. 390). Lemos 1991, no. 366 (BM GR 1888.6-1.496d). Lemos 1991, 75. This is Lemos 1991, no. 253 (BM GR 1888.6-1.460k and 1924.12-1.84). For such plastic additions see Boardman 1976, especially 287-8. The unpublished Cahn fr. is Basel, Cahn HC 476. They are also sometimes shown attached to the side of skyphoi , kantharoi and rhyta, see Boardman 1976, 289; add that on the rhyton shown on the Kleophrades Painter’s psykter, Princeton y1989-69, Guy 1990, 46-7. For a fragment from a skyphos see Getty 86 AE 585. Lemos 1991, no. 480 (BM GR 1924.12-1.186). Lemos 1991, no. 391 (BM GR 1924.12-1.203). Lemos 1991, no. 1254 (Brussels inv. A 1788: CVA Brussels pl. 3.7, Belgium 105 – upside-down). Aphrodite bowl: Lemos 1991, no. 252 (BM GR 1888.6-1.456). Joining fragment: BM GR 1924.12-1.418. Lemos 1991, 243-4, nos 252-8. There is also a group of fragments from later, smaller lekanai, all found at Rizari (Lemos 1991, 244-5, nos 26972), that have mock rivets. Lemos 1991, no. 154 (BM GR 1888.6-1.475g = GR 1924.12-1.119). Cf. Lemos 1991, nos 151 (BM GR 1888.6-1.475d), 170 (BM GR 1924.121.115) and 222 (Cambridge N 48 and 49: CVA Cambridge 2, pl. 17, nos 18 and 20, GB 496). One group of joining fragments is BM GR 1888.6-1.499g and b, 1924.12-1.21 and 1888.6-1.443. A second group is BM GR 1888.61.499e, 1924.12-1.20 and 1888.6-1.459, to which also belong 1888.61.460 e, f and g. Compare changes elsewhere – e.g. from Wild Goat into Fikellura, see Schlotzhauer (forthcoming b). For the date note the Chian fragment from the Lion and Sphinx workshop found in the destruction deposit at Old Smyrna – Lemos 1991, no. 1457. Lemos 1991, no. 1272 (Istanbul, Arch. Mus.). Lemos 1991, no. 1449 (Cambridge, Mus. Arch. and Anth. NA 103). This may also have been the case on the fragment from a stemmed skyphos bowl, Lemos 1991, no. 1278 (from Chios town). Cf. also perhaps, if it is Chian, a fragment in Leiden: Prins de Jong 1925, pl. 1.9 (not in Lemos 1991). Lemos 1991, no. 1338 (BM GR 1886.4-1.998). Lemos 1991, no. 1419, pls 182-3. Plate fragments: Cambridge, Mus. Arch and Anth NA 98 (Lemos 1991, no. 1411) and BM GR 1924.12-1.16 (Lemos 1991, no. 1406). Ring vase: Lemos 1991, no. 1440 (BM GR 1888.6-1.763). Lemos 1991, no. 552 (BM GR 1888.6-1.465a,b,d,e etc.). Boston 88.830.7 (Lemos 1991, no. 582) joins BM GR 1888.6-1.465a; Univ. Coll. London 751 (Lemos 1991, no. 682) joins BM GR 1888.6-1.465b (Lemos 1991, pl. 70 row 3). Lemos 1991, no. 439 (BM GR 1888.6-1.466d,f,e). Lemos 1991, no. 458 (BM GR 1888.6-1.464). Note that the Chios plate (Lemos 1991, no. 684) and the Berezan fragments (Lemos 1991, no. 799) still have filling ornaments and should be early. Boardman 1956, 55-62.

132 | Naukratis: Greek Diversity in Egypt

42 Lemos 1991, no. 516 (BM GR 1924.12-1.343 – subject wrongly identified by Lemos) and Lemos 1991, no. 796 (Aigina, Kolonna fr.). 43 Bonn 697.7: Lemos 1991, no. 788; Piekarski 2001b, no. A 11, pl. 3.3. 44 On the iconography see most recently Tempesta 1998. 45 Lemos 1991, 111-2. 46 Lemos 1991, nos. 800 (Istanbul inv. no. 8904) and 801. See also Lemos 2000, 384-5. 47 Red lion inside: BM GR 1888.6-1.790. 48 E.g. Lemos 1991, no. 972. 49 Lemos 1991, no. 964. 50 St. Petersburg, Hermitage, Berezan B 465: Solovyov 2005, 65 no. 95. 51 BM GR 1924.12-1.366; 1924.12-1.367; 1888.6-1.548p (=Lemos 1991, no. 1495). 52 Lemos 1991, 173-5. Add the twin reed-handled kantharos from Berezan: Lemos 1991, no. 1635. 53 Lemos 1991, nos 1603, 1597, 1598, and 1614. 54 I am very grateful to Dmitry Chistov, of the Hermitage Museum and the current director of the Berezan excavations, for bringing this new find to my attention. 55 Lemos 1991, nos 1600 and 1601, 1602, 1611, 1605. 56 Tocra chalices: Lemos 1991, nos 927 and 928, and 1600 and 1601. 57 Lemos 1991, nos 1617-1619 and 1625-1634. 58 Lemos 1991, nos 1458-1460 (cups), 1461-1465 (chalices), 1466 (bowl) and 1467 (fragment). 59 The current whereabouts of the third fragment in her list, Lemos 1991, no. 1459, once in Berlin (Lane 1933/4, 186 fig. 26), is not known: it might well have come from one of the two other cups. 60 Cf. Lemos 1991, nos 1331, 1351, 1302, 1296 and 1279. 61 I am very grateful to Yulia Ilyina of the State Hermitage in St Petersburg for confirming that though the interior is slipped, there is no white slip under the black. She also notes that the pale clay makes one think of Corinthian. 62 Oxford G.133.2 and 6: Lemos 1991, no. 1458. BM GR 1886.4-1.1283: Lemos 1991, no. 1460. 63 The class was noted in Price 1924 (p. 183 with fig. 59 on p. 215 = BM GR 1924.12-1.188; cf. Boardman 1956, 61 with n 9). Add: BM GR 1924.12-1.169,170,171,172, and 176; BM GR 1888.6-1.561u (= GR 1924.12-1.456 – elaborate meander inside rim); BM GR 1965.930.400a; BM GR 1965.9-30.400 (complete profile); BM GR 1910.222.57 (foot fragment). 64 Boardman 1956, 60-1; Boardman 1986, 254; Lemos 1991, 154-62. 65 BM GR 1888.6-1.550a: Lemos 1991, no. 1463 66 BM GR 1924.12-1.206 (Lemos 1991, no. 1462), 204 (Lemos 1991, no. 1464), and 342 (not in Lemos 1991). BM GR 1924.12-1.205 should probably come from another chalice by this painter; so, too, perhaps a fragment in Oxford, 1925.608e (Lemos 1991, no. 162), which preserves traces of a horseman with a spear. 67 Cambridge, Fitzwilliam Mus. GR 97.1894: Lemos 1991, no. 1461. BM GR 1888.6-1.550b (Lemos 1991, no. 1465) joins, as does GR 1924.121.352 (not in Lemos 1991). 68 See Pipili 1987, 71-3; also 76 and 41. Winged demons do occur on other East Greek vases, but they are very different – see Schaus 1986, 275 ns 71-7. Add the many more Milesian examples excavated in recent years at Miletos, and the fragment in the collection of William Suddaby – Padgett 2003, 276-7 no. 69 (creature wrongly identified). 69 Stibbe 1972, 88-89 (the Oxford fragment is illustrated on pl. 60.2-3). 70 BM GR 1888.6-1.549. 71 See Lemos 1991, 209-22; and Lemos 1992, 157-73. 72 See Sabine Weber, this volume. 73 See Donald Bailey, this volume. 74 For the question of clay sources, see Mommsen and Kerschner, this volume, n.7. 75 The hawk on a nb (Neb) basket (as identified by Petrie) – BM GR 1888.2-8.3: Weber, this volume (Weber Fig. 16); CVA British Museum 8, pl. 2.2 (GB 597) and 9.14 (GB 604). See further Schlotzhauer and Weber 2005, 88-91. 76 Opet’s boat (?): Oxford 1924.64 (from Karnak) – CVA Oxford 2, pl. 10, nos 24 a-f, GB 401; Boardman 1958, 4-12; Boardman 1999a, 138 fig. 162. 77 African: BM GR 1886.4-1.1282 (BM Cat Vases B 102.33); Lemos 1991, no. 1657. This seems more North Ionian than Chian. 78 At home on Chios there are even examples of painted dedications on coarse wares (Lamb 1934/5, 161 and fig. 13 on p. 162) – the island of Homer enjoyed its literacy! 79 See Williams 1983a, 183-6, for Aigyptis see 185; see also Boardman 1986, 254 and 257 fig. 5. 80 BM GR 1888.6-1.531: Gardner 1888, 64-5 and pl. 21 (inscr. no. 768); Möller 2000a, 178 no. 4.

Some Observations on Milesian Pottery Udo Schlotzhauer with contributions by P. Herrmann (†) and S.Weber Abstract Compared with the immense importance attributed to the Archaic polis of Miletos during the Archaic period, only very little information about this epoch, gathered from archaeological excavations in Miletos itself, has been available until now. Questions regarding the numerous colonies and emporia of the city, for example, are therefore somewhat difficult to answer. Due to mass production and the virtually indestructible nature of pottery, this category of finds in particular has been at the centre of attention when questions regarding dating, trade and artistry have arisen. The following contribution intends to portray the current state of archaeological research on Miletos as concisely as possible, with pottery as its main focus.* Introduction Miletos was one of the most important Greek centres in the East Aegean during the Archaic period. In addition to its extensive colonial activities in the Black Sea area,1 the polis, together with other Greek city states, also operated the port of trade Naukratis at the Canopic mouth of the Nile.2 However, archaeological evidence that would confirm the role ascribed to Miletos at Naukratis by literary tradition has so far failed to materialise. This gap in the archaeological record was often overrated in the past, with Miletos’ role in Naukratis even being questioned altogether.3 Sir John Boardman, however, had already realised that this discrepancy between the literary and archaeological traditions must be attributed foremost to the lack of knowledge about pottery from the East Aegean: ‘When we know more, it seems likely that the Milesian share will be recognised at the expense of the Rhodian.’4 Thus, the state of research on pottery production at Miletos is of immediate relevance to research in Naukratis.5 Until only a few years ago, knowledge of Milesian pottery and the pottery of the entire East Aegean was extremely meagre. Yet it is precisely pottery that is at the heart of archaeological research. In contrast to other categories of finds, their statistical components are hardly affected by chance preservation or through displacement and reuse in other contexts. In addition, pottery is found in large quantities in all locations, as it is a trade good or commodity, on the one hand, and, on the other, sometimes even an artistic work itself, the canvas for vasepainting and coroplastic ornamentation. Thereby, pottery contributes significantly to our understanding of trade, the arts and crafts and cultural transfer during any particular period. Furthermore, inscriptions are occasionally found on pottery in sanctuaries, and they aid in the understanding of the activities of individual persons in the past. In recent years there have been rapid advances in research on Archaic pottery from the East Aegean.6 Although extensive excavations have been carried out for many years in various locations – including in Archaic layers – it is a well-known fact

that publications do not keep pace with the finds.7 Theories based on individual archaeological results should, therefore, for the time being, not be generalised, as they are far from conclusive for many regions and locations. The increase in fundamental information about pottery from the East Aegean is, however, not only the result of excavations. Important advances, particularly in the determination of the origin of individual pottery groups and types, have also been made thanks to cooperation with the natural sciences in the form of archaeometry.8 Thus, it is now possible to determine a surprising diversity in production within a few ceramic production centres.9 Scientific studies have their place in many contributions in this volume10 or are the actual subject of the work.11 A third aspect of the progress in research on the ceramic production in the East Aegean lies in the renewal of outdated terminological systems of classification. The development and establishment of new systems of classification and terminology are closely linked to the innovations mentioned earlier, that are due to the increase in finds and the results from scientific studies.12 Communicating the new information and achieving an adequate conformity across the results is essentially only possible with the consistent dismissal of obsolete systems, which are based on fundamental misinterpretations. This moreover prevents the danger of outdated research opinions being carried on within systems that have merely been modified. At the same time, however, the possibility of a new terminology has to be approached with particular care. Only where it seems inevitable and where sufficient arguments for its use exist, should this path be chosen. A further danger lies in depriving categories of material of their history of research, or in establishing parallel, rival strands of research. With all this in mind, devising a new classification system for the East Aegean, which moreover makes the region comparable internally as well as with other Greek landscapes, seemed an inevitable conclusion; this system has already been introduced and explained in detail elsewhere, and will be applied also in following (cf. also the overview chart Fig. 10).13 Miletos and its hinterland While previous research had always regarded Samos as the prominent centre of the arts in southern Ionia, the results of recent excavations in Miletos show that this site was in no way inferior to its neighbouring island.14 This can be concluded from a large number of objects, which were discovered in the settlement, the sanctuaries and the necropoleis of Miletos in recent years.15 However, very little understanding of Milesian pottery production, particularly of painted pottery, exists as yet, even though the main focus of excavations in the city and its hinterland over the last one and a half decades has been on the Archaic period.16 The following pages are intended to provide an overview of Naukratis: Greek Diversity in Egypt | 133

Schlotzhauer Milesian pottery, as far as it is understood at present, as well as add some new results from the German excavations in Miletos. The polis of Miletos During the past 15 years the most important findspots for Archaic pottery within Miletos have been the well known sanctuary of Athena in the centre of the ancient city, the settlement and Artemis temple on the Kalabaktepe, and the Aphrodite sanctuary on the Zeytintepe.17 Intensive excavation activities took place in the settlement at the southern edge of the walled Archaic city, at the Kalabaktepe, under the direction of V. von Graeve between 1986 and 1995.18 Here it was possible to establish a stratigraphic sequence for the settlement layers, especially on the southern slope of the Kalabaktepe,19 through which important information regarding the relative and absolute chronology for the large quantity of associated pottery could be gained. Excavations on a smaller scale by M. Kerschner, who was able to identify the temple of Artemis Chitone on the eastern hill of the Kalabaktepe, yielded further stratigraphically relevant pottery.20 A further focus of recent excavation activity has been the Zeytintepe, a hill next to the gates of the city in the area of an ancient settlement called Oikous.21 The sanctuary of Aphrodite of Oikous was discovered here in 1990 and excavated subsequently. Thechora of Miletos22 Three important findspots for Archaic material are to be noted in the chora of Miletos.23 Firstly, Assesos, presumably a Milesian frurion at the Eastern edge of Milesia.24 In a rather restricted trial trench, pottery was found, mainly of Archaic date, which essentially resembled that found in Miletos. The finds are being published by G. Kalaitzoglou.25 It is tempting to assume that this may be the sanctuary of Athena at Assesos mentioned by Herodotus (1.19) and other ancient authors.26 However, the topography of this site has not yet been clarified to an extent that would permit its definitive identification. The proportion of fine ceramics, particularly that of painted vases, is relatively large, so that these can hardly be classified as settlement finds. The absence, however, of votive inscriptions in general as well as upon the pottery itself, otherwise a common occurrence in Milesian sanctuaries, argues against these finds being sanctuary pottery. The question must thus remain open for the time being. Another important sanctuary in the chora, connected to the polis of Miletos by a prominent processional road, is the widely renowned oracular temple of Apollo at Didyma.27 Despite many years of excavation, only very little pottery pre-dating the 5th century bc was found at this location.28 Only in recent years finds from the Archaic period have been accumulating. Thanks to the archaeologists excavating and studying the material in Didyma, the discovery of one specific find from this site can be placed in a wider context here.29 Further finds, some Archaic, have been made along the processional road mentioned earlier.30 An intensive survey was also carried out in the remaining areas of the chora of Miletos.31 It did not, however, yield any noteworthy information relating to the Archaic chora or pottery from this period.32

134 | Naukratis: Greek Diversity in Egypt

The pottery of Miletos: state of research on painting style and types of vessels33 A new classification system has been developed for pottery of the East Aegean, which allows the integration of results from recent excavations, the natural sciences and an expanded material basis.34 The system distinguishes undecorated as well as decorated fine and everyday pottery according to larger and smaller phases of development within the production site or larger regions. This terminological convention has the advantage that the spatial and chronological placement of pottery can be readily determined by means of a standard selection of terms and code structure. The consistent application of this model will enable developments in different regions to be correlated in a generally intelligible way. Thus, developments in a workshop in Aiolis can be easily and intelligibly compared with developments in northern Ionia, southern Ionia or eastern Doris as well as Caria and Lydia. Until now correlations such as these were only achievable by a specialist who was immediately able to recognize that, for instance, the northern Ionian so-called ‘Tübingen group’35 and the Chian ‘Comast group’,36 the groups S and R, as well as the ‘Altenburg painter’ of the so-called Fikellura style37 from Miletos represent parallel developments from the later second third of the 6th century bc. The geographical localisation as well as the stylistic classification and the chronological horizon of the categories of Knipovich group38 and Enmann group,39 too, cannot simply be presumed. The new classification makes it easy to recognise, by means of the second, chronological component of the code, ‘A II’, that these are stylistically (often also chronologically) parallel developments, whether in northern Ionia (NiA II), on the island of Chios (ChiA II) or in south Ionian Miletos (MileA II). This system is an open model, which can be adapted to conditions in different regions or parts of regions, in poleis and as far as workshops. The course of development is not necessarily parallel everywhere, nor are developments structured in the same way. A phase may, for example, be absent in one particular landscape or smaller region, while another workshop elsewhere could, in turn, feature a phase more than its neighbour. A developmental structure has yet to be drawn up for many locations and periods, as research on ceramics in the East Aegean has yet to reach the stage of that on the Archaic and Classical periods of Athens. In Miletos, the division between the two main phases of Milesian vase-painting is the change from the Wild Goat (henceforth ‘WG’) style, MileA I of the 7th and beginning of the 6th century bc, to the so-called Fikellura style (MileA II), which, according to my opinion, emerged during the first third of the 6th century bc. Figural and ornamental friezes, in which large portions of the figural body (e.g. the heads – except for birds – wings and the belly) were reserved and the ornaments (e.g. buds and blossoms, as well as dividing bands) were outlined, had always been grouped together under the WG style. By contrast, the outlining technique became unusual in the Fikellura style (MileA II).40 Until now a clear definition of both of these two main phases of Milesian painting has not been established in any studies of these styles, which has often led to different ascriptions of individual examples. Unlike other Greek regions, Milesian painters retained their reserving technique even during the phase of black-figured vase-painting and did not switch to the incising technique. Therefore, in the case of

Some Observations on Milesian Pottery Milesian products, it is not always easy to distinguish between the two main phases AI and AII, particularly in cases of small fragments. For the incising technique is regarded as an indicator for the clear division between the earlier and the later stylistic phases. Only recently was an attempt made to establish a definition that can now be used as a starting point for a better differentiation between the phases of Milesian vase-painting.41 A comprehensive proposal so far exists only for the early phase, the Archaic I period (MileA I), of Milesian painted pottery. It suggests a division into four sub-phases, which approximately span the time from the second quarter of the 7th to the first quarter of the 6th century bc, based on the development of shapes and decorations.42 This kind of internal division does not as yet exist for the second period of Milesian vase-painting, MileA II.43 Ceramics from the Geometric44 and Classical periods will also need to be classified accordingly in the future. The early period of Archaic vase-painting in Miletos: MileA I The state of publications on vase-painting of the phase MileA I, which essentially encompasses the styles known so far as ‘Subgeometric’ and ‘Wild Goat’ style,45 is extremely disparate.46 Noticeable in Milesian vase-painting of MileA I is the limited range of variation of ornaments and figures. Simplification and a limited figural repertoire occur in those phases in particular, in which there is an increase in production and standardisation (phases MileA Ic-d).47 Whereas sphinxes, griffins, boars, bulls, lions, panthers, foxes, hares and dogs were still represented more frequently in the first two phases MileA Ia/b, in MileA Ic/d, almost exclusively ibexes, deer and geese are depicted. Representations of humans, too, are very rare and have only been documented for the late phase MileA Id, several times in form of a head protome48 and once as a horseman on a sherd from Daskyleion.49 Other representations of humans (presumably mythical or divine beings), in contrast, can only be cited for the earliest phase MileA Ia.50 The newly proposed development is, according to the current state of research, divided further into four phases. Phase MileA Ia51 (c. 670–650 bc) is characterised as being the transition from the Geometric to the Early Archaic period. The distinction between it and the Geometric period is best manifested by the presence of several figures in the central metope field. Dot-filled volutes appear for the first time; in the following phase they develop into the virtually mandatory central motif of the volute blossom, primarily in the shoulderfrieze on jugs. The lowest zone of decoration consists only of a row of vertical strokes or a wreath of rays. Sometimes the figures still have filled-in faces as in late Geometric, but usually they are already reserved. In phase MileA Ia painted pottery shapes are limited to oinochoai with a round mouth, kraters, amphorae and cups with everted rim. In phase MileA Ib52 (c. 650–630 bc) the decoration of vessels changes to include friezes encircling the whole vase. Aside from the row of strokes and the wreath of rays, lotus-and-bud chains emerge and soon become canonical in the zone above the foot. Ornamental volutes are sometimes found on the underside of dinoi or in the centre of plates and bowls. Filling ornaments and fauna attain their richest diversity during this phase. So far it has been possible to detect phase MileA Ib on the following shapes: kraters, dinoi, lids, cups with everted rim, plates, and trefoil and

round-mouthed oinochoai. MileA Ic53 (c. 630–610 bc) is characterised in particular by the fact that the hanging and standing triangles and the halfrosettes, which had already emerged in MileA Ib, are now linked with animals in the friezes and, thus, slow down its movement. The number of friezes on a vessel’s body also increases in some examples, so that the body of the Lévy oinochoe,54 for instance, is covered by five animal friezes. There is a decrease in the variety of figures and filling ornaments, the heights of the friezes are reduced and friezes become more standardised. Likewise, the lotus-and-bud chain in the zone just above the foot is only rarely replaced by a wreath of rays. The repertoire of painted shapes in MileA Ic comprises oinochoai with a trefoil mouth, kraters, dinoi, cups with everted rim, lids, plates and bowls. The final phase MileA Id55 (c. 610–580 bc) leads to a simplification in decoration with elongated animals and ornaments of exaggerated size. It is an economising development, in which Milesian vase-painters fill out a frieze with only few animals or ornaments, an obvious result of mass production. This characteristic feature of phase MileA Id is easily noted. On the other hand, a superior figural painting style persists, which continues to feature numerous figures, often interacting, as well as complicated ornamental friezes. Thus far this it has not been possible yet to clearly differentiate this feature in all its aspects from the preceding phase MileA Ic. The fact that this superior trend continues to exist even in the last phase of MileA Id, alongside mass goods affected by simplification, is indicated by vessels that represent the transition to the late phase MileA II. Phase MileA Id is also the earliest Milesian phase, the ceramic produce of which is found at Naukratis. Amongst these is the so-called Polemarchos-krater (Schlotzhauer and Villing Fig. 19), which scientific analyses have identified as a product of the Kalabaktepe workshops.56 Further Milesian finds, including those of phase MileA Id, were made in Egypt, but outside of Naukratis. One example57 is the amphora from Thebes/Gurna.58 The late period of Archaic vase-painting in Miletos: MileA II Vessels which combine both stylistic stages MileA I and II, socalled bilinguals (cat. no. 1, Fig. 1), display in their MileA II friezes features of the first phase of the new period, MileA IIa.59 They should therefore be consistently attributed to MileA IIa, according to the archaeological principle ‘the youngest element dates the find’, even if elements of the previous stylistic phase MileA Id sometimes still prevail. Surprisingly, these early examples already feature human representations with a narrative potential alongside conservative MileA I-style animal friezes with animal-fighting scenes or rows of animals, all on the same vessel. The further development has not yet been worked out. However, there is now a much larger basis of material known from excavations at Miletos, that will enable us to go beyond the limits of the previously recognised groups,60 painters61 and repertoire of shapes.62 The resulting increased diversity in known designs in the period MileA II is easily explained by the fact that our previous knowledge was merely based on the pottery that was exported from Miletos.63 Only certain vase shapes decorated in this style and only certain workshops and painters, however, appear to be represented in export markets.64 Naukratis: Greek Diversity in Egypt | 135

Schlotzhauer

Figure 1 MileA IIa, 'Aphrodite-Cup', cup with everted rim: Miletos; temenos of Aphrodite (cat. no. 1)

Figure 2 MileA II, cup with everted rim: Miletos; temenos of Aphrodite (cat. no. 2)

136 | Naukratis: Greek Diversity in Egypt

Some Observations on Milesian Pottery

Figure 3 MileA II, amphora:Tell Defenneh, fort (BM GR 1888.2-8.46a; sample Defe 11: group D) (cat. no. 5)

Others vases, particularly those of high quality, seem to have remained within the most immediate geographical surroundings. Vessels in this style have now been discovered in Miletos, and they feature an abundance of representations of myths, gods, everyday life and religious festivals, as well as names. This may be exemplified by a piece that was recently discovered in the sanctuary of Aphrodite at Miletos (cat. no. 2, Fig. 2).65 It is presumably the goddess herself who appears as the mistress of nature in the centre of the finely made vessel, a cup with everted rim. She holds a panther in her left hand.66 Floral motifs complete the reference to nature. The ‘bent-knee’ running pose of the goddess is so far unique and distinguishes her from the static, frontally orientated and winged mistress of the animals, who usually also holds an animal in both hands. Only a few examples of unusual vase forms or unparalleled decoration from phase MileA II reached the Milesian colonies and emporia. Mugs with trefoil mouth67 and plastic headkantharoi or -mugs68 should be recalled here.69 A prime example for this phenomenon is a long-known vase from Egypt that had been thought to be unique but that can now be associated with a Milesian workshop. The first indication of its Milesian origin was provided by the scientific determination of origin via NAA in 2004. This was later confirmed by new finds in Miletos in the summer of 2005, thus confirming the reliability of the scientific procedure. This piece will be discussed in some detail in the following paragraphs.

An amphora from Tell Defenneh S. Weber (NAA carried out by H. Mommsen and A. Schwedt, Bonn) The amphora (cat. no. 3, Fig. 3) was found at an undesignated place in the ‘kasr’ during the excavation at Tell Defenneh, a site in the eastern Nile delta, which was conducted by William M. Flinders Petrie in 1886.70 The vessel is decorated with a row of strokes on the lip and a triple cable on the neck, which is surmounted, on side A, by a simple gear pattern and, on side B, by chevrons. On the shoulder is a small lotus-frieze and a broad frieze of water birds facing to the right. The belly is painted with a meander-and-square band; below that is a frieze of animals (seven goats and a deer are preserved), followed below by a hook meander and enclosed palmettes; and finally below that a narrow band with a simple gear pattern and a lotus band. Filling ornaments consist of simple four-dot-rosettes in the figural friezes and of very hastily painted small lotus flowers between the feet of the animals. The amphora, together with seven other examples decorated in the so-called ‘Fikellura style’, was assigned by R.M. Cook to his group C, named after our amphora ‘group of B.M. B 117’.71 Cook suggested a date not long after 550 bc. It is a very heterogeneous group related to his group B or ‘Lion group’, but the rendering of the animals of our amphora is clumsier than that on vases of this group. Thus, this amphora stands stylistically somewhat apart from the other vases painted in this style. This led to speculations about the amphora’s place of origin. The possibility that Fikellura (= MileA II),72 like the Orientalizing animal frieze style, the so-called Wild Goat style (= EA I),73 was produced in different cities in the East Greek area with Samos or Rhodes as the main centres, was considered by R.M. Cook74 and also favoured by E. Walter-Karydi.75 For the amphora B 117, however, Walter-Karydi tentatively proposed an Ephesian origin.76 As a characteristic feature of Ephesian manufacture she cited the band of black and white squares,77 a decorative element, however, not painted on the amphora B 117. Walter-Karydi, of course, based her attributions upon stylistic studies long before clay analyses were carried out. In the last 20 years analyses have shown that patterns of the chemical elements in the clay paste of vases in the Fikellura style conform to Milesian clay beds.78 In consequence, in 1986 G.P. Schaus rejected the hypothesis of different production centres for Fikellura vases, basing his arguments upon the clay analyses and his stylistic study of two major Fikellura painters. He concluded that vases in the Fikellura style must have been made in one city only: Miletos.79 Nevertheless, we still have the Ephesian origin for the amphora from Tell Defenneh once suggested by WalterKarydi. New clay analyses could be the key to solving this problem. In fact, NAA analyses of three MileA II (= Fikellura) sherds found in Ephesos, which are close in style to two of the vases from Ephesos listed by Walter-Karydi and assumed by her to be Ephesian, have been carried out by M. Kerschner and H. Mommsen.80 The analyses revealed that the vases belong to the chemical provenance groups A and D. Group A is definitely a Milesian pattern and belongs to the so-called Kalabaktepe workshops; D can be located in all probability in Miletos as well.81 The NAA analysis of amphora B 117 now carried out by H. Mommsen and A. Schwedt in Bonn, sample number Defe 11, has proven that the clay paste of our vessel belongs to the chemical Naukratis: Greek Diversity in Egypt | 137

Schlotzhauer

Figure 4 MileA II, mug: Miletos; temenos of Aphrodite (cat. no. 4)

Figure 5 MileA II, cup: Miletos; temenos of Aphrodite (cat. no. 5)

group D.82 With this result we can conclude that at the moment there is no evidence for the production of vases belonging to the MileA II period ( = Fikellura) in Ephesos. Vases in this style were produced in different workshops in Miletos and its chora. In addition local imitations can be detected in some of its colonies and in Naukratis.83 The results attained by S. Weber and H. Mommsen on the amphora from Tell Defenneh have thus led to quite surprising conclusions that stand in a long line of similar new discoveries having been made through scientific clay analyses. Earlier investigations by R.E. Jones, M.J. Hughes and P. Dupont, as well as more recent work by J.N. Coldstream together with D.J. Liddy and by M. Seifert with Ü. Yalçýn had already provided evidence that the so-called animal frieze and Fikellura styles were produced in Miletos, as were cups with everted rim, transport amphorae and other shapes.84 This evidence and a detailed discussion have been presented by M. Kerschner.85 Augmented by new research initiated by M. Kerschner and H. Mommsen using the NAA method, all this now allows a much clearer and differentiated picture of Milesian pottery to be drawn. Investigations continue as colleagues with pottery finds from different sites join the project of M. Kerschner and H. Mommsen.86 As a result, a multitude of different vessels can now be added to the Milesian groups A and presumably D,87 a 138 | Naukratis: Greek Diversity in Egypt

new overview of archaeometric results on Milesian ceramics is therefore of utmost urgency. In addition to scientific results, new finds from Miletos also provide added support for the stylistic attribution to a Milesian workshop of the amphora from Tell Defenneh, which with its peculiar decoration has stood in isolation thus far. These recent finds show that the creator (or his workshop) of the amphora from Tell Defenneh also decorated other shapes, and that perhaps he/they also potted them. One cup (cat. no. 4, Fig. 4) is decorated with young leaping deer, which clearly displays the same style in painting as the deer and goats on the amphora from Tell Defenneh (Fig. 3). The same painter’s hand can be recognised in the clumsy depiction of the animals with their characteristically exaggerated spots, as well as in the peculiarity of the composition despite the difference in the type of animals. Although filling ornaments are absent on the cup from the sanctuary of Aphrodite near Miletos, nevertheless the same decorative scheme can be recognised as on the amphora. Simple, sometimes double dividing lines border the frieze and the dividing bands, and the painter’s preference for simple separating bands with a gear pattern is more than obvious on both vessels. Furthermore, a third vessel, a bowl (cat. no. 5, Fig. 5), provides the perfect analogy for the depictions of birds on the amphora (Fig. 3), which until now had been unique. Here, too, the similarities are astonishing, in detail as well as in general conception. Although, again, the filling ornaments that decorate the amphora are absent, a simple band with a gear patter divides the bird frieze from the band of flower buds on the cup from Miletos, just as on the amphora. Further observations on painted mugs of the phase MileA II from Miletos The group of painted mugs and their variety of shapes has received little attention thus far. Yet mugs constitute a substantial part of the finds from Miletos and quantitatively are hardly less significant than cups with everted rim (Ionian cups) or hemispherical cups (‘bowls’). The abundance of variations in shape and type is likewise notable. Some differ considerably in the shape of the foot, the rim or the number of handles. For example, there are mugs with a disk base,88 with a ring-base that can be offset (Figs 8, 9)89 or not,90 and mugs with a simple flat base91 (Fig. 7). Likewise, the rim can be shaped differently. Some rims are straight and others flaring.92 In another variety the rim is undulating. In view of the similarity to oinochoai with trefoil mouth, these mugs are designated ‘with trefoil mouth’ (Fig. 9).93 Their mouth, however, is much flatter and has more lobes than the mouth of trefoil-mouthed oinochoai, which have only three pronounced lobes. Recently, a further significant and almost complete example of the type of ‘mug with trefoil mouth’ has been published, from the area of the sanctuary of the Milesian oracular temple of Apollo at Didyma.94 In view of its almost complete state of preservation, it is now apparent that this group of trefoil mugs usually had two handles.95 Other varieties of mugs possess up to three handles, although the mug with only one vertical handle is most common. However, no examples of such mugs have been found thus far in Naukratis. In spite of the unusually large multitude of pottery types from the East Aegean at this site, only South Ionian mugs with a simple flat base and a flaring rim are known from the site. It can be proven that these were produced not just at Miletos but also at Samos, as attested

Some Observations on Milesian Pottery

Figure 6 MileA II, mug: Didyma, sanctuary on the Taxiarchis hill (cat. no. 6)

Figure 7 MileA II, mug with flat base: Miletos; temenos of Aphrodite (cat. no. 7)

Figure 8 MileA II, mug with an offset ring-base: Miletos; temenos of Aphrodite (cat. no. 8)

by their dipinto HRH as well as by scientific analyses of their clay (e.g. samples Nauk 1-3, Schlotzhauer and Villing Figs 14–16).96 By contrast, it had long been assumed that an example of another South Ionian/Milesian group of high quality mugs or kantharoi was represented in Naukratis, namely a Milesian mug or kantharos in the form of a head (sample Nauk 57, group KROP). The long development of this type can be traced in Miletos through a series of new finds.97 However, it has been shown that the unusual stylistic features of the fragment, which until now had been attributed to its date, must be interpreted differently, for the fragment found in Naukratis is presumably from a painted mould-made head-kantharos with additional modelling that was produced in Athens.98 The rich variety of mugs presented here represents only a part of the Milesian production of painted and plastically decorated MileA II mugs of the 6th century bc.99 Mugs of the phases MileG and MileA I from the preceding Geometric and early Archaic periods, especially the painted examples, will be discussed elsewhere.100 Here, mention will be made, however, of a further aspect, which is connected with the category of late painted MileA II cups and which is significant especially in connection with Naukratis. Decorated cups with votive inscriptions in the form of dipinti have recently been discovered at Miletos in unprecedented large numbers.101 Three of these mugs, from different find contexts, are almost identical in form and on the basis of their similar decoration are presumably closely connected in date. Moreover, in three cases (cat. nos 7–9; Figs 7, 8, 9) the name of the dedicator is the same, so that it seems conclusive that the three mugs were ordered by the same person from the same workshop within a short span of time. Precisely this aspect is encountered in Naukratis as well. Here, vessels with the name of the dedicator AIGUPTIS, likewise painted before firing on Chios, arrived at the sanctuary of Aphrodite, most likely have been transported there by a trader acting on a customer’s order.102 A further argument for the theory or a special order to a workshop in Miletos is the essential correspondence in decoration and form observable in two of the four mugs presented here (cat. nos 6–9, Figs 6–9). Of special interest is the fact that the four vessels had already been determined for dedication to specific deities prior to their completion. In fact, the mugs were found in different sanctuaries at Miletos, in the ancient city as well as in the surrounding area, but only one vessel preserves the name of the deity (cat. no. 7, Fig. 7). The fragment of the mug (cat. no. 6, Fig. 6) from the Taxiarchis hill near the oracular sanctuary of Apollo Didymaios has already been published.103 The excavators assume that the hill is the location of one of the historically documented sanctuaries of Didyma.104 In view of the find context there is no doubt that the pottery was meant for the sanctuary. The dedicatory inscription A]NEQHK[E(N) on the mug fragment illustrated here adds certainty to this assumption. Although in this case the deity to whom the dedicator offered the mug cannot be determined, nevertheless the fragment of the Milesian mug from Didyma can be added to a series of votive mugs. A better preserved mug (cat. no. 7, Fig. 7) from the site of the temple of Athena in Miletos corresponds to the mug in Didyma in form and decoration. On this mug as well there is a dipinto in the same place as on the mug from Didyma, between the likewise identical decoration, and this dipinto also attests Naukratis: Greek Diversity in Egypt | 139

Schlotzhauer

Figure 9 MileA II, mug with trefoil mouth and offset ring-base: Miletos; temenos of Aphrodite (cat. no. 9)

that the dedicator already had determined the intended deity prior to ordering it in the workshop. But even more information can be gained from the mug from the temple of Athena. The mug was dedicated to Apollo by a man by the name of )/Alsioj or La/sioj, although it was allegedly found ‘near the temple of Athena’. Unfortunately, not all the documentation of the excavation of P. Hommel in 1957 at the site is accessible, nor did Hommel publish many of the important finds. Therefore, the exact spot of the mug’s discovery cannot be determined with any certainty. However, the fact that the mug was found near the sanctuary of Athena is clearly to be understood from the excavation diary.105 Other, still extant finds from the context AT 57.O.191 range in date from late Geometric to the Classical periods. Dedications to other deities in a sanctuary of one particular deity are attested in other places as well. Thus, in the Milesian colony at Olbia a dedication to Athena was found in the sanctuary of Apollo.106 This might suggest that same cult partnership was present in Miletos’ colony as might be deduced for Miletos from the discovery of the mug. The veneration of one or several deities in a sanctuary dedicated to one specific god is not exceptional. But this is not the only possible solution. It is worth contemplating whether other sanctuaries stood to the east of the temple of Athena. Indeed, a sanctuary of Demeter has been postulated after the 5th century bc in the area close to where the mug was found.107 However, its existence is not attested with any certainty either through architecture or inscriptions. In view of the mug’s good preservation and the sherds’ clean breaks, it does not seem likely to have been brought from the distant sanctuary of Apollo Delphinios near the South Market. Overall, it is thus perhaps most likely that yet another sanctuary of Apollo is to be located in the unexcavated area to the east of the so-called Southeast building108 or in the southeast section,109 about whose epithet, however, as little can be said at present as about the architectural form of the presumed sanctuary.110 And what about the inscription on the mug? Following the author’s rediscovery of the long forgotten mug in the storerooms of the Miletos excavation, P. Herrmann – shortly before his sadly premature death – wrote the following short commentary on the inscription:111 Der Textaufbau ist klar, allein der Name des Dedikanten bleibt problematisch. Nach meinen Notizen führen die Schriftreste am Anfang eher auf LA als auf AL, also )/A?lsioj m[a)ne&q]hken tw)p?[o/l]lwni oder 140 | Naukratis: Greek Diversity in Egypt

)/Alsio/j m[e a)ne&q]hken tw~i? [ 0Apo/l]lwni Für einen Namen )/Alsioj konnte ich keinen Beleg finden. Es gibt in Kos und Kalymna einen Monatsnamen )/Alseioj oder )Alsei=oj, der auf a)/Lsoj zurückgeführt wird. a)/Lsoj selbst wiederum wird etymologisch nach einer unter mehreren Annahmen mit dem Stamm al- ‘nähren’ verbunden, danach auch eine lexikalische Glosse a)/Lsij ‘Wachstum’, die mit dem Wort a)ldai/nw ‘zum Wachsen bringen’ kombiniert wird.112 Von der Bildung her hielte ich diese Namensform also für möglich. Sollte die Lesung aber doch La/sioj lauten, hätte man einen Beleg für diese Namensform in Tegea,113 dazu wohl auch Lasi/aj auf Münzen von Katane und Achaia. La/sioj wird von F. Bechtel114 von dem Adjektiv La/sioj abgeleitet und als ‘mit behaarter Brust’ interpretiert. Ich finde übrigens dasselbe Dilemma bei einem Namen auf Chios: eine Inschrift hat einen Namen )/Alswn.115 Auf Münzen der Alexanderzeit aus Chios ist dagegen ein La/swn bezeugt.116 Beide Belege dürften aber denselben Namen meinen117 – nur: welche Form ist richtig? Man sollte im Falle der Tasse von Milet vielleicht schreiben: )/A?lsioj oder L?a?&sioj, eine sichere Entscheidung scheint mir nicht möglich. Beitrag P. Herrmann † In the year 2005 a further mug (cat. no. 8, Fig. 8) was discovered in the sanctuary of Aphrodite of Oikous, which was located outside the walls of Miletos, to the northwest on the hill known today as ‘Zeytintepe’. It differs in form and decoration from the two preceding mugs and belongs to the group of mugs with incurving ring-base.118 It nevertheless represents the stylistic group MileA II and, hence, was produced during the same period. This is clearly confirmed by the dipinto as well. For, not only does the form of the letters correspond most closely, but the name )/A?lsioj or L?a?&sioj (Fig. 8) – unknown until now – appears again on this mug. The appearance of such a rare name on two almost coeval mugs found at the same site would render it unlikely that two different persons were the dedicators. N. Ehrhardt points out, furthermore, that this person does not use an ethnikon nor a surname, so that there is no possibility of a mix-up in dedicators.119 This new inscription also solves P. Herrmann’s problem of how to reconstruct the inscription: it can now be considered as certain that the inscriptions on all three mugs must derive from one and the same person, and that this person presumably chose the same form with krasis, that is, )/A?lsioj or L?a?&sioj ma)ne&qhken . In the year 2005 yet another find appeared, which elucidates

Some Observations on Milesian Pottery

New classification of East Greek pottery (demonstrated on the example of South Ionia) New Classification

Classification of R. M. Cook 1998

SiG 675 670

Early Orientalising

a 650

650

EWG

b

MWG I

630

SiA I

625

c 610

MWG II

600

d

590 E.g. Bilinguals

a

580

Hiatus? or

570

MWG III? 560

SiA II Fikellura

494

494

SiC Figure 10 Classification system of South Ionian pottery

the final remaining question as to the person behind the three dedicated mugs with dipinto inscriptions. This time the fragment of a trefoil-mouthed mug (cat. no. 9, Fig. 9)120 with eye-decoration121 displays a graffito on the inside of the rim. Here, finally, both of the disputed first letters are preserved. Thus, the name is now confirmed at Miletos in three cases: )/Alsioj. With the graffito, in contrast to the dipinto, we have a presumably spontaneous expression of the same person, who once more dedicated a mug in the sanctuary of Aphrodite near Miletos. It cannot be determined whether or not the person carried out this act before or after offering the other mugs. But it does indeed illustrate the variety of forms and decorations that were in vogue on mugs alone during this short period, the period in which )/Alsioj ordered mugs painted in the style of MileA II in a Milesian workshop and then dedicated them in Milesian sanctuaries.

Catalogue Abbreviations Diam. L. Th. W. NAA

Diameter Length Thickness Width Neutron Activation Analysis

1. Aphrodite-cup, cup with everted rim; MileA IIa (Fig. 1) Sanctuary of Aphrodite at Zeytintepe near Miletos Balat Depot, Inv. Z 01.15.3– Z 01.70.2 – Z 02.20.3 – Z 02.56.1 Ht. as restored 7.7cm; Diam. of rim 17.8cm; Diam. of foot 6.8cm; Th. 0.26–0.34cm. Clay: 7.5 YR 7/4-6 to 2.5 Y 7-6/2; soft consistency; very fine temper, dense, fine black particles, much fine dark mica; surface well smoothed; paint: reddish-black, varying on the outside, in some places ‘lacquer red’, in others (inside) dark reddish-black; applied colour: red; slip: 2.5 Y 8/1 to 5 Y 7/1. Sanctuary of Aphrodite at Zeytintepe. Find context Z 01.15 from Q 01.10: Archaic layer, undisturbed, on the virgin soil (‘Feste Erdschicht direkt auf dem Felsen’). Z 01.70 from Q 01.14: also contained worked marble fragments, therefore post-Archaic layer (‘Schutt der persischen Zerstörung’). Z 02.20 from Q 01.14: continuation of Z 01.75 (= late Archaic dislodged and recently disturbed?). Z 02.56 from Q 01.14: oldest layer, eroded limestone, above the natural depression in the centre of the trench, perhaps the northern extension. Published: Schlotzhauer (forthcoming b); Kerschner and Schlotzhauer 2005, 50-1 figs 50-1.

Naukratis: Greek Diversity in Egypt | 141

Schlotzhauer 2. Cup with everted rim; MileA II (Fig. 2) Sanctuary of Aphrodite at Zeytintepe near Miletos Balat Depot, Inv. Z 04.14.60 – Z 04.21.19 Ht. 4.1cm; Diam. of rim 12.8cm; Diam. of foot 6.2cm; Th. 0.2–0.4cm Clay: 5-7.5 YR 6-7/4-6; firm consistency; very fine temper; fine compact matrix; much fine dark mica; surface: well smoothed; paint: reddishbrown to black, varying, light metallic sheen; applied colour: red; slip: 10 YR 8/3. Sanctuary of Aphrodite at Zeytintepe. Find context Z 04.14 from Q 04.3: firm, light brown, layer of earth mixed with many pieces of limestone and marble; much pottery at the southern edge, increasing in number towards the adjoining find context Z 04.12, a lamp depot. Building debris and rubbish from the sanctuary, Late Archaic period, end of the 6th century bc. Z 04.21 from Q 04.3: sandy, whitish-grey layer with limestone chips and ash; strong concentrations of ash in several places, some with pieces of charcoal; worked marble blocks; lots of bones and building debris: Building debris and rubbish from the sanctuary, Late Archaic period, end of the 6th century bc. Unpublished. 3. Amphora from Tell Defenneh; MileA IIa (Fig. 3) From Tell Defenneh London, British Museum, GR 1888.2-8.46a including GR 1924.12-1.1080 (Vase B 117) Ht. as restored 31.5cm; Diam. of rim 15.6–14.7cm Clay: light reddish brown (2.5 YR 6/3), contains mica, very pale brown slip (10 YR 8/4), contains mica; paint: reddish brown. Lower part of the belly and the foot are missing; one reed of the triple handle on each side is missing; on side B most part of the belly is restored. The restoration renders too squat an impression of the shape. Published: Petrie 1888, pl. 27.3 and 3a; Petrie 1891a, 56 fig. 1; Walters 1893, 92 no. B 117; Cook 1933/4, 8 (C1), 9, 65, 73, fig. 13.9, pl. 4c; Cook 1954, 7 pl. G.B. 574, 1, details of shoulder pl. G.B. 569, 3-4; Walter-Karydi 1973, 137 no. 683, pl. 89; Weber in Schlotzhauer and Weber (forthcoming). 4. Fragment of a mug; MileA II (Fig. 4) Sanctuary of Aphrodite at Zeytintepe near Miletos Balat Depot, Inv. Z 04.23.13 – Z 04.71.11 L. 7.9cm; W. 9.3cm; Th. 0.3–0.65cm Clay: 5-7.5 YR 7/6; firm consistency; fine temper; dense matrix; much fine dark mica; surface: well smoothed; paint: dull reddish ‘lacquer-like’; slip: 2.5 Y 8/2. Sanctuary of Aphrodite at Zeytintepe. Find context Z 04.23 from Q 04.3: large sherds of amphorae and bowl, densely packed: sanctuary rubbish, Late Archaic period, end of the 6th century bc. Z 04.71 from Q 04.1: layer of greyish brown sandy clay, little marble, only few worked pieces: fill of building debris and sanctuary rubbish, Late Archaic period, end of the 6th century bc. Unpublished. 5. Fragment of a cup; MileA II (Fig. 5) Sanctuary of Aphrodite at Zeytintepe near Miletos Balat Depot, Inv. Z 04.7.51 – Z 04.75.102 L. 6.2cm; Diam. of foot 10.1cm; Th. 0.3–0.35cm Clay: 5 YR 7/4; firm consistency; very fine temper; fine dense matrix; much fine dark mica; surface: very well smoothed; paint: brownishblack, dull, dense; applied colour: red; slip: 10 YR 8/3. Sanctuary of Aphrodite at Zeytintepe. Find context Z 04.51 from Q 04.3: light brown, in places whitish earth: sanctuary rubbish, Late Archaic period. Z 04.75 from Q 04.5: loose, medium grey layer of fill with lots of worn pottery, medium-sized to large pieces of limestone and marble: building debris and sanctuary rubbish, Late Archaic period, end of the 6th century bc. Unpublished. 6. Fragment of a mug from Didyma with dipinti-dedication (Fig. 6) Sanctuary on the Taxiarchis near the sanctuary of Apollo at Didyma Didyma Depot, Inv. Ke 00-110 Published: Bumke and Röver 2002, 98 fig. 18, 99 fig. 21. 7. Fragment of a mug with simple flat base;122 MileA II (Fig. 7) Near the sanctuary of Athena at Miletos Balat Depot, Inv. AT 57.O.191.1 Ht. as restored 7.3cm; Diam. of foot 6.8cm; Th. 0.4 - 0.6cm Clay: 7.5 YR 7/6; firm consistency; very fine temper; with fine inclusions, dense clay mass; much fine and coarse dark glimmer; surface: well smoothed; paint: visible strokes, varying reddish-brown to black, light metallic sheen. Near the sanctuary of Athena. Find context 191: Archaic layers. Southeast 142 | Naukratis: Greek Diversity in Egypt

section, north-eastern part. Upper layers disturbed, lower layers mixed: from the Archaic to late Mycenaean periods.123 Unpublished. 8. Fragment of a mug with offset ring-base, type b;124 MileA II (Fig. 8) Sanctuary of Aphrodite at Zeytintepe near Miletos Balat Depot, Inv. Z 05.19.18 Ht. as restored 7.8cm; W. 7.7cm; Diam. of foot 6.6cm; Th. 0.25–0.6cm Clay: 5-7.5 YR 7/6; firm consistency; fine temper; fine dense matrix; much fine and some coarse dark mica; surface: well smoothed; paint: varying black to brown, dull sheen, visible brush strokes. Sanctuary of Aphrodite at Zeytintepe. Find context Z 05.19 from Q 05.3: grey earth with much pottery and building debris: rubbish pit, Late Archaic, end of the 6th century bc. Unpublished. 9. Fragment of a mug with trefoil mouth125 and offset ring-base, type b126 MileA II (Fig. 9) Sanctuary of Aphrodite at Zeytintepe near Miletos Balat Depot, Inv. Z 05.20.36 – Z 05.88.100 – Z 05.106.46 H. 13.2cm; Diam. of rim c. 16.5cm; Diam. of foot 14cm; Th. 0.38–1.02cm Clay: 7.5-10 YR 7/6-4; firm consistency; fine temper; fine dense matrix; much fine dark mica; surface: well smoothed; paint: varying black to brown, dull sheen, visible brush strokes; slip: 10 YR 8/2. Sanctuary of Aphrodite at Zeytintepe. Find context: Z 05.20 from Q 05.4: part of the thick, densely packed pottery layer found in most of this quadrant as well as in the neighbouring quadrants Q 04.3, Q 05.1 and Q 05.2. Firmly bonded mixture of largish pieces of limestone and chalky soil. Pottery and bone finds correspond to Z 05.6. Z 05.88 from Q 05.2: high concentration of pottery, bone and small finds as well as many marble chips. Dark humus-rich soil, with limestone chips spilling in from the east. Part of the pottery deposition of the terrace fill, with diagonal layers of fill. Borders in the W on Q 05.4 and is connected with Z 05.6 and Z 05.20. Z 05.106 from Q 05.2: underneath Z 05.88. Loose humus soil with a large proportion of ash and thereby intensely grey-ish in colour. Interspersed with small stones, much pottery and marble. Borders on Q 05.4 in the W and is connected with Z 05.6 and Z 05.20. Unpublished.

Illustration credits Fig. 1a author, 1b–d H. Grönwald, Berlin; Fig. 2a–b author, 2c–e H. Grönwald, Berlin; Fig. 3a-b British Museum; Fig. 4a author, 4b H. Grönwald, Berlin; Fig. 5a author, 5b H. Grönwald, Berlin; Fig. 6a–b excavation at Didyma; Fig. 7a–c author, 7d H. Grönwald, Berlin; Fig. 8a–b author, 8c H. Grönwald, Berlin; Fig. 9a–b author; Fig. 9c H. Grönwald, Berlin; Fig. 10 U. Schlotzhauer.

Notes *

1 2 3 4 5 6

My gratitude is extended to the director of the excavation at Miletos, Prof. Dr. Volkmar von Graeve (Bochum), who gave me this material from Miletos for research and made it possible for me to report on it here. I also express my thanks to H. Bumke (Bonn) and to F. Heinrich (Bonn), who researches Archaic pottery from Didyma, for permission to publish and for providing the illustration of the mug (cat. no. 6, Fig. 6). For taking on the task of translating and checking my text I am grateful to E. Schalk (Berlin) as well as A. Villing and D. Williams (London). With regard to terminology, there have been regrettable alterations compared to the preliminary report (for example, Kerschner in Akurgal et al. 2002, 10). The terminology used here is based upon the publication by Kerschner and Schlotzhauer 2005 (for example, ‘Mile’ instead of ‘Mil’, ‘A Ia’ instead of ‘SubG’, etc.), and this is the terminology which should also be used by future studies. Ehrhardt 1988. Hdt. 2.178; Strabo 17.1.18; but see the discussion of the role of Miletos in Naukratis from the viewpoint of ancient literature; e.g. Möller 2001, 1–21; 2005; Bresson 2005. See, e.g., Austin 1970, 51 n. 4; Sullivan 1996, 190; Möller 2000b, 747; for the other position favouring a leading role of Miletos, see Haider 1996, 97; Herda (forthcoming b). Boardman 1980, 49. See e.g. Schlotzhauer 2001a. Important articles on pottery in the East Aegean that have appeared in recent years: Akurgal et al. 2002; Attula 2006; Ersoy 2003, 2004; Hürmüzlü 2004a; Kerschner et al. 2002; Kerschner 2003, Ýren 2002, 2003; Özer 2004.

Some Observations on Milesian Pottery 7 8

9

10 11 12 13 14

15

16

17 18 19 20 21

22 23 24

25 26 27

A fact that is rarely brought to speech, but which presents a serious problem for many excavations. Akurgal et al. 2002; Coldstream and Liddy 1996, 480-1; Dupont 1983, 1986, 2000; Hertel et al. 2001; Hughes et al. 1988; Jones 1986; Kerschner et al. 1993, 2002; Kerschner and Mommsen 2005; Kerschner 2006; Mommsen, Kerschner and Posamentir 2006; Mommsen, Schwedt and Attula 2006; Posamentir and Solovyov 2006; Schlotzhauer 2006; Seifert 1998, 2004; Seifert and Yalçýn 1996. One need only compare the surprising results relating to group G and g, which have been localised in Aiolian Kyme. With the aid of archaeometry an especially broad range of production can be proven; see Kerschner and Mommsen in this volume. The scientific results found archaeological confirmation in the fill of a kiln at Klazomenai, which contained a large number of different fabrics and decorative styles. See the preliminary report by Y. Ersoy 2003. See the papers in this volume by Attula, Bailey, Schlotzhauer and Villing, Weber, and Williams and Villing. See the papers in this volume by Attula and Mommsen; Dupont and Thomas; Kerschner; Kerschner and Mommsen; and Mommsen et al. See, e.g., Kerschner and Schlotzhauer 2005. See Kerschner and Schlotzhauer 2005, 1-4. Thus far there are few summarising monographs on Miletos. The most recent attempts are Gorman 2001 and Greaves 2002 (for Greaves see the reviews: Osborne 2003; Cobet 2004; Posamentir 2006), which includes older literature on Miletos and publications until 1999. A very comprehensive bibliography on Miletos is forthcoming by Ehrhardt, Lohmann and Weber in Cobet et al. (forthcoming). Important recent works on urban history, excavations and finds (excluding Milesian pottery), are Barrandon and Marcellesi 2005; Brize 2001; Cobet 2000, (forthcoming); Ehrhardt 1998, 2003a, 2003b; Forbeck 2002; Forbeck and Heres 1997; Günther 2003; Donder 2002; Graeve et al. 1999, 2001, 2005; Graeve 1997/8, 1999, 2000a, 2000b, 2001, 2005, (forthcoming); Heinz (forthcoming); Held 2000, 2002, 2004; Henke 2005; Herda 1998, 2005, (forthcoming a), (forthcoming b); Herrmann et al. 2006; Köster 2004; Krumme (forthcoming); Kunisch (forthcoming); Panteleon 2005; Pfisterer-Haas 1999; Schneider 1999; Selesnow 2002; Senff 2000, 2002, 2003, 2004, 2006, (forthcoming); Starke 2000; Stümpel et al. 2005; Weber 2002, 2004; and Zimmermann (forthcoming). See e.g. on aegyptiaca: Hölbl 1999, (forthcoming); bronzes: Brize 2001; Donder 2002; terracotta figurines: Graeve 1999, (forthcoming); stone sculpture: Graeve 1983, 1985, 1986a, 1986b, 1996, 2005. The most important publications on Milesian pottery, mainly of the Archaic period, are the following: A. finds: Graeve 1971, 1978; Graeve et al. 1986, 1987; Heinz 1990; Hommel 1959/60; Kleine 1979; Kleiner 1959/60; Mallwitz and Schiering 1968; Niemeier 1999, 381-414. B. Individual categories and observations: Aydemir 2005; Carl (forthcoming); Käufler 1999; Kerschner 1999, 2002; Kerschner and Mommsen 2005; Ketterer 1999; Krumme 2003; Naso 2005a; Posamentir 2002; Schlotzhauer 1999, 2000, 2001a, 2001b, (forthcoming a), (forthcoming b); Seifert 2004; Villing 1999; Voigtländer 1982, 1986. See Held 2000, 2004; Niemeier 1999. Brinkmann 1990; Graeve et al. 1986, 1987, 1990b; Seifert 1991; Senff 1997c. Graeve 1997/8, 75-80; Graeve and Senff 1990, 1991; Heinrich and Senff 1992; Senff 1995a, 1997b, 2000. See Kerschner 1995, 1999; Kerschner and Senff 1997. For cult, see Ehrhardt 2003b, 280-9; Herda 1998. See e.g. Gans 1991; Graeve 1997/8, 83-7; Graeve et al. 2005; Heinz and Senff 1995; Senff 1992, 1997a, 2003. For the settlement at Oikous, see Herrmann 1995, 285; Ehrhardt 2003b, 270-80; Herda 2005, 288-9 with ns 216-18, 291 with n. 230; (forthcoming b, chap. I with n. 12, chap. IX with ns 289, 292-6); Lohmann 2002, 232-3. Most recently on the hinterland of Miletos, with extensive literature: Lohmann 2002. See also on Teichioussa: Voigtländer 2004. A general overview is presented by Senff 2006. Lohmann 1995, 311-22, 1997a, 2002, 179-81; Senff 1995b; B.F. Weber 1995. O. Rayet and A. Thomas already localised the site in 1877 as well as U. v. Wilamowitz in 1906 (detailed descriptions by Herrmann 1995, 291-2 with ns 152-3). Kalaitzoglou (forthcoming). See Lohmann 1995, 1997b; Kalaitzoglou (forthcoming). The bibliography on Didyma is very comprehensive. Here only the

28 29

30

31 32 33 34 35 36 37

38 39 40

41

42 43

44 45 46 47 48 49 50 51 52

53 54 55 56

57

most recent are mentioned; through them older works can also be accessed: Bumke et al. 2000, 2002; Filges 2004; Filges and Tuchelt 2002; Tuchelt (forthcoming). See e.g. Naumann and Tuchelt 1963/4, 42-62 pls 8-26; Tuchelt et al. 1971, 57-87 pls 3-17, 1973/4, 149-51, Schattner 1989, 1992. This excavation is part of a project being conducted by H. Bumke (Bonn), E. Röver (Berlin) and A. Filges (Frankfurt) at the Taxiarchis hill. See preliminary reports by Bumke and Röver 2002; Filges 2004; Filges and Tuchelt 2002. My gratitude to H. Bumke (Bonn) and F. Heinrich (Bonn), who is studying the Archaic pottery, for the opportunity to see the excavation material and illustrations of the mug fragment cat. no. 6 (Fig. 6). Tuchelt 1996; Bumke et al. 2000. For a summary on the Processional Avenue, see Herda (fortcoming a), (forthcoming b). For Archaic and later pottery from the cultic area on the Sacred Road, see Schattner in Tuchelt 1996. Lohmann 1995, 1996, 1997b, 1999a, 1999b, 2000, 2001, 2002, 2004, (forthcoming); Berndt 2003. In particular, thus far no Archaic kilns have been found outside of the ancient city of Miletos. In the following the new classification system for East Greek pottery according to Kerschner and Schlotzhauer 2005 is used. Kerschner and Schlotzhauer 2005. A group classified within the Klazomenian black-figured style, according to R.M. Cook. See Cook 1998, 95-8; Özer 2004, 202-3. Likewise a group of black-figured vases, presumably produced foremost on the island of Chios. See Lemos 1991, 169-75, 189-90; Cook 1998, 75; see also Williams, this volume. In which details were reserved and not incised. For the divergence from the technique of black-figured vase-painting, see Cook 1998, 78–82, 89; for a different opinion on the dating of MileA II (Fikellura) see Schlotzhauer (forthcoming b). See Cook 1998, 101-3; Özer 2004, 204-5. See Cook 1998, 103-5; Özer 2004, 204-5. Where they were still used, they experienced a change in meaning, as for instance in the characterisation of layers of cloth: the differentiation between the upper garment and the trousers underneath or contours of the body visible through the cloth. See the symposiast: Walter-Karydi 1973, pl. 72.555. Schlotzhauer (forthcoming b) attempts to define the distinction between the two main phases, while Kerschner and Schlotzhauer 2005 provide definitions for the fine classification of the early phase SiA I. Kerschner and Schlotzhauer 2005. Forthcoming are monographs with the following corresponding sections on: amphorae in the Fikellura style (MileA II): F. Wascheck (Bochum); lids: R. Posamentir (Ýstanbul); trade amphorae: A. Naso (Campobasso); cooking vessels: A. Aydemir (Bochum); oinochoe: S. Käufler (Bochum); bowls: A. Villing (London); and mugs and cups with everted rim: U. Schlotzhauer (Berlin). Pottery of the Geometric period is being prepared for publication by M. Krumme (Athens). For the older systems of classification and their correlation with the new division, see Kerschner and Schlotzhauer 2005, 9, 17, 25, 33. Kerschner and Schlotzhauer 2005. Among the recent publications, special mention is made here of the detailed study by Käufler 1999 as well as the general description by Cook 1998, 29-45. Further essential publications on this topic can be found in the article on the classification of the phase SiA I by Kerschner und Schlotzhauer 2005, 33-45. See e.g. Walter-Karydi 1973, pl. 65.530, 67.530. Kerschner 2002, 40, 172 fig. 52. See Graeve 1971 and the satyr from excavations at Kalabaktepe (K 1992.696.2), illustrated in Simon 1997, 1114 no.29c, pl. 751.29c . For MileA Ia, see Kerschner and Schlotzhauer 2005, 9-16. For MileA Ib, see Kerschner and Schlotzhauer 2005, 17-25. Käufler (1999) demonstrates the early development of the phase MileA Ib by way of some examples, however, still without the new classification and terminology. For MileA Ic, see Kerschner and Schlotzhauer 2005, 25-33. Walter 1968, pls 116-17.592. For MileA Id, see Kerschner and Schlotzhauer 2005, 33-45. See Kerschner and Schlotzhauer 2005, 43 fig. 43. For the results of the chemical analysis of the krater, see Mommsen et al. this volume and Schlotzhauer and Villing this volume; for the Kalabaktepe workshop, see Kerschner 2002, 37-42. Kerschner and Schlotzhauer 2005, 44 fig. 44. Naukratis: Greek Diversity in Egypt | 143

Schlotzhauer 58 In addition, for the fragment of a plate of the phase MileA Id from Edfu, see Weber 2001, pl. 23.3. 59 For this phase, see Schlotzhauer (forthcoming b). 60 For the traditional classification into groups, see Cook 1933/4. 61 See Schaus 1986, with a bibliography of older publications. 62 For a list of known shapes in Fikellura (= MileA II), see Cook 1998, 77-8; Schaus 1986, 268-70, 281-2 (on the two painters discussed by Schaus in his paper). 63 Cook 1998, 78, who concludes his discussion of the shapes of Fikellura (MileA II) vases with: ‘Further excavation at Miletus may well add more.’ 64 For the distribution of painted pottery in the style of MileA II, see Cook 1933/4, 85-9; 1998, 88-9; Schaus 1996, 31, 34-6, 40 fig. 3 (limited to the East Greek region). G. Schaus discusses in detail two vase painters from two different generations known from exported vessels (Schaus 1986). See also Schlotzhauer 1999; Posamentir 2002. 65 The best comparison for this image thus far comes from the Heraion on Samos, where it decorates the same kind of vase, a cup with everted rim, but of a more elaborate type, similar to the well-known ‘Little-Master’ cup in the Louvre (Walter-Karydi 1973, pl. 46); see Walter-Karydi 1973, pl. 47.424. Nevertheless, there is one important difference: the figure from the Heraion is winged and is not represented in the ‘bent-knee’ running position. 66 The theme of the potnia theron in Miletos will be discussed more fully in a planned study of the iconography of MileA II pottery from Miletos. 67 Schlotzhauer 1999, 235-6. 68 Schlotzhauer (forthcoming a). 69 See infra with ns 97-98. 70 BM GR 1888.2-8.46; Petrie 1888, pl. 27, 3 and 3 a; Petrie 1891a, 56 fig. 1; Walters 1893, 92 no. B 117; Cook 1954, 7 pl. G.B. 574.1, details of shoulder pl. G.B. 569.3–4; Walter-Karydi 1973, 137 no. 683 pl. 89. 71 Cook 1933/4, 8 (C1), 9, 65, 73 fig. 13.9, pl. 4c. 72 Kerschner and Schlotzhauer 2005, 7, 46. 73 Different places of production are one of the grounds for defining the new classification system of Archaic East Greek pottery suggested recently by Kerschner and Schlotzhauer 2005, 4-5, 7. 74 Cook 1960, 118-26. 75 Walter-Karydi 1973, 2-52. Kerschner 2002, 41-4, and Schlotzhauer (forthcoming b) give good overviews of the history of research on Fikellura pottery. 76 Walter-Karydi 1973, 137 nos 677-84 pl. 89. Walter-Karydi (1973, 66) postulated ‘[…] das scheint ephesische Art zu sein […]’, p. 107 n. 180. However, she qualified this by adding: ‘Die Technik ist der milesischen sehr ähnlich; das Gefäß könnte auch milesisch sein.’ 77 Walter-Karydi 1973, 137. Her nos 677, 678 and 679 belong to the socalled ‘Ephesian Ware’, a distinctive ware found in Sardis, Ephesos and, marginally, Miletos. Cf. Greenewalt, Jr. 1973, 91-122. Greenewalt, Jr., no. 1 = Walter-Karydi 1973, no. 677, Greenewalt, Jr., no. 2 = Walter-Karydi no. 678. Analyses show that some of the vessels of the ‘Ephesian ware’ were manufactured in Sardis: Kerschner 2005, 139. For locally produced Archaic vessels in Ephesos (groups H and I), cf. Kerschner 1997, 211; 2002, 189-205. Bands of squares are not found exclusively on vases of the ‘Ephesian ware’, but also on vases of the late animal frieze style (SIA Id) and the Fikellura style (MileA II). Cf. Cook 1933/4, 71, 75 fig. 10.7, pl. 16. 78 Dupont 1983, 37-9; 1986, 57-71; Jones 1986, 665-6. 79 Schaus 1986, 283-4. 80 Kerschner 2001, 82 pl. 8.3; Kerschner 2002, 43-4. 81 For group D, cf. Kerschner 2002, 44-7, 137, 143. 82 Cf. the paper by H. Mommsen et al. in this volume. The analyses will be published in detail in Schlotzhauer and Weber (forthcoming). 83 Dupont 1983, 37-9; Cook 1998, 77-90; Kerschner 2002, 204-5; Mannack 2002, 98; Kerschner and Schlotzhauer 2005, 46; cf. also the papers by Dupont and Thomas and by Attula in this volume. For locally produced pottery, imitations of the South Ionian Archaic style II (SiA II), found in Tell Defenneh cf. Weber (forthcoming) and in Naukratis, see Schlotzhauer and Villing in this volume. 84 Cf. note 8. 85 Kerschner 2002, 34-6. 86 M. Akurgal (Ýzmir), R. Attula (Greifswald), M. Frasca (Catania), W.-D. Niemeier (Athens), R. Posamentir (Ýstanbul) and S. Solovyov

144 | Naukratis: Greek Diversity in Egypt

(St. Petersburg), U. Schlotzhauer (Berlin), A. Villing (London), S. Weber (Mainz) and D. Williams (London). 87 Cf. Schlotzhauer and Villing Figs 18–22. 88 Schlotzhauer 1999, 224-5. 89 Schlotzhauer 1999, 225-7. 90 Schlotzhauer 1999, 227-8. 91 Schlotzhauer 1999, 228-32. 92 Schlotzhauer 1999, 232-5. 93 Schlotzhauer 1999, 235-6. 94 Filges 2004, 152 fig. 3 (= Filges and Tuchelt 2002, 8 fig. 5.2 = Bumke and Röver 2002, 98 fig. 18.2). See also the new fragments of a trefoil mug from Istros: Bîrzescu Figs 10–11. 95 During a visit in 2005 the author was able to view the comparable mug from Olbia, now in the Museum in Kiev (Schlotzhauer 1999, 224 fig. 1). Thereby he could discern that the mug originally had two handles (cf. also the recent publication of the mug, with two handles, by Reeder 1999, 178 no. 65), whereas in the abovementioned article (Schlotzhauer 1999) it is pictured with only one handle; see also Lemos 1991, 179-80, pl. 220.1658, who refers to it as a one-handled mug and tentatively thinks of Chian production; the first publication of the piece, Shtitelman 1977, no. 12, had assigned it to Rhodes. 96 For this group of mugs with dipinto, see Schlotzhauer 2006, 311-3. 97 Schlotzhauer 1999, 236-8; (forthcoming a). 98 See the preliminary discussion of this problem in: Schlotzhauer (forthcoming a). 99 In the contribution by Schlotzhauer (1999) the well-known decorated MileA II-mug from Istros (Alexandrescu 1978, 52 fig. 153) is unfortunately omitted (as kindly pointed out by P. Alexandrescu during the Panionion conference in Güzelçamlý 1999 on the occasion of the 100th anniversary of the German excavations at Miletos). 100 The Geometric forerunners will be discussed by M. Krumme (Athens) and the early Archaic examples by the author. 101 The dipinti and graffiti from Miletos have been studied by N. Ehrhardt (Münster). I am grateful to N. Ehrhardt for permission to present here these inscriptions from Miletos, some of which are as yet unpublished. 102 Chian chalices or kantharoi with dipinti of Aigyptis (first identified by D. Williams) in the BM: GR 1924.12-1.755 and 808; GR 1924.121.720; GR 1924.12-1.827; see Boardman 1986, 254-6; Williams 1983a, 185 with n. 59; Williams 1999, 138 and fig. 52d. 103 Bumke and Röver 2002, 98 fig. 18, 99 fig. 21. 104 Bumke and Röver 2002, 86 and n. 5. 105 It can be gathered from the excavation reports (Hommel 1959/60, 1967) that in 1957 Hommel excavated in trenches H/J XII/XIII and J XIV, directly below and next to the foundations of the late Athena temple. Through the explicit note in the diary about find context 191, in which the fragment of the mug was found: ‘191: 25./26.9.57 SOAbschnitt, NOTeil...’ (excavation diary 1957, p. 27 [P. Hommel], Miletos Archive, Bochum), the exact location of the mug can be pinpointed. 106 See Ehrhardt 1988, 164 with n. 757. 107 See Held 1993, 371-5, esp. 375. 108 Held 1993. 109 Hommel 1959/60, 31-2. 110 Nevertheless, one architectural fragment discovered during excavations in 1968, conducted to the northeast of the Southeast building, could stem from an altar, which need not necessarily connected with the Athena sanctuary. Cf. Schiering 1979, 90-1, 96-7, figs 4a-b. 111 Letter of October 27, 1998. 112 See Frisk 1960, 65; Chantraine 1968, 55. 113 IG V 2,6 Z.117, mit der Lesung L?asi/w. 114 Bechtel 1917, 494. 115 SEG XXII 508 A 20. 116 Belege: Sarikakis 1989, 24, 286. 117 See Haussoullier 1879, 244 n. 2. 118 See Schlotzhauer 1999, 225-7. 119 Pers. comm. 120 See Schlotzhauer 1999, 235-6. 121 For eye-decoration see Schlotzhauer (forthcoming a) n. 82. Cf. also Kunisch 1990, 20-7; Steinhart 1995, 55-6.

East Greek ‘Situlae’ from Egypt Sabine Weber With an Appendix: Neutron Activation Analysis Results by H. Mommsen,A. Schwedt, S.Weber and M.R. Cowell Abstract A distinctive group of East Greek vases with elongated shape and wide open mouth called ‘situlae’ was mainly found in Tell Defenneh, Egypt. The shape and some painted Egyptianizing subjects indicate that the Greek potters and painters focused on a clientele living in Egypt. The place of manufacture of the ‘situlae’ is a question of debate. NA-analyses carried out in cooperation with H. Mommsen and A. Schwedt, Bonn, showed that with the same clay paste not only ‘situlae’ but also other shapes like stamnoi and an amphora had been made. The place of manufacture, however, could not be precisely located because the samples belong to a new chemical group not as yet represented in the databank of Greek pottery compiled by H. Mommsen and M. Kerschner.* The evidence The so-called East Greek ‘situlae’ raise questions of shape, iconography, and place of manufacture. Fundamental to the study of East Greek ‘Situlae’ is the chapter in the Corpus Vasorum of the British Museum, fasc. 8 by Robert M. Cook.1 East Greek ‘situlae’ were found in East Greece (Rhodes and Samos) and in Egypt (Tell Defenneh, Memphis and, just recently noted, possibly also at Naukratis [Fig. 1]).2 These three sites also yielded the greatest quantity of the Greek painted pottery from the late 7th to the late 6th century bc found in Egypt (Naukratis more than 7,000 vessels and sherds, Tell Defenneh about 340, Memphis with its necropoleis about 30). In other places in Egypt Greek painted pottery has been found in much smaller quantities.3 The study of East Greek ‘situlae’ began with the excavation of a site in the eastern Delta at the Pelusian branch of the Nile by W.M. Flinders Petrie in 1886, because the first vessels of this shape were found here.4 Petrie identified the place from the

Arabic name, the location, and the name in the Old Testament, Tahpanhes (Jeremiah 43.5–13), as the cosmopolitan city of Pelusian Daphnai mentioned by Herodotus 2.30. He equated this city with the stratopeda, or the camps, where Psammetichos I, the first king of the 26th Dynasty (reigned 664– 610 bc), settled the Ionian and Carian mercenaries who helped him to establish his reign. It is, however, more plausible that the stratopeda are sites archaeologically not located up to now and thus not identical with Daphnai.5 The high percentage of Greek pottery and the position at the eastern border of the Delta let Petrie conclude that Greek mercenaries were stationed at Tell Defenneh and that it was the Greeks living there that used the pottery. But Greek cooking pots, a good indicator of different culinary practices and therefore indicative of the presence of different ethnic groups, have not been found.6 The function of the main excavated architectural structure with a casemate 2 foundation of about 43.5 m is not quite clear.7 It is very probable that it was an official building, constructed under the reign of Psammetichos I, as attested by the foundation deposits. The purpose of the building is still debated: a secure centre for local administrative purposes and control,8 a royal palace, a treasury, a temple or a temple storehouse. Annexes with many small chambers were added successively to the main structure. In these adjacent rooms most of the Greek pottery was found, the greatest quantity of Greek pottery outside of Naukratis, together with Egyptian material of various kinds (pottery, scarabs, gold and silver objects, faience objects, military and non-military bronze and iron objects like knifes, scale-armour, arrow-heads and weights) mostly without any significant stratigraphy, just lying in the dust of the desert. It is not possible to give the percentage of the Greek pottery in relation to the Egyptian material because Petrie often just noted that a shape was ‘common’ without quoting the number. In two rooms sink-jars were found, suggesting that they might have served as kitchens and the others as storage rooms. In this connection we may note that Greek and Phoenician amphorae were reused as water containers as attested by Herodotus (3.5-7) for the Persian Period: the amphorae were collected at Memphis, refilled with water and than sent to the arid areas in the Eastern Delta.9 Perhaps this was already the case in the 6th century bc and could be the reason for the high percentage of container vessels in Tell Defenneh. The Greek painted pottery found there dates from the last decade of the 7th century to about 525 bc – then it ends. This could be explained by the Persian conquest of Egypt. Other Greek objects said to be found in Tell Defenneh are an East Greek gem, two terracotta antefixes, and a bronze bowl (the last three objects not mentioned by Petrie).10 That Greeks possibly still lived there in the first half of the 5th century bc is suggested by the graffito at Abydos of a certain Timarchos, who describes himself as being from Daphnai’.11

Figure 1 Map of find places of East Greek ‘situlae’

Naukratis: Greek Diversity in Egypt | 145

Weber Egyptian bronze situlae and the question of the origin of the shape of the East Greek ‘situlae’ Petrie named the vessels of a distinctive group as ‘situla-form vases’ because the shape reminded him of the Egyptian bronze situlae (Fig. 2), which were very popular during the Late Period.12 These are relatively small bronze vessels that served a different purpose from the big Greek storage vessels (the height of the so-called Typhon ‘situla’ is 54.0cm; the others can be restored to a height between 39.5 and over 40.0cm depending on the state of preservation and possible reconstruction; the Egyptian bronze situlae in contrast are usually between 6.0 and 20.0cm high and rarely taller than 30.0cm). East Greek ‘situlae’ cannot easily be carried around but are practical for storage and mixing. The small Egyptian vessels with their characteristic omega-shaped bail handle and bag-like shape, sometimes ending in a nipple (perhaps reminiscent of the female breast), were used in Egyptian cult practice (in the ritual of Djeme and in funerary ritual, involving Amenope and Isis) for carrying and for the subsequent libations of water or milk to the dead, at least from the time of the New Kingdom onwards.13 They bear a distinctive decorative scheme in three registers: in the upper register below the rim the solar barque, with jackals and baboons, represents the transit of the sun across the sky. The middle register represents the earth and depicts a worshipper in front of gods: an ithyphallic god (Amun-Min or Amenope), Isis, Nephthys, Horus and other deities. The lower register represents the water, and the bottom of the vase is decorated by a lotus flower. Some Egyptian bronze situlae have also been found outside Egypt, in the Near East,14 Italy and Greece: Lefkandi in Euboia,15 Pherai in Thessaly,16 the sanctuary of Malophoros in Selinus/Sicily,17 sanctuaries and necropoleis in Cyprus,18 and the Samian Heraion (Fig. 2).19 Greek artists were therefore able to become acquainted with the Egyptian shape in Greece. But obviously they did not copy the Egyptian bronze situla itself. When copying or adopting a foreign vase into their repertoire, Greek potters always kept the shape either because it was a handy addition to their repertoire or because they could increase their export volume to the place of origin of this shape.20 They usually transformed the foreign shape merely by painting it or by using another material.

Figure 2 Egyptian bronze situla from Samos,Vathy,Archaeological Museum

146 | Naukratis: Greek Diversity in Egypt

The term ‘situla’ is Latin.21 In Greek, vessels with the same function were called kados, antlion or gaulos.22 The expression ‘situla’ is in use for various bucket-shaped metal and clay vessels with bail handles which were produced all over the Mediterranean23 and in Central Europe.24 But in nearly all of these cases the type of handle of these vessels is indeed that of a modern little bucket, carrying and pouring being the main characteristic features. In the Eastern Mediterranean a big pyxis shape with a conical body, horizontal banded handles and a wide flat bottom without a separate foot is also called a ‘situla’.25 The rendering of the lip, with a distinctive ridge underneath, is comparable to our ‘situlae’. A Late Geometric example from Smyrna shows a similar decorative scheme as the early ‘situlae’: a broad frieze underneath the ridge with three fields, the middle one with figural decoration (water bird to right), the outer ones geometric patterns. Below this frieze is a banded decoration.26 A krater-like Lydian vessel found in Rhodes was also put forward as a possible prototype for the Greek ‘situlae’.27 But the rendering of the handles applied directly at the lip differs from the ‘situlae’ and the body is not as elongated as in the ‘situlae’. Possible ancestors of the shape of the ‘situla’ might be undecorated Greek pithoi like the ones from Rhodes, Samos or Nisyros (Fig. 3).28 In Egypt, too, we find a wide range of parallels in storage vessels with long, bulging bodies but without a foot (Fig. 4).29 For the ‘situlae’ the Greek potters just added a foot and copied the figural and ornamental decoration from other shapes. The name ‘situla’ for our special group is thus misleading, but has been well established for over a century. I would like to suggest putting it in quotation marks to symbolize the ‘so-called’ status. East Greek ‘situlae’ The shape is a tall narrow tube, swelling gradually towards the base and then curving quickly in to a low foot; the lip is flat and wide, and below it, separated by a narrow ridge, are two vertical handles. These Greek ‘situlae’ are containers for storage and mixing.30 It is not surprising to find them in an Egyptian context, because the percentage of containers among the Greek finds in Egypt is overwhelming. Being painted, these vessels were a more exclusive version of a container.

Figure 3 Greek pithos from Samos, necropolis

Figure 4 Egyptian jar from Thebes, Medinet Habu, burial 9, 25th dynasty

East Greek ‘Situlae’ from Egypt Figure 5 East Greek ‘situla’ from Vroulia (Group A)

East Greek ‘situlae’ were mainly found in Egypt (Tell Defenneh, Memphis and Naukratis) but are as well attested in Greece: on Samos and Rhodes from the end of the 7th to the end of the 6th century bc where they were found in settlement contexts in Vroulia, in the sanctuary of the Samian Hera, and as grave goods on Rhodes.31 The earliest ‘situlae’ came to light in Vroulia, Rhodes (Fig. 5), and in the Heraion of Samos (Fig. 6). The shape continues in Southern Ionia and Eastern Doris for over a century. Whether some fragments mentioned in this context from Miletos,32 Samos,33 Burgas (Datça)34 and Rhodes35 belong to this group of vases is, however, doubtful because of their bad state of preservation. They could also belong to other shapes like the ‘pyxis situla’ and are therefore not included here in the map. R.M. Cook divided the ‘situlae’ into three different, essentially chronological groups as follows. Group A comprises ‘situlae’ with banded decoration and wavy lines between the handles (Fig. 5). This is a distinctive group of five examples from Vroulia (they can be dated by their association with North Ionian bird bowls in contexts of the end of the 7th/ beginning of 6th century bc).36 G.R. Schaus suggested that perhaps one fragment with banded decoration from Tell Defenneh belongs to this group.37 Group B (Figs 6–9): To this group belong the so-called Typhon ‘situla’ (Figs 8–9), the ‘situla’ from Memphis and a fragment from Rhodes with the depiction of a griffin (Fig. 22). The decoration between the handles is divided into three fields, the middle one figurative, the outer ones ornamental. The handles are composed of three reeds. The lower part of the vessel is painted with broad black bands interrupted by small reserved stripes. Maybe the ‘situla’ from Samos (Figs 6-7), rendered in the animal frieze style (Wild Goat Style II/ SiA I c) and therefore

to be dated to the end of the 7th or beginning of the 6th century bc, is an earlier variant of this group.38 The ‘situlae’ from Egypt are all in black figure style with added red and can be dated to the second quarter of the 6th century bc. The pieces of this group are always made very finely. In proportion to the height the wall is very thin. The rendering of the mouth of the Typhon ‘situla’ is remarkable. The mouth plate juts out on the outside as well as on the inside. Group C (Figs 10–13): We can count about 32 ‘situlae’ belonging to this group, most of them badly preserved and restored. Except for two examples from Rhodes,39 the vases of group C were all found in Tell Defenneh. The three-field-decoration is given up in favour of a broad panel with figural decoration. The lower part of the body is painted with two or more friezes of alternating incised lotus flowers and palmettes highlighted by added red. This colourful decoration recalls the ‘Vroulian’ ware.40 The handles may be three or four reeded. The evolution of the shape shows a tendency from a more bulky to a more slender, tubular body. A common, never neglected feature of all ‘situlae’ of Group A–C is the small ridge some centimetres beneath the lip (cf. Figs 7, 9, 11). On some examples the ridge on the outside corresponds to a groove on the inside as well.41 The handles are applied to the body always at the level of this ridge. Late examples from the necropolis of Ialysos still preserve the lids and it can be assumed that a lid belonged to every single vase of

Figure 6 East Greek ‘situla’ from Samos, Samos, Heraion, Inv. K 1590 (Group B)

Figure 10 East Greek ‘situla’ from Tell Defenneh, London, British Museum, GR 1888.2-8.27 (Vase B 106.2) (Group C)

Figure 7 Profile of East Greek ‘situla’ from Samos, Samos, Heraion, Inv. K 1590

Figure 8 East Greek ‘situla’ from Tell Defenneh, London, British Museum, GR 1888.2-8.1 (Vase B104) (group B)

Figure 9 East Greek ‘situla’ from Tell Defenneh, London, British Museum, GR 1888.2-8.1 (Vase B104)

Figure 11 Profile of East Greek ‘situla’ from Tell Defenneh, London, British Museum, GR 1888.2-8.27 (Vase B 106.2)

Naukratis: Greek Diversity in Egypt | 147

Weber

Figure 16 Fragment of ‘situla’ from Tell Defenneh, London, British Museum, GR 1888.2-8.3

Figure 12 East Greek ‘situla’ from Rhodes, Rhodes,Archaeological Museum, Inv. No. 10641 (group C)

Figure 13 East Greek ‘situla’ from Rhodes, Rhodes,Archaeological Museum, Inv. No. 10641

this shape.42 The ‘situla’ from Ialysos (Figs 12–13) can be dated by an Attic late black-figure olpe in the same grave context to the last decade of the 6th century or around 500 bc. So we can trace the shape for almost a century. The new ‘situla’ fragment from Naukratis (Figs 14–15) does not fit in one of these groups. Being a more slender vase it comes close to ‘situlae’ of the younger group C but the decoration is different. Subjects on East Greek ‘situlae’ from Egypt Some of the subjects on the ‘situlae’ found in Egypt are remarkable, as already noted by Petrie and other scholars.43 Subjects suitable for both cultures are depictions of lions, rams, bulls, birds and sphinxes.44 But there are also subjects with a distinctive Egyptian flavour. On a small flat fragment, flaked off from the surface of a ‘situla’, a falcon on nb-basket is painted (Fig. 16).45 The hieroglyphic sign nb (basket) means ‘lord’ or ‘all, everybody’. The falcon is the emblem of the god Horus who was equated with Greek Apollo. Perhaps it could be completed as ‘Horus, lord of…’ or ‘the two lords of Upper- and Lower Egypt’, but in that case the Seth animal or a second falcon would also have had to be depicted on the nb-basket. This sign can also be part of the name of the pharaoh. The painter could have known this hieroglyphic sign either from having seen it in Egypt or via small-scale Egyptian and/or Egyptianizing objects traded to, or manufactured in Greece, such as the silver cartouche from a cremation burial in Ialysos, Rhodes, with the depiction of a bird on a basket.46 The bird on this ‘situla’ fragment must have been an integral part of the figural decoration of the ‘situla’, but due to the bad state of preservation we cannot judge what the original picture looked like. Besides the cartouches on the

Figure 14 Fragment of ‘situla’ from Naukratis, London, British Museum, GR 1886.4-1.1311

Figure 15 Fragment of ‘situla’ from Naukratis, London, British Museum, GR 1886.4-1.1311

148 | Naukratis: Greek Diversity in Egypt

Figure 17 Fragment of ‘situla’ from Tell Defenneh, London, British Museum, GR 1888.2-8.5 (Vase B106.1)

amphora in Basle and the Petrie Museum, London (Bailey Figs 1–5),47 this depiction is the only known painted hieroglyph on a Greek vase from Egypt.48 Another subject is two men fighting with clubs (Fig. 17).49 The better preserved man is characterised as Egyptian by his shaved head and circumcision. Herodotus (2.36) tells us that the Egyptians used to be circumcised for hygienic reasons. In the second quarter of the 5th century bc, on the Attic red figure pelike by the Pan Painter, Egyptian priests of the Egyptian king Bousiris are also characterised this way.50 This feature was therefore judged appropriate to indicate an Egyptian. Club fighting or fencing was a sport depicted in Egyptian art from the time of the Old Kingdom, and mainly in the New Kingdom (1551–1085 bc).51 The picture of this Egyptian sport is, however, rendered in a Greek manner on the ‘situla’. The sportsman is naked, a feature not often found on Egyptian representations where a loin cloth is usually worn52. The rendering of nudity gave the Greek painter the opportunity to indicate the ethnicity of the sportsman by means of the circumcision. In Egypt representations of athletics in the Late Period are rare and there are no illustrations of fighting contemporary with our vase;53 thus this picture is an important reference, showing that fighting was either performed in the 6th century bc or that the Greek painter had seen ancient Egyptian illustrations of club fighting. On the ‘situla’ from Memphis,54 found next to the Ptah enclosure where the Apis bull was worshipped, the subject of two bulls on one side seems appropriate. The rendering of a man on the other side wearing a dress with broad borders, striding to the right carrying a long staff in the right hand and holding some object in the left looks unfamiliar to a Greek eye.55 The painter had obviously difficulty in rendering this figure correctly: he erased a part of the right foot and leg and a tip of a cloak.56 If we take the representation ‘man with staff’ as a sign or code,57 then the picture could be understood in a Greek as well as in an Egyptian context. In both cultures the staff is the sign of dignity or age.58 This figure may vaguely recall Egyptian dignitaries.59 Even more puzzling is the rendering of a rectangular frame with two wavy lines in the upper right corner. It is not a Greek decorative pattern but placed in front of the head of the dignitary. Could it be an Egyptian sign, a hieroglyphic pattern, to give additional information on the man? If this were the case, the painter was not able to read or write Egyptian hieroglyphs but intended to paint a sign that looked like a hieroglyph.60 In G. Schaus’s opinion the man could be a worshipper of Apis.61 If so, the painter had selected a subject that would well suit this location. This phenomenon has already been noted in connection with Attic pottery for other areas.62 Few other Greek vases of the 6th century bc found in Egypt

East Greek ‘Situlae’ from Egypt show Egyptian influence and were therefore intentionally produced for a market with a clientele living in Egypt.63 The representation of the snake-bodied demon on side A of the best-preserved ‘situla’ from Tell Defenneh (Fig. 8) could also have been painted for the Egyptian market, or at least fits very well its find spot.64 The mixed creature is usually called Typhon.65 In Greek mythology Typhon was the son of Gaia and Tartaros, battling against the established order with the aim of dethroning Zeus.66 He was slain by Zeus and thrown into Tartaros where he revolted by means of earthquakes and volcanic eruptions. Typhon is rendered in Greek art as a mixed monster with a winged human body and snake legs, as on the lost Amyklaean throne, described by Pausanias 3.18.10. The battle between Zeus and Typhon was a subject on bronze shield-bands from Olympia in the first half of the 6th century bc.67 The depiction of the demon on the ‘situla’ recalls representations on late Protocorinthian vases, mainly alabastra from the 7th century bc onwards, and on other Greek vases.68 From the 6th century bc onwards the name Typhon was also used for the Egyptian god Seth (Hdt. 2.144, 156).69 There are hints that the Greek name was also accepted in Egypt.70 The name was written with the hieroglyph of the so-called Seth-animal, a jackal-like animal with long ears, very probably a mythical animal for it is not possible to determine its zoological identity.71 The Egyptian representation of Seth was the Seth-animal or a semianthropomorphic form with the head of the Seth-animal. He was the god of confusion and chaos who disturbs order. He was both an enemy and friend of Horus and murderer of Osiris. Seth was set in opposition to Horus. Horus is lord of Lower Egypt and lord of the home country. Seth is the lord of Upper Egypt and lord of foreign countries and the desert. He was venerated on the border of the desert and where caravan routes began, e.g. in the vicinity of Tell Defenneh. The Egyptians accepted him as a god up to the 20th Dynasty; later worship turned into demonization. His physical strength is characteristic for Seth. He is able to conquer the great chaos serpent Apophis,72 but he could also be identified with her. The Greek painter of the ‘situla’ chose to depict the Greek version of a demon known in both cultures. A further indication that the vase might have been highly esteemed in an Egyptian context is Petrie’s note of traces of a Demotic inscription in ink.73 Unfortunately Petrie gave no exact description of it and neither Cook nor I could detect traces of this inscription.74 Place of manufacture One important question is where the ‘situlae’ from Tell Defenneh may have been made. The ‘situlae’ of Cook’s group A from Vroulia on Rhodes are supposed to have been produced locally, a claim never refuted.75 As to the place of manufacture of the Group B ‘situlae’, Rhodes or another place in the Eastern Doris was favoured.76 Shape and Egyptian subjects on some of the fragments were the reasons why M.W. Flinders Petrie supposed that the ‘situlae’ were produced locally at Tell Defenneh.77 R.M. Cook also favoured the idea of a production centre in Egypt but only for his group C ‘situlae’, except for the two late pieces from Rhodes, which he thought to be manufactured near the place where they had been found.78 E.R. Price suggested that the ‘situlae’ were made for the Daphnae market by Ionian potters, but not on the spot.79 She favoured a Rhodian origin, because the earliest example known to her came from Rhodes and the

youngest are stylistically close to the ‘Vroulian ware’, which she believed to be Rhodian. In E. Walter-Karydi’s opinion the ‘situlae’ of Cook’s group C should be Aiolian.80 She compared the figures on some of the ‘situlae’ in London with the so-called Caeretan hydriae. G. Schaus noted that the ‘situlae’ are not homogeneous but made in more than one fabric, and put forward the suggestion that they were produced in Rhodes and Egypt.81 J.Y. Carrez-Maratray followed Cook in assuming an Egyptian origin for the ‘situlae’ of group C,82 a hypothesis thought possible also by A. Möller83 and K. Smoláriková.84 They favoured the idea that Rhodian clay must have been transported to Tell Defenneh or its vicinity and there made into ‘situlae’ by Rhodian potters.85 The clay of Cook’s group C, to which most of the ‘situlae’ from Tell Defenneh belong, is not the local Egyptian Nile silt nor is it marl; it is dense and usually of light brown colour (most sherds correspond to Munsell 7.5YR 6/4 and 10YR 6/4 = light brown to light yellowish brown). Petrie described it as ‘finegrained hard pale buff clay’, some examples being harder and browner, some paler and soft light grey. He remarked further that a group of stamnoi from Tell Defenneh share exactly the same clay, colouring and designs, but never have figured decoration.86 This observation led to the conclusion that the clay for both the ‘situlae’ and the stamnoi was transported from somewhere in Eastern Greece to Egypt and that the vessels were manufactured at Tell Defenneh or its environs.87 But the principal find place of a stylistically homogeneous group of ceramics is a weak indicator for the determination of the place of manufacture; this has been obvious ever since the true place of manufacture of Chian pottery, formerly called ‘Naukratite’, was revealed.88 J. Boardman and R. Jones carried out scientific analyses on ‘situla’ fragments in Oxford belonging to Cook’s group C and suggested a Rhodian origin.89 The ‘situlae’ matched Jones’s cluster II, comprising also 8th century bc pottery and mid-5th century bc terracotta statuettes found on Rhodes and thought to have been produced locally. The style of the analysed ‘situla’ fragments is close to Vroulian vases, believed to have been produced on Rhodes. The analyses carried out by P. Dupont also led to an assumed Rhodian origin. He took a sample of a sherd perhaps belonging to a ‘situla’ with the provenance ‘from Egypt’ (sample no. DEF 1; Dupont and Thomas Fig. 6) that could be associated with samples of two Vroulian cups (sample nos NAU 58 and 59; Dupont and Thomas Fig. 5) (Dupont’s group C2).90 These two analyses point to Rhodes as place of manufacture for some of the ‘situlae’ of Cook’s group C. We had the opportunity to make NA-analyses from fragments in the British Museum and took samples from three situlae of Cook’s group C (sample nos Defe 1-3, Figs 18–20), from two fragments belonging to stamnoid vessels (sample nos Figure 18 Fragment of ‘situla’ from Tell Defenneh, London, British Museum, GR 1888.2-8.65 (Vase B106.19)

Naukratis: Greek Diversity in Egypt | 149

Weber

Figure 19 Fragment of ‘situla’ from Tell Defenneh, London, British Museum, GR 1888.2-8.16+17

Figure 20 Fragment of East Greek ‘situla’ from Tell Defenneh, London, British Museum, GR 1888.2-8.20 (Vase B106.11)

Figure 21 Fragment of stamnos from Tell Defenneh, London, British Museum, GR 1888.2-8.42a

Figure 22 Fragment of stamnos from Tell Defenneh, London,The British Museum, GR Reg. No. 1888.2-8.44a

Figure 23 Fragment of amphora from Tell Defenneh, London, British Museum, GR 1888.2-8.25 (Vase B106.15)

150 | Naukratis: Greek Diversity in Egypt

Defe 4-5, Figs 21–22)91 and from a black-figured amphora (sample no. Defe 8, Fig. 23), all found at Tell Defenneh.92 Our aim was to check a) a possible assumed Rhodian provenance and b) whether all the fragments were made with the same clay paste, as suggested by their appearance. As comparanda a ‘situla’ fragment from Cook’s group B, found in Rhodes, with the depiction of a griffin (Fig. 24) and a Vroulian cup from Naukratis (Fig. 25) were also analysed. The analyses of the pieces from Tell Defenneh (sample nos Defe 1-5, 8) showed that all belong to one chemical group, named TD. Our analysis therefore proved that the group does not consist only of ‘situlae’ but also of stamnoi and an amphora. But it is a hitherto unknown group in the data bank of H. Mommsen and M. Kerschner, and not close to any of the known Greek groups.93 In addition, the other two pieces did not match any of the groups; they were made from different clay pastes. The analysis of the new ‘situla’ from Naukratis (sample no. Nauk 78; Figs 14–15) has shown that this piece, too, is a chemical single that does not correspond to the other samples of Group TD. The Bonn data bank does not yet possess a large corpus of Archaic vases from Rhodes or other places in the East Dorian region for comparative purposes. On the other hand, thanks to the databank we know where the ‘situlae’ could not have come from. It seems, however, very unlikely that in Egypt, in addition to the workshop of the vases of group QANN (very probably a workshop in Naukratis using local Nile silt),94 a second workshop for Greek vases was established in the third quarter of the 6th century bc using imported clay. A more likely place of manufacture would be the East Dorian region, where a workshop might have specialized in these vessels. Close contacts with Egypt could explain the Egyptianizing subjects on some of the ‘situlae’. The incorrect renderings on some of the vases could be better understood if the place of origin was outside of Egypt.95 Archaic ceramic material from the Dodecanese is still little known.96 Analyses of pottery of the East Dorian region will provide further information and one day, perhaps, the result that the ‘situlae’ from Tell Defenneh were manufactured at a place in the Eastern Doris will emerge. An amphora handle from among the material from Emecik (Old-Knidos) that was analysed very recently 97 falls into group TD.98 Our analyses have made it obvious that the ‘situlae’ come from at least three different as yet unlocated workshops (single from Rhodes, group TD and single from Naukratis). The statement of R.M. Cook is therefore still valid: ‘More analyses would be helpful’.99

Figure 24 Fragment of ‘situla’ from Rhodes, London, British Museum, GR 1868.4-5.78

Figure 25 Fragments of Vroulian cup from Naukratis, London, British Museum, GR 1888.61.569a-c

East Greek ‘Situlae’ from Egypt

Appendix: Neutron Activation Analysis Results H. Mommsen,A. Schwedt, S.Weber and M.R. Cowell The samples taken from sherds in the British Museum, Greek and Roman Department, London,100 have been analysed by Neutron Activation (NAA), a method routinely applied in Bonn (run P064, irradiated September 2003).101 The composition data have been grouped and compared with the Greek databank of more than 5300 samples.102 The program SEARCH was used, which works like a filter sorting out of a large databank all samples with statistically similar concentration values, taking into consideration experimental errors and also a possible ‘dilution’ of the clay paste by varying amounts of non-plastic parts poor in trace elements. This dilution factor (= best relative fit-factor to the mean concentration values M of the group) is given for each of the grouped samples (= factor in the tables). Table 1: Group TD, pattern unknown Element concentrations C in µg/g (ppm), if not indicated otherwise, average errors, also in percent of C, average values M and spreads, also in percent of M Sample Defe 1 Defe 2 Defe 3 Defe 4 Defe 5 Defe 8 av.meas.error in % av.value M spread in %

factor 0.903 1.008 1.069 1.063 1.021 0.933

As 4.61 4.04 4.11 6.42 5.00 3.53 0.11 2.4 4.6 1.0 22.

Ba 217. 180. 180. 115. 187. 115. 39. 23. 166. 42. 26.

Ca % 5.71 9.82 8.37 8.84 8.47 7.46 0.20 2.5 8.1 1.4 17.

Ce 51.0 47.9 46.5 44.3 48.5 49.0 0.51 1.1 48. 2.3 4.7

Co 52.7 56.5 60.8 54.6 54.4 63.8 0.22 0.4 57. 4.3 7.5

Cr 524. 530. 650. 674. 551. 761. 2.2 0.4 615. 96. 16.

Sample Defe 1 Defe 2 Defe 3 Defe 4 Defe 5 Defe 8 av.meas.error in % av.value M spread in %

factor 0.903 1.008 1.069 1.063 1.021 0.933

Hf 3.21 2.79 2.83 2.77 3.02 2.72 0.056 1.9 2.9 0.19 6.6

K% 1.66 1.43 1.55 1.62 1.65 1.42 0.028 1.8 1.6 0.11 7.0

La 25.1 23.1 22.3 21.1 23.5 24.9 0.091 0.4 23. 1.5 6.6

Lu 0.34 0.33 0.30 0.30 0.32 0.29 0.015 4.8 0.31 0.019 6.0

Na % 0.63 0.69 0.69 0.77 0.58 0.52 0.004 0.7 0.65 0.087 13.

Nd 14.9 15.0 17.5 9.99 11.7 15.9 1.9 14. 14. 2.7 19.

Sample Defe 1 Defe 2 Defe 3 Defe 4 Defe 5 Defe 8 av.meas.error in % av.value M spread in %

factor 0.903 1.008 1.069 1.063 1.021 0.933

Sm 3.15 3.09 2.79 2.59 3.00 3.01 0.017 0.6 2.9 0.21 7.2

Ta 0.80 0.66 0.64 0.66 0.73 0.68 0.030 4.3 0.70 0.060 8.6

Tb 0.58 0.52 0.48 0.42 0.49 0.51 0.043 8.5 0.50 0.053 11.

Th 8.70 8.05 7.76 7.28 8.16 9.59 0.060 0.7 8.3 0.80 9.7

Ti % 0.43 0.43 0.26 0.43 0.42 0.46 0.068 17. 0.41 0.068 17.

U 1.73 2.22 1.48 1.66 1.61 1.52 0.11 6.2 1.7 0.27 16.

Cs 5.52 4.87 4.76 4.20 5.94 5.15 0.081 1.6 5.1 0.61 12. Ni 829. 773. 1041. 935. 874. 1081. 50. 5.4 919. 121. 13. W 1.76 1.73 1.80 1.75 1.33 1.09 0.17 11. 1.6 0.30 19.

Eu 0.91 0.82 0.76 0.77 0.86 0.83 0.020 2.4 0.82 0.057 6.9 Rb 86.7 65.7 74.3 67.6 84.6 74.0 2.2 2.9 75. 8.6 11. Yb 2.08 1.96 1.82 1.86 1.94 1.89 0.056 2.9 1.9 0.089 4.6

Fe % 4.89 5.40 5.47 5.59 5.19 5.29 0.018 0.3 5.3 0.25 4.6 Sb 0.44 0.35 0.29 0.30 0.41 0.40 0.020 5.6 0.37 0.060 17. Zn 72.9 98.7 99.9 115. 84.4 90.9 2.4 2.6 94. 14. 15.

Ga 12.2 15.0 12.8 14.9 15.8 15.4 2.4 17. 14. 2.4 17. Sc 15.8 16.4 16.5 16.7 16.4 15.4 0.021 0.1 16. 0.48 2.9 Zr 107. 60.9 147. 91.7 132. 153. 24. 21. 115. 35. 31.

Naukratis: Greek Diversity in Egypt | 151

Weber Table 2: Raw concentration data of sample Rhod 20 (chemical single) Element concentrations C in µg/g (ppm), if not indicated otherwise, average experimental errors, also in percent of C Sample factor Rhod 20 1.000 av. meas. error in %

As 4.90 0.11 1.6

Sample factor Rhod 20 1.000 av. meas. error in %

Hf 3.36 0.068 2.3

Sample factor Rhod 20 1.000 av. meas. error in %

Sm 3.11 0.020 0.11

Ba 186. 1.0 0.5

Ca % 8.03 46. 0.11

Ce 51.5 10 1.2

Co 56.6 0.21 0.026

Cr 995. 3.4 1.8

Cs 6.07 0.63 0.020

K% 2.29 1.3 7.5

La 25.8 0.025 42.

Lu 0.32 1.2 10.0

Na % 0.31 0.11 2.5

Nd 16.4 0.3 2.2

Ni 859. 0.016 0.023

Ta 1.15 0.3 4.4

Tb 0.51 0.033 0.16

Th 9.73 2.9 6.3

Ti % 0.63 0.051 0.063

U 1.85 6.0 2.0

W 2.18 0.071 2.5

Eu 1.00 0.8 0.3 Rb 106. 3.3 1.8 Yb 1.90 0.5 2.4

Fe % 5.57 0.18 2.1

Ga 16.8 0.4 9.2

Sb 0.45 0.005 0.025

Sc 18.5 0.5 0.1

Zn 80.6 0.064 27

Zr 112 9.3 15

The sample numbers correspond to the following registration numbers in the British Museum, Greek and Roman Department: Defe 1 = fragment of situla, Reg. no. GR 1888.2-8.65 (Vase B106.19) (CVA British Museum [8] II. D. m pl. 4.3), Fig. 18 Defe 2 = fragment of situla, Reg. no.GR 1888.2-8.16 + 17 (Vase B106.12-13) (CVA British Museum [8] II. D. m pl. 7.1-2), Fig. 19 Defe 3 = fragment of situla, Reg. no. GR 1888.2-8.20 (Vase B106.11) (CVA British Museum [8] II. D. m pl. 8.2), Fig. 20 Defe 4 = fragment of stamnos, Reg. no. GR 1888.2-8.42a (CVA British Museum [8] II. D. m pl. 10.5), Fig. 21 Defe 5 = fragment of stamnos, Reg. no.GR 1888.2-8.44a (CVA British Museum [8] II. D. m pl. 10.4), Fig. 22 Defe 8 = fragment of amphora, Reg. no. GR 1888.2-8.25 (Vase B106.15) (CVA British Museum [8] II. D. m pl. 10.6), Fig. 23 Rhod 20 = fragment of situla, Reg. no. GR 1868.4-5.78 (CVA British Museum [8] II. D. m pl. 4.1), Fig. 24 The samples of the sherds of three ‘situlae’, two stamnoid vessels and fragments from an amphora make a new group (named as TD).To this group a sherd of an amphora handle from Emecik can be added and the data show a close relationship with three other samples also found in Emecik (group EMED).103 The sample of the ‘situla’ sherd found in Rhodes (cf.Table 2) is a chemical loner. It has high Cr and Ni values.104 The sample of the ‘situla’ from Naukratis (Nauk 78) is a single as well as the sample of the Vroulian cup (Nauk 59).

Illustration credits

Figs 1, 7, 9, 11, 13, 16 the author; Fig. 2 DAI Athens Neg. no. 1983/1014, photo: Hellner; Fig. 3 after Boehlau 1898, 17 fig. 14; Fig. 4 after Hölscher 1954, pl. 47 F 3; Fig. 5 after Kinch 1914, pl. 28, 8; Fig. 6 DAI Athens Neg.no. Samos 2305, photo: Wagner; Figs 8, 10, 17–24 the British Museum; Fig. 12 22nd Ephorate of Prehistorical and Classical Antiquities, Rhodes; Figs 14, 25 photo U. Schlotzhauer; Fig. 15 author after drawing U. Schlotzhauer.

Notes *

1 2

I should like to express my gratitude to the members of the staff of the Department of Greek and Roman Antiquities of the British Museum, London, namely A. Villing and D. Williams, who enabled me to study the material and provided me with information and photographs. I am also glad to have the opportunity to thank the 22nd Ephorate of Prehistorical and Classical Antiquities, Rhodes (M. Filimonos), the German Archaeological Institute, Athens (M. Krumme, B. Konnemann), and the University Museum, Philadelphia (A. Blair Brownlee) for the opportunity to study objects, for photographs and the permission to reproduce them in this paper. For help and various suggestions I should like to thank R. Attula (Greifswald), M. Kerschner (Wien), F. Meynersen (Mainz), H. Mommsen (Bonn), R. Posamentir (Istanbul), and U. Schlotzhauer (Berlin). This contribution emerged from the project ‘Griechische Kunst und Kunsthandwerk in Ägypten’ within the Sonderforschungsbereich 295 ‘Kulturelle und sprachliche Kontakte’ at the Johannes Gutenberg-Universität, Mainz, and I owe a great debt of gratitude to the heads of this project, U. Höckmann and D. Kreikenbom. Cook 1954, 29-37 pl. G.B.596-605 (with earlier literature); Cook 1960, 139-40; Boardman 1998b, 144 figs 303-4; Cook and Dupont 1998, 116-8. London, British Museum GR 1886.4-1.1311. During one of the handling sessions at the colloquium in London, Dyfri Williams showed me an unpublished piece of the lower part of the body of a closed vessel that in my opinion could come from a ‘situla’. It would be the only known example of this shape from this site; Petrie (1888, 62) stated in his list of situla-form vases: Naukratis = none, Defenneh = common). L. 10.3; W. 14.2; Th. of wall 0.35-0.6cm; very pale brown clay (10YR 8/3); two joining fragments, mended. The lower part is painted in a dull grey on the outside. Greek letters were incised upside down after firing; the painted surface is flaked away

152 | Naukratis: Greek Diversity in Egypt

3 4 5 6

7 8 9 10

11 12

13 14 15 16 17 18 19

where the incisions were made. The reading of the preserved letters is difficult: [?] M O [N?] [.] or [A?] M O [N?] [.]. Clairmont 1954/5, 85-141; Venit 1982, passim; Weber 2001, 127-50; Smoláriková 2002, 23-46. Petrie 1888, 62-3 pls 25-6. A good overview of the history of research on Tell Defenneh is presented by Carrez-Maratray (1999, 274-86). Cf. also Leclère 1997. Carrez-Maratray 1999, 275-9. The only non-Egyptian cooking pots are Levantine: cf. Maeir 2002, 235-46. On the importance of cooking pots as indicator of Greek presence in the Levant cf. Fantalkin 2001b, 84-6,116-26; Wenning 2001, 262, 267 fig. 3; Niemeier and Niemeier 2002, 238. Spencer 1999, 297 n. 9. Spencer 1999, 299-300. Smoláriková 2002, 70. East Greek gem: London, British Museum GR 1888.2-1.161 (Gem 321); Walters 1926, no. 321 pl. 6; Zazoff 1983, pl. 22.6. Terracotta antefixes: Cairo, Egyptian Museum; Maspero 1914, 528 nos 5570-1; Empereur 2003, 32 fig. 11b-c. Bronze bowl: Cairo, Egyptian Museum, JE 31665 and 25212; Bissing 1901, 62-3. Jeffery 1998, 355. 358 no. 51 pl. 70: 500–450 bc (?); Carrez-Maratray 2000, 165. 170. Petrie 1888, 62: labelled as ‘situla-form vases’ in the table, as ‘situlatype of vase’ in the text. Petrie (1891a, 55 fig. 40) even depicted both classes of vessels side by side. The derivation of the ‘situla’ from the Egyptian bronze situla was accepted by Walters 1893, 42 but contested by Zahn 1898, 51 n. 1, and Walter-Karydi 1973, 100 n. 23. Cf. in general on Egyptian bronze situlae: Bissing 1901, 7-58; Lichtheim 1947, 169-79; Green 1987, 66-115; Nicholson 2004, 7-9. Teeter 1994, 259-63; on the use of the Egyptian bronze situla see most recently Bommas 2005, 257-72, esp. 264 with further literature. Montet 1928, 254 pl. 153 no. 965; Woolley 1921, 119 pl. 2.1-3; Stager 1996, 69-70; Frankel and Ventura 1998, 39-55; Kamlah 1999, 163-90. From the Toumba Cemetery at Lefkandi in Euboia: Popham et al. 1982, 238 fig. 8. p. 239 pl. 33 a and h; Popham and Lemos 1996, pl. 132; 143 a (T.42,17 = LPG or SPG I). f (T.70,17 and 20= LPG). Cambridge, Fitzwilliam Museum; H 12cm; Pendlebury 1930, 91-2 pl. 3 no. 227. Gàbrici 1927, 357, 359 fig. 154b: undecorated miniature situla, H with handles about 6.0cm. Matthäus 1985, 226-8. Bronze situla, Vathy, Archaeological Museum; AR 1983/4, 59 fig. 114;

East Greek ‘Situlae’ from Egypt

20 21 22 23

24 25

26 27

28 29

30

Kyrieleis 1991, 129 pl. 29, 2, found together with other votives and ritual pottery in a well in the sanctuary of Hera that was given up in the beginning of the 6th century bc. The shape of the situla corresponds to types I or III in Lichtheim 1947, 174-5 pl. 4.10, 16, 17. Below the rim are five ornamental bands (two friezes with triangles, one with floral ornaments and three friezes with triangles. The middle register is a figural frieze depicting a standing worshipper wearing a long tunic in front of Amun-Min (or Amenope), a rectangular altar, Horus or Haroeris with ankh-sign, Isis with wadjsceptre and ankh-sign, Nephthys with wadj-sceptre and ankh-sign. The space in front of the heads of the gods is left empty, like on the example in Fribourg, Switzerland (Page Gasser 2001, 139-42) and on some situlae from the animal necropolis at Saqqara. The next register is decorated with a scale pattern, an unusual feature in Egyptian situlae. A parallel for the scales beneath the figural scene is only found on the example from Lefkandi (Popham and Lemos 1996, pl. 132, T. 42, 17). The rounded bottom of the vessel is decorated with lotus leaves. Cf. e.g. the case of the Nikosthenic amphora. The potter Nikosthenes produced this Etruscan shape for export to Etruria. Tosto 1999, 95-8; Shapiro 2000, 313-37. Hilgers 1969, 77-9, 282-3. Zahlhaas 1971, 7; Hurschmann 2001, 605. Greek metal situlae cf. Zahlhaas 1971, 109-12 fig. 1 D (her ‘Form D’ with rounded bottom and without foot in some respects recalls Egyptian situlae); Gauer 1991, 110-23. Etruscan bronze situlae from Spina cf. Hostetter 2001, 19-34. Kastelic 1964, 18; Kromer 1969, 72-80; Frey 1969, 83-7; Megaw and Megaw 1989, 37-9. Late Geometric pyxis situla from Smyrna, cf. Özkan 1999, 36 no. 65. Orientalizing pyxis situlae with palmette decoration from Klazomenai, http\: klazomenai.tripod.com/resim04-01.html and Hürmüzlü 1995, 61 pl. 22 fig. 82 pl. 23 figs 83-4. Izmir, Archaeological Museum, Inv. 10946, Özkan 1999, 36 no. 65. Istanbul, Archaeological Museum, Inv. 2894, H 36cm; first half of the 6th century bc, A: two confronted bulls in the handle zone, B: goat and lion; below the wave pattern typical of Lydian art. The connection was made by Cook 1954, 32. Good illustrations in Akurgal 1961, 151-3 figs 102-3 and Amandry 1962, 54, 68 Beil. 14.2 [side B]. Walter-Karydi (1973, 100 n. 23) denied the possibility that the ‘situlae’ could depend on a shape like this. A vase close in shape to the krater in Istanbul is in Rhodes, Archaeological Museum, Inv. 14222, H 15, 5cm; Dm mouth 21cm; only one half of the vessel is preserved. It was found in a grave in Nisyros, cf. Jacopi 1932/3-41b, 522-23 figs 501, pl. 2. The vessel has two broad handles attached directly to the rim. Figure-decorated ‘metopes’ feature on side A a dog, on side B a bird. Jacopi (1932/3-41b, 522) considered a possible local manufacture of the vessel and called the shape ‘specie di corta situla’. Cook 1954, 32. From Kamiros: Jacopi 1931/9, pl. 8.139, 159. From Samos: Boehlau 1898, 17 fig. 14. From Nisyros: Jacopi 1932/3-41, 51315 figs 41-3; 519 fig. 48. From Histria: Lambrino 1938, 41-9 figs 96-7. Schäfer 1908, 127. Shapes with a bulging bottom have a long tradition in Egypt and were manufactured in various materials, cf. Hölscher 1954, pl. 47; Radwan 1983, pls 64-6 nos 347-60 (small vessels, the highest about 30cm high, mainly without handles, from Middle Kingdom onwards, many pieces from New Kingdom for wine or beer); French 1988, 82 fig. 1; on Egyptian storage jars cf. Aston 1996, 45, 231 fig. 129.10 (from Matmar), 64 ‘group 37’, 301 fig. 199a, datable to the 12th–9th centuries bc; 65 ‘group 42’, 303 fig. 201d; meat jars 66 ‘group 50’ 307 fig. 205f; 76 ‘group 30’ Phase III south 8th–7th centuries bc, 323 fig. 221c; 326 fig. 224f. The vases might have been used for preparing kykeon, a mixture of wine, herbs and cheese: in Ialysos, grave 183 a situla (Rhodes, Archaeological Museum, Inv. 10641) was found together with Attic and local pottery and a bronze cheese grater (Rhodes, Archaeological Museum, Inv. 10642, Jacopi 1929, 192 fig. 186; Jacobsthal 1932, fig. 1; Jacopi 1933, II D m pl. 1.4-5), like the ones from Pyres 13 and 14 at Lefkandi, Toumba cemetery: Popham and Lemos 1996, pl. 146c (Pyre 14,18 = SPG IIIa) and d (T.79B,2 = SPG II) and T.79A (three graters, SPG II). One of these cremations may be connected with a warrior and therefore the grater could have been used in the funeral feast, cf. Popham et al. 1982, 213-48, esp. 240-1. Graters are not connected so much with female burials but mainly with male ones. Perhaps they were used during the symposion to mix wine with grated cheese or to prepare the kykeon, a mixture of Pramnian wine, barley groats, grated cheese and honey, cf. Homer,

31 32

33 34 35

36

37 38

39

40 41 42

43 44

45

46 47

Od. 10, 243-4; 11, 624. 628ff.; Ar. V. 938; Ar. Lys. 231; Kroll 1919, 1494; Jacobsthal 1932, 1-7, 6; Liddell - Scott: 1837 s.v. turo/knhstij; Bruns 1970, 2, 15, 37. Other graters, votive or household utensils, are known from Samos (Gehrig 1964, 9-10 no. 50-4; p. 97-8), Perachora (Payne 1940, 182 pl. 81.11 and Olynthus (Robinson 1941, 191), the Enodia sanctuary in Pherai (Kilian 1975, pl. 94.33). Further cheese grater in animal form: Hoffmann 1964, no. 12. Kinch 1914, 105-6, 125-6; 1929, figs 186-9. 198; Cook 1954, 29-37; Walter 1968, no. 591 pl. 115. Cook and Dupont 1998, 116-18. Ketterer 1999, 217 fig. 9, 221 cat. no. 6: two small joining fragments from the body of an open vase in Fikellura style (MileA II); sphinx and another figure with wings. No rim or foot fragments are preserved, therefore it is doubtful whether the fragments belong to a vase of ‘situla’ shape. Walter 1968, 57, 116 nos 435-7, pl. 79. Ýren (2003, 141, 184 no. 284) attributed one of these rim fragments mentioned by Walter 1968 (no. 437) to Aiolis because of the angular rendering of the loop ornament. Three unpublished rim fragments were found in Burgas (Datça), Özer 1998, 36-9. I owe information about these fragments R. Attula. Rhodes, Archaeological Museum, Inv. C/310 A-H, from Ialysos, clay: 7.5YR 7/4 (pink), on fragment c/310 H parts of a graffito are preserved: H L; unpublished black-figured fragments of neck or body of a straight closed vase from the stipe votive of the sanctuary of Athena Polias, mentioned by Vita 1985, 368 (‘[…] frammenti di una situla tipo Daphne con raffigurazione di guerriero […]’) and by Cook and Dupont 1998, 205 n. 3 and 5. The decorative system does not correspond to the ‘situlae’ from Tell Defenneh: two figural friezes are separated by a hook meander. In the upper frieze warriors are depicted, in the lower chariots. There are no fragments from rim or base, so the shape could be another type of vessel, e.g. a slender hydria or a ‘situla pyxis’. Five examples in Istanbul from Vroulia: Kinch 1914, 105 no. 11 pl. 23.12 (from the main sanctuary); 125-6 nos 2-5 pl. 28.8, 9, 11 (from houses of the settlement). Cook 1960, 140 supposed that these vessels may be local Vroulian. Philadelphia, University Museum, E147.3: Schaus 1995, 25-6 pl. 11.1. Cf. Kinch 1914, pl. 28.11. Samos, Heraion, Inv. K 1590. Our Fig. 6 shows an old state of preservation. Today, the ‘situla’ is broken again into fragments (14, two still glued together). Some of the parts, still visible on the photograph, are missing, for instance the head of the water bird. The surface is much worn. The old illustration was chosen because it shows best the original shape of the vase. 1. Rhodes, Archaeological Museum, Inv. 10641, H 35.5cm; Dm mouth 18.2–18.5cm; Dm foot 11.95–12.2cm; reddish yellow clay (Munsell 7.5YR 8/6) with many small dark and fewer white grits; A: woman to right, B: stylized tree with birds on the volutes. Below two broad registers with lotus flowers and palmettes, 1929, 192 fig. 186; 1933, II D m pl. 1.4-5.– 2. Rhodes, Archaeological Museum, Inv. 10773, H as restored 39cm; Dm mouth 18.3cm; pink clay (between Munsell 7.5YR 8/4 and 7.5YR 7/4); A: horseman to right, B: warrior to right; two lower registers with carelessly drawn lotus flowers and palmettes, 1929, 204-7 fig. 198. Cook and Dupont 1998, 114-5. This feature is not only found with situlae but also with other vessels, cf. Johnston 1993, 351 cat. no. 53 (amphora, Laconian ?), 353 fig. 5.A. The local Egyptian ceramic repertoire consists of many lids in various sizes, see Petrie 1888, pl. 36, but only one Greek lid of East Greek grey ware was found (London, British Museum GR 1888.2-8.139, unpublished). Petrie 1888, 62-8; Cook 1937, 227-37; 1954, 31; Boardman 1958, 4-12; 2000, 133-53; Schaus 1995, 27; Hoffmann and Steinhart 1998, 49-61; Carrez-Maratray 1999, 283-6; Schlotzhauer and Weber 2005, 89-91. Sphinx: Cook 1954, pl. G.B. 598.3-4. The winged, crouching female sphinx is a Greek type, in Egyptian art the sphinx is depicted as a reclining male sphinx, symbolizing either the sun god Re-Harmachis or the pharaoh. On the differentiation between Greek and Egyptian sphinxes see Höckmann and Winkler-Horacvek 2005, 90-6. London, British Museum GR 1888.2-8.3; preserved H 6.7cm; pink clay (Munsell 5YR 7/4) with a wash (Munsell 10YR 7/4); Petrie 1888, 62 pl. 26.1; Cook 1954, pl. G.B. 597.2; Weber in Schlotzhauer and Weber 2005, 88, 110 fig. 17. Rhodes, Archaeological Museum, Inv. 10696, 1929, 50-3. Cf. also the Egyptian objects from the stipe votiva di Kamiros: Jacopi 1932/3-41, 317 fig. 57 (falcons from faience). Cf. Bailey, this volume.

Naukratis: Greek Diversity in Egypt | 153

Weber 48 Other hieroglyphs or Egyptian signs are graffiti scratched after firing on two fragments of Greek trade vases from Tell Defenneh: London, British Museum, Department of Ancient Egyptian and Sudan, AES 23761 (Petrie could read three signs: k m khu. I can confirm two signs: Gardiner G 17 [m] and N 27 [khu]); AES 23762 (shuttle of Neith = Gardiner list R 24/25 25); on fragments of two trade amphorae, kept in the same department, are painted demotic signs: EA 23775; EA 22343 (Johnston Fig. 16). All vases are unpublished but mentioned by Petrie 1888, 74 without distinguishing Egyptian or Greek fabric. A further Demotic inscription in black ink is on a Levantine amphora from Tell Defenneh; possibly a votive inscription: London, Petrie Museum, UC19250, cf. www.petrie.ucl.ac.uk under the Museum number. 49 London, British Museum GR 1888.2-8.5 (Vase B106.1); H (restored) 16.4cm; Dm lip (restored) 19.3cm; light yellowish brown clay (Munsell 10YR 6/4); Petrie 1888, 62 pl. 26.3; Walters 1893, 88 no. B106.1; Cook 1954, pl. G.B. 598.6; Weber in Schlotzhauer and Weber 2005, 89, 110-12 figs 20-1. Petrie (1888, 62) notes: ‘[…] the lotus group between the two fighters is again not a Greek lotus pattern, but like the lotus flowers on piles of Egyptian offerings. It cannot be doubted that this was painted with living Egyptians under the artist’s eyes.’ 50 Athens, National Museum, Inv. 9683; ARV² 554.2. 51 Decker 1987, 90-5; Decker and Herb 1994, 564-71; Weber in Schlotzhauer and Weber 2005, 89-92. There is only a later written source for ritual club fighting in the Saite Period: Hdt. 2.63. 52 On Greek athletic nudity cf. Decker 2003, 51-2. In Egyptian art only children or youths performing sports or playing are rendered naked, cf. Decker and Herb 1994, pls 291, 302, 314-5, 340, 345-6, 348. 53 Decker 2003, 55. 54 Philadelphia, University Museum of Archaeology and Anthropology, Inv. 29.71.189. Cf. the illustrations in Schaus (1995, pl. 11.2-5) and Weber (2001, pl. 21.3 a-b). 55 The object in the left hand could be a flower or a small stone vessel like the jar for holy oil, cf. on the vessel Arnold 1977, 485-6 no. 41. 56 On Egyptian shawls and dress: Bianchi 1978, 95-102. 57 This cipher corresponds to the hieroglyphic signs A 21 (sr or smr = official or noble) and A 107 (with outstretched right hand). Gardiner 1973, 444. 58 Hassan 1976, passim; Steuernagel 1991, 35-48; Fehr 2000, 139. 59 Cf. the representation of the mayor of Thebes, Sennefer, 18th Dyn., in the pillared hall of his tomb, Strouhal 1996, 165 fig. 182. 60 A quadrangular sign could stand for a building or a precinct, cf. the sign-list by Gardiner 1973, 492-8 no. O. The reading of the sign on the ‘situla’ as a pseudo-hieroglyph was doubted by Hoffmann and Steinhart 1998, 57. For incorrectly written but legible Egyptian signs on Greek vases cf. the cartouches on the amphora in Basle, collection Cahn, HC 1175/London, Petrie Museum (Bailey Figs 1–5), cf. Decker (2003, 49-56 with reference to older literature) and Bailey in this volume. Decker (2003, 56) hypothetically put forward, like other scholars, an Egyptian place of manufacture for this vase. This hypothesis has now been refuted on the basis of clay analysis, unless, of course, one assumes the import of clay. Cf. the paper by Bailey in this volume. 61 Schaus 1995, 26-7. 62 Brommer 1984, 178-84; Shapiro (2000, 318-37) on the stamnoi and kantharoi of the Perizoma group designed for the Etruscan market. Special commission for the Persians in Egypt: Kahil 1972, 271-84. Lezzi-Hafter (1997, 353-69) published two Attic red-figured mugs with special shape and subjects for a Thracian clientele. 63 Boardman 1958, 4-12. 64 Carrez-Maratray 1999, 284-6. 65 Walters 1893, no. B 104; Cook 1954, 32-3 pl. G.B. 596; Walter-Karydi 1973, no. 1060 pls 135-6; Touchefeu-Meynier 1997, 149 s.v. Typhon no. 11; Tempesta 1998, 71-2, 147, 172 no. 80, pl. 38.1-2. 66 Schmidt 1916-24, 1426-54; Schefold 1978, 53-4; Schefold 1993, 196-9.

154 | Naukratis: Greek Diversity in Egypt

67 Kunze 1950, 82-8. 68 Touchefeu-Meynier 1997, 148-9 s.v. Typhon nos 1-10. 69 Griffiths 1970, 259; Kranz 1934, 114-15; Kolta 1968, 161-8; Velde 1986, 816-17 s.v. Typhon. 70 Erman and Grapow 1982, 262 (tbh). 71 Velde 1967, 1-26; Donadoni 1981, 115-23; Velde 1984, 908-11; Wilkinson 2003, 197-9. 72 Hornung and Badawy 1975, 350-2 s.v. Apophis. 73 Petrie 1888, 74 § 71; Petrie 1891a, 57-8; Leclère 1997. 74 Cook 1954, 32-3. I had the opportunity to study the vase in 2001. 75 Cook 1960, 140. 76 Cook 1960, 140; Walter-Karydi 1973, 90-1 nos 1061-63, pl. 136. 77 Petrie 1888, 62. 78 Cook 1954, 32; Cook 1960, 140. 79 Price 1928, 9. 80 Walter-Karydi 1970, 8 with n. 27 and 31; Walter-Karydi 1973, 100 n. 23. 81 Schaus 1995, 25. 82 Carrez-Maratray 1999, 284. 83 Möller 2000a, 145. 84 Smoláriková 2002, 64. 85 The question of clay imports to Naukratis is also addressed by Williams (this volume), Bailey (this volume), and Schlotzhauer and Villing (this volume). 86 Petrie 1888, 63. 87 Cook 1960, 140 put forward as the place of manufacture the ‘stratopeda’, or camps, given to the Ionian and Carian mercenaries by Pharaoh Psammetichos I. 88 Lemos 1991, 191-4; Mannack 2002, 93-4. 89 Oxford, Ashmolean Museum, Inv. 1925.608 a-c; Beazley et al. 1931, pl. G.B. 401.25-7; Jones 1986, 669-70. 90 As discussed by P. Dupont and A. Thomas in this volume; cf. also Dupont 1983, 29. 91 Cf. similar stamnoi from Ialysos: Rhodes, Archaeological Museum, Inv. 1320: Jacopi 1933, II D m pl. 1.3, Inv. 12062: Jacopi 1934, II D m pl. 3.5, Inv. 10487: Jacopi 1934., II D h pl. 8.3. 92 Cf. Mommsen et al., this volume. 93 Cf. Mommsen et al. Fig. 1. 94 Cf. the papers by Schlotzhauer and Villing and by Mommsen et al. in this volume. 95 As it is the case with the amphora in Basel with the cartouches of pharaoh Apries, Basel, Cahn HC 1175/London, Petrie Museum UC 30035A-B, cf. the paper by Bailey (this volume). 96 Schefold 1942, 128-30; Walter-Karydi 1998, 287-96; Berges and Tuna 2000, 198, 212. 97 R. Attula pers. comm. 98 To be published by R. Attula. 99 Cook and Dupont 1998, 118. 100 The samples were taken by Michael R. Cowell, Department of Conservation, Documentation and Science, British Museum, London. 101 On the method in general cf. Mommsen et al. 1991, 57-64 and Mommsen 2005, 40-1. 102 Cf. samples of various East Greek wares in Akurgal et al. 2002 and the paper by Mommsen et al. in this volume. 103 The samples 17, 23 and 117 from Emecik belong to group EMED. Samples 17 and 117, both body fragments, are published by Attula 2006, 130-1 cat. no. 245 fig. 26 pl. 67.8 and 140 cat. no. 287 fig. 30 pl. 75.2. Cf. also Mommsen, Schwedt and Attula 2006, 199-204. 104 NAA carried out in Bonn and by other scientists showed that an unusually high level of Cr and Ni is present in Mycenaean and later pottery made in Rhodes, a feature hitherto not known from other workshops of the Eastern Aegean, thus hinting at a Rhodian source, cf. Jones 1986, 669-70; Leonard et al. 1993, 118 with n. 36; Marketou et al. (forthcoming).

The Apries Amphora – Another Cartouche Donald M. Bailey

Abstract Two joining sherds with a cartouche of the Egyptian Pharaoh Apries were recently noted in the collections of the Petrie Museum of Egyptian Archaeology in London. They have been shown to fit into a well-known, but fragmentary East Greek amphora with two of the names of Apries, a king of the 26th Egyptian Dynasty, painted round its neck. The Petrie sherds have ‘Thebes’ written upon them in pencil, thus suggesting the findspot of the vase itself.* As the present volume emphasises, and the many publications of the last decades have shown, the study of East Greek vases is in a healthy state, but there are still groups that are difficult to place, and the vase under discussion, bearing the names of the Pharaoh Apries, falls into one of these (Fig. 1).1 Although the vase itself has not been analysed, a joining sherd in the collections of the Petrie Museum has recently been traced, and due to the kindness of the Curator, Stephen Quirke, has been examined by Neutron Activation Analysis. It has been found to belong amongst Mommsen’s Pattern B, of the ‘bird bowl workshops’ and is probably North Ionian.2 The Apries Amphora, a neck-amphora, assembled from fragments, has black-figure decoration. On one side confronted boxers flank a prize dinos on a stand; a bird, probably a raptor,

standing on an ionic column, survives behind one of the boxers. On the other side of the pot the heads of two standing women remain, behind whom are plants, a tree or a bush. Only one handle is present and the lower body is almost completely gone, together with the base. Around the neck of the vase, interrupted by the upper springs of both handles, is a band of four cartouches. First published by John Boardman in 19803 several subsequent articles have mentioned and discussed it.4 Very full descriptions of the vase, defining the subjects of the figural decoration and the hieroglyphs, and also discussions of the vessel’s date, are given by Thomas Schattner and Norbert Dürring, and by Friedhelm Hoffmann and Matthias Steinhart.5 The Petrie Museum fragment, UC30035a-b, consists of two joining sherds from a neck-amphora with a thickened squared rim (diam. 17.0cm: Fig. 2). The fabric is orange with a buff slip outside and on top of the rim. There is black glaze on the outside of the rim and in a narrow band on the top of the inner rim. Below this band, inside the mouth, is a thinned black glaze merging with a red glaze; at about 5.5cm below the rim inside are two narrow bands of white ceramic colour, not quite parallel with the rim. On the outside, below the rim, a cartouche is painted in black-glaze medium, with hieroglyphs laid over in white fired-on ceramic colour (Figs 3–4). Both fragments have

Figure 1 The Apries Amphora, Cahn Collection HC 1175, with the joining fragments Petrie Museum UC30035a-b

Naukratis: Greek Diversity in Egypt | 155

Bailey

Figure 2 Section drawing of the Petrie fragments, Petrie Museum UC30035a-b

Figure 3 The Petrie fragments, Petrie Museum UC30035a-b

‘Thebes’ written in pencil. The new cartouche was noticed when the writer was browsing the Petrie Museum website and he was encouraged by John Boardman and Alexandra Villing to put this short note into the present publication. It seemed probable that the fragment was either from the Apries Amphora or from one closely resembling it: Alexandra first pointed out the place where it fitted the Apries Amphora, where part of the cartouche and a part of a r‘ sign matched. The cartouche band (Fig. 5) encompasses the Throne Name and the Birth Name of the Saite king Apries, one of each on the front and on the back of the neck of the vase, between the handles. The new cartouche is on the side with the major (remaining) scene of the boxing contest and (despite considerable errors on the part of the painter) is the Throne Name of Apries. The painter appears to have inserted titles (with a lavishly loaded brush), not only between cartouches, but also, in one surviving example, between a cartouche and an upper handle-spring. Thus, reading right to left, we have probably the lower stroke of the nt-r of the Good God title between the handlespring and the new cartouche with the Throne Name Haaibre; after this is the Son of Re title and the Birth Name Wahibre, followed by a Good God title (signs transposed) in front of the scar of the other handle-spring; on the further side of this handle-spring scar is another Good God title (again with transposed signs) in front of the Throne Name; following the latter name is the Son of Re title in front of a largely lost cartouche, no doubt containing the Birth Name of Apries; between this cartouche and the handle spring it seems very likely that there was another title. I am grateful to Kate Morton

for adapting Norbert Dürring’s illustration of the layout of the cartouche-band. It can be seen (Figs 3–4) that the new Throne Name is less accurately depicted than that in the cartouche already on the vase, and in both Throne Names the letter h. is omitted. One useful aspect of the identification of the new fragment with the Apries Amphora is the probability that the vase comes from Thebes. One cannot be sure that the pencilled words ‘Thebes’ are in Petrie’s hand; however, comparison with documents written by him has convinced me that he did indeed write on the Petrie Museum’s sherds. Petrie collected and purchased anything that was offered to him that he felt was of interest and that he could afford, and he would be certain to regard a royal cartouche on a vessel as worth acquiring. Boardman and Sabine Weber both have mentioned the presence of a small number of Greek painted pots in Thebes.6 It is not known when the fragmentary Apries Amphora was found, but the new fragment must have been recovered at the same time, presumably before Petrie abandoned Egypt in 1926. The Egyptian dealer who acquired it may not have received all the sherds, the finder retaining the new fragment for disposal elsewhere: there are several possible scenarios, but it eventually came into Petrie’s hands. Other as yet unanswerable questions include where and when the vase was manufactured, and why it had royal cartouches. Dyfri Williams, in this volume, has discussed the difficulties (with Chian pottery, found sometimes in large quantity outside Chios) of deciding whether pots or potters moved about, with the attendant possibility that raw clay moved with the potters.

handle spring

handle scar

handle spring

handle spring

Figure 5 Cartouche band adapted from Norbert Dürring’s version.A: on vase; B: transcribed; C: reconstructed

156 | Naukratis: Greek Diversity in Egypt

Figure 4 Cartouche on Petrie fragment UC30035a

handle spring

handle spring

handle spring

The Apries Amphora – Another Cartouche He tends to favour a sophisticated and well-ordered system of trade, with merchantmen crossing wide areas of the eastern Mediterranean, carrying safely, on the whole, shiploads of desirable finewares, some of which had been specially ordered by individuals in Egypt for votive purposes in particular shrines. He cannot wholly reject the idea that potters from Chios used Chian clay while resident in Naukratis. The small North Ionian grouping into which the Apries Amphora falls seems even less likely to have been made in Egypt. Many of those discussing the date within the 26th Dynasty of the Apries Amphora believe that it was not made before Apries’ deposition by Amasis in 570 bc; his death in 567 bc does not add much to the period that his name might or might not have been used officially, although Leahy in 1988 has shown that it was employed after Amasis claimed the throne. Indeed there may have been parallel rule of the two kings for three years after 570 bc.7 Stylistically, dates of about 550-540 and as late as the 530s bc have been suggested for the vase, but I have not the expertise to know whether they are plausible; some of the comparanda brought forward, of widely differing sources, materials, decoration and dating, however, show signs of desperation. Decker 2003 prefers a date before the death of Apries and discusses at length the possibility that the vessel was a prizevase for the sport shown on it, but cannot arrive at certainty. A vase from Klazomenai with a black-glaze neck and scale patterns on the body has a closely similar body shape and rimform, and also has a moulding at the base of the neck.8 It is from a context of c.520–500 bc. A small number of black-figure sherds from Klazomenai published by Özer are perhaps stylistically similar to the Apries Amphora and may date between 570 and 560 bc; the author also mentions amphorae of the same shape from a kiln of the middle of the 6th century bc.9 The cartouches, while not wholly accurate, are sufficiently well rendered for the names of Apries to be read. A speculative potter in North Ionia might have painted them on a pot for export to Egypt as an inducement to persuade rich Egyptians to purchase it (and others similar?). He may not have realised (or even cared) that the king was dead at the hands of his still living successor when he copied the names from an inscribed object, perhaps a sealing from a commercial import or from an Egyptian object dedicated at a Greek shrine.10 A band of cartouches was what he wanted, not a particular pharaoh. This procedure may have been followed whether the vase was painted before or after 567 bc. The presence in Thebes of a vase with Apries’ names may not have been as dangerous as may be thought. The real ruler of

the Thebaid at this time was not Amasis, but the God’s Wife of Amun, the Divine Adoratrice Ankhnesneferibre, who reigned for over 60 years (586–525 bc).11 Daughter of Psammetichus II, she was Apries’ half-sister; Budge12 regarded her as his full sister. Had she known of such vessels as the Apries Amphora, she may have liked him enough to have tolerated the use of his name well after his death. The names of previous pharaohs were legion throughout Egypt and few were, as Hatshepsut was with Tuthmosis III, subjected to damnatio memoriae. There is little evidence that Apries’ name was erased by Amasis,13 who buried him at Sais with full honours. It would seem likely also that Amasis, possibly legitimised on the throne of Egypt by marriage to the God’s Wife, was more often than not in his northern capital of Sais, and may never have come across vases such as this, particularly if they were in a batch that reached Thebes. Illustration credits

Fig. 1 the British Museum; Fig. 2 drawing D.M. Bailey; Fig. 3 photo D.M. Bailey, ed. C.M. Johns; Fig. 4 photo A. Villing; Fig. 5 drawing K. Morton.

Notes *

1 2 3 4 5 6 7 8 9 10 11

12 13

Concerning the names within the cartouches, I am most grateful to Jeffrey Spencer and Stephen Quirke for discussing them, and to Stephen for allowing me to photograph the fragments. Alexandra Villing had much advice to offer about East Greek painted vases. I am very grateful to David Cahn for making the Apries Amphora available for examination. Vase HC 1175, Kreuzer 1992, 52-4. It was acquired by the late Herbert Cahn on the Paris art market and is said to be from Egypt. Mommsen et al., this volume, TbEgy 1. Boardman 1980, 138-9. Boardman 1987, 147-8, fig. 4; Cook 1989, 167; M. Weber 1995, 163-70; Gill and Vickers 1996, 7; Cook and Dupont 1998, 107; Decker 2003, 49-56. Schattner and Dürring 1995, 65-93; Hoffmann and Steinhart 1998, 49-61. Boardman 1958, 4-12; Boardman 1980, 137-8; Weber 2001, 139-40. James 1981, 736. Ersoy 1993, 539-40, pls 40-41; Ersoy 2004, 58, fig. 17a. Özer 2004, 201-2; 215, n. 15. I am grateful to Bilge Hürmüzlü for pointing out this reference. Cf. most recently Ebbinghaus 2006, 189-202. A selection of references to Ankhnesneferibre includes Budge 1909, 224, pl. XXX; James 1981, 733; Leahy 1988; Clayton 1994, 197; Elwood 1994, 93, 103; Lloyd 1983, 303; Dodson 1995, 194; Mys´liwiec 2000, 130-31; Dodson 2002, 186: only Budge and Elwood refer to her marriage to Amasis. Ankhnesneferibre’s splendid sarcophagus, only slightly usurped, is in the British Museum (EA 907): Budge 1885. Budge 1885, p. x. Leahy 1988, 198.

Naukratis: Greek Diversity in Egypt | 157

The Greeks in Berezan and Naukratis: A Similar Story? Richard Posamentir Abstract Ionian Greeks founded hundreds of colonies in many different places and along various coastlines, always having to face problems such as unknown territories, climate or local cultures. Even though these cities and settlements to the west, north or south of Ionia lie far from each other, several similar phenomena can be observed in most of these places – which means that similar questions have to be answered in order to understand the processes taking place at every new spot where Greeks arrived. Among these colonies there are some places which have been investigated archaeologically for a long time, providing us with countless artefacts of Ionian style that are among the finest ever found, and giving us the chance to learn more about their Ionian mother cities as well. Berezan and Naukratis are both sites of major importance in this respect and, by showing remarkable parallels, offer a most welcome opportunity to study early Greek colonies in different parts of the world. On the occasion of a conference held at Mainz in 1999, M. Kerschner1 offered a comprehensive account of the essential problems one encounters when studying Archaic Ionian pottery found in Naukratis. Assessing his contribution one easily reaches the conclusion that in fact similarities between Naukratis and Berezan are not only limited to the involvement of Ionian people in their foundation history – but continue up to the present time. They also relate to the long history of modern excavations at both sites as well as the structural phenomena which will be focused on in this contribution.2 As is the case with Naukratis, Berezan (most likely once called Borysthenes) has long been of major interest to archaeologists, and numerous generations have already explored the site, while simultaneously great damage was done to the area by non-archaeological interference; in Berezan this was mostly due to military action.3 Apart from the usual loss of archaeological information caused by early excavation and conservation techniques, which are common at every site of the Greek and Roman world, the island suffered additionally from subsequent changes of the sea level, submerging substantial parts of one of the earliest Greek colonies along the northern Black Sea shore. Furthermore, the excavated material has been distributed among different museums, making it difficult to present a comprehensive study and quantification of the pottery to the public – again most unfortunately a parallel with the situation regarding Naukratis. Museums in St. Petersburg, Moscow, Odessa, Cherson and Ochakiv hold collections of various sizes – the Halle and Bonn collections (with smaller amounts of pottery from excavations of Ernst von Stern, one of the first excavators of this site) must also be mentioned. Additionally, frequent changes in scholarly approach and applied techniques must necessarily lead to a high degree of diversity concerning the choice of which shapes, wares or fragments of pottery are

collected and which are left on the site. This results in unintentional and unpredictable differences between the collections mentioned above. Thus also the choice of fragments already published most naturally depended on personal opinion or rating, or was sometimes made with the aim of supporting certain theories about the site. In this contribution more emphasis is put on those fragments which have never before been published or have even escaped the attention of scholars working on this material. Excavations in Berezan continue up to the present,4 and fortunately a lot has changed since the early days. Still – working on a limited amount of material, such as the Archaic Ionian pottery5 kept in the State Hermitage in St. Petersburg, some problems immediately become apparent: the thousands of sherds found in cemeteries and in the settlement between 1963 and 1991 represent fine-ware pottery only; almost no ‘Greek’ coarse-ware (except transport amphorae) was found6 or collected if not showing dipinti or graffiti. The lack of coarseware (as well as unpainted fragments of fine-ware) should already be an obvious warning not to expect a strict adherence to proportions in the collection of certain wares – the so-called ‘schwarzbunte’ pottery, mostly represented by jugs which are already well-known from other Mediterranean sites,7 can serve as an example: about 50 different shoulder parts (Fig. 1) with applied red and white colour found their way to the Hermitage and other museum collections, while undecorated parts of these jugs are almost completely missing from the earlier years – which means that they were obviously previously sorted out at the site. Theoretically the loss of these undecorated parts would be of minor importance since information on the total number of different vessels of this type might certainly seem of higher value from a statistic point of view – but during the processing of all these sherds it increasingly emerged that, surprisingly, a lot of fragments could at least partly be assembled and that a

Figure 1 Shoulder parts (inv. nos. B65–36; B73–97; B76–14; B80–29) of four socalled ‘schwarzbunte’ jugs

Naukratis:Greek Diversity in Egypt | 159

Posamentir Figure 2 Places of origin of Archaic Ionian pottery found on Berezan

Figure 3 Places of origin of Archaic Ionian pottery from the 7th century BC found on Berezan

Figure 4 Places of origin of Archaic Ionian pottery from the 6th century BC found on Berezan

160 | Naukratis: Greek Diversity in Egypt

The Greeks in Berezan and Naukratis: A Similar Story? number of additional complete or almost complete vessels would have been gained if all sherds had been kept – even though Berezan is already extremely rich in this regard. Most important, nevertheless, is still the quantity – it means that an amazingly high number of such more or less identical vessels, being representative for a certain time span, existed at Berezan at a certain time. Still, the remaining material, even if only a selective crosssection of fine-ware pottery from the 7th/6th centuries bc, represents one of the finest complexes of Archaic Ionian pottery ever found – this alone would justify the attempt to produce a comprehensive study. Furthermore, the sherds are in amazingly good condition as regards their surface and their sharp breaks – giving the impression of pots that, once broken, had not been moved much subsequently. Even more remarkable is the size of the sherds, which is very impressive compared to what can be found in the mother cities of Ionia. It must therefore be a declared aim of this study not only to present a general survey of the whole complex but also to restore as much context information as possible about the ‘where’, ‘how’ and ‘with what’ of each single fragment. Apart from these general considerations a couple of essential questions need to be addressed, as is the case regarding other colonies such as Naukratis. Where exactly does the imported Archaic Ionian pottery derive from? This question might be of major importance concerning the ethnic composition of the Greek settlers and the trading habits of these people, wherever they might have originated. Even though many important sites, especially in northern Ionia, and other hitherto unknown sites still await further investigation, our knowledge in this field has been considerably expanded due to recent and well-established archaeometric results:8 it has become feasible to determine places of manufacture more precisely. Understandably, scholars generally tend to look for intensive contacts between mother or founding city and colony – e.g. Miletos in the case of Berezan – yet we have to ask ourselves without prejudice whether this concept is always valid. Many pottery exports might well be more dependent on the rules of a

free market in which price, quality and aesthetics are more important than the actual ties to the founding city.9 Additionally, sources of influence might change and strong directional connections in the early days of a colony might not exist at a later time. We also have to consider whether traditional views such as the uncritical en gros assignation of countless colonies to the colonising spirit of the powerful Archaic city of Miletos alone stands firm in the light of archaeological evidence.10 At first sight some facts seem to be not only obvious but also alarmingly surprising: North Ionian products by far predominate the whole complex of Archaic Ionian pottery from Berezan (Fig. 2)11 – but this result must be critically considered, since proportions change substantially when one looks at the material divided according to its date. By focusing on 7th century bc pottery it turns out that Milesian or South Ionian pottery is still predominant (Fig. 3) – while things change dramatically in the first half of the 6th century bc in favour of North Ionian products (Fig. 4). The same pattern has recently been established by M. Kerschner for the western colonies, but it might also hold true for the northern colonies:12 the late 7th century bc imported pottery is generally dominated by South Ionian, mainly Milesian products, while the market in the early 6th century bc is dominated by North Ionian products. Superfluous to mention that we are still not able to assign precisely every single fragment to its point of origin – but we definitely should be able to discern tendencies by now. Yet a pottery kiln, discovered on the southern slope of the acropolis of Klazomenai and providing important evidence for the local production in Klazomenai should be taken into consideration: the kiln was in use during the 2nd quarter of the 6th century bc and the range of material found inside seems to contradict generalizations concerning the definition of fine-ware pottery produced in the Eastern Greek world.13 Nevertheless, these tendencies lead to another important conclusion: the undeniable parallel, contemporary existence of South Ionian and North Ionian pottery, which cannot usually be studied in the Ionian centres where one tends to find just one or the other type of pottery, but not both at once, is definitely of major importance. Since pottery trade seems to be amazingly rare between the various important sites in Ionia,14 we are

Figure 5 Detail of North Ionian amphora (inv. no. B408) showing North Ionian Wild Goat style in combination with black-figure style from the early 6th century bc found on Berezan

Naukratis: Greek Diversity in Egypt | 161

Posamentir

Figure 6 North Ionian bird bowls (inv. nos B69–60; B70–56; B89–28) from the late 7th century BC found on Berezan

Figure 7 South Ionian cup (Knickrandschale) (inv. no. B69–138) from the late 7th century BC found on Berezan

heavily dependent on colonies like Naukratis or Berezan in order to synchronize stylistic development and certain changes of substantial significance – such as the transition from the so called Wild Goat to the Fikellura style15 in the southern part of Ionia and respectively the transition from the so called Wild Goat to the black-figure style16 in the northern part of Ionia. Even though this most fascinating period of changes remains unclear in many aspects,17 it seems very unlikely that these two transition processes should be seen as independent and local phenomena. Unfortunately, it is easier in Berezan to follow the change between North Ionian Wild Goat and black-figure style (Fig. 5), while transitional pieces of the South Ionian products are rather scarce. Obviously, they are far more prominent in the area where the style was invented; in Miletos. This circumstance is definitely also due to the fact that around the time of change during the first half of the 6th century bc, North Ionian pottery already dominated the ceramic spectrum in Berezan. The typical North Ionian table amphorae or meander rim plates, being produced not only in one but most likely several places, had obviously already overrun the market. Yet the question whether the mutable partition between South Ionian and North Ionian products reflects a free market only or also indicates changes in the origin of probably newly arrived settlers18 remains open to debate – fine-ware pottery alone cannot provide a solution to a problem which in a similar way concerns the considerable amount of rough and locally produced coarse-ware,19 as well as other, non-pottery evidence.

How can the material be dated and what are the consequences for our understanding of the foundation, development and progress of a flourishing colony? The foundation of Berezan has been set, according to written sources,20 at a very early date and a small amount of apparently early material has been published in order to support this view.21 Recently this opinion has rightly been put to further discussion.22 Now, that important centres of Archaic pottery production such as Miletos23 for the South Ionian area or Klazomenai24 for the North Ionian area have been investigated more intensively, we are able to view earlier assessments on a more solid basis. At first glance it is remarkable that the surface of the earlier sherds is mostly not so well preserved as that of the following period even though the quality of production is usually higher in the 7th century bc; also the size of the sherds is generally noticeably smaller. Taking a closer look, it becomes perfectly clear that the amount of pottery found in Berezan dating from the 7th century bc is furthermore significantly smaller than that dating to the first half of the 6th century bc. This circumstance is not surprising since most colonies need at least two generations in order to reach a first period of higher accomplishments; the settlement might have been much smaller and the number of inhabitants might have been significantly lower. Nevertheless, the state of preservation and size of sherds seem to indicate that a lot more imported pottery was at hand at the site within the 6th century bc and broke after a shorter period of use.25 But an additional reason for this fact is revealed by focusing on the earliest pieces which should – according to the written sources – go back to the middle of the 7th century bc: the pottery does not attest Greek inhabitation before the last third of this century. There are practically no early types of the so-called bird bowls (Fig. 6; unlike the finds from Taganrok, for example)26 and there are also no early or even subgeometric

Figure 8 Fragments (inv. nos B254; B69–79; B83–15) of South Ionian Wild Goat style vessels from the end of the third quarter of the 7th century BC (SiA Ib/c) found on Berezan; the lower sherd probably of North Ionian origin and slightly later (?)

Figure 9 Fragments (inv. nos B172; B451) of South Ionian Wild Goat style pottery from the late 7th century BC (SiA Ic/d) found on Berezan

162 | Naukratis: Greek Diversity in Egypt

The Greeks in Berezan and Naukratis: A Similar Story? types of the so-called Ionian cups, or Knickrandschalen(Fig. 7):27 even though one exceptional piece has been considered by Kopeikina28 to be one of the oldest sherds found in Berezan. Early pieces of the Middle Wild Goat style in the Black Sea area, such as the beautiful pieces from Nemirov29 or the well-known jug from Temir Gora30 find only few and fragmentary parallels among the Berezan material (Fig. 8).31 What we do have from the 7th century bc consists mostly of so-called Milesian Middle Wild Goat II jugs and plates (or so called stemmed dishes, Fig. 9), accompanied by a number of North Ionian bird bowls dating not earlier than 630 bc. It has to be emphasised again that an early foundation date of around 650 bc is not supported by any firm evidence and is even contradicted by it; pottery does not indicate the presence of Ionian settlers before around 630 bc. In the following decades of the 6th century bc, pottery seems to reflect a time of comparative steadiness and importance as a trading emporium for the northern Black Sea coast, as we will see below.

Does the profile of imported East Greek pottery varieties at Berezan show the same features as in the cities where these kinds of pottery were produced? Even though full statistic analysis from major Ionian sites is still not available, such fundamental differences as exist can be recognized.32 A comprehensive overview of the material kept in the Hermitage of St. Petersburg shows a surprisingly limited variety of shapes (Fig. 10) and at the same time an even more surprising amount of almost identical objects. What is remarkable, for example, is the fairly high number of ‘extraordinary’ shapes such as askoi, alabastra, lydia etc. – this is without parallel in the mother city of Miletos. The fact that the majority of the vessels was obviously imported from northern Ionia (or even brought by North Ionian people?) offers a reasonable explanation, since most of these shapes are by far more common in the area of Klazomenai, for example. Still, one might wonder for what purpose the considerable number of askoi, for example, were dispatched to Berezan. Figure 10 Variety of shapes of Archaic Ionian pottery from the 7th and 6th centuries BC found on Berezan

Figure 11 Variety of Archaic drinking cups from the 7th and 6th centuries BC found on Berezan

Naukratis: Greek Diversity in Egypt | 163

Posamentir

Figure 12 North Ionian plates (inv. nos B101; B66–27; B67–68; B71–60) from the 6th century bc with floral decoration found on Berezan

More difficult to explain is the circumstance that there is not a single example of the typical Milesian one-handled drinking cup or mug among the thousands of sherds. Since these thinwalled mugs were more than common in 7th/6th century bc Miletos,33 this might seem rather surprising but should possibly not be overrated considering the fact that these vessels are mostly undecorated. Besides these unusual features the spectrum of shapes is dominated by table amphorae, jugs, kraters, plates or stemmed dishes, and drinking cups. Taking a closer look at specific shapes (such as drinking cups [Fig. 11]) one finds mostly well-known types of the North Ionian area, such as bird-, rosette-, meander-, lotus-, eye-, banded-ware- and animal-frieze bowls next to the so-called Ionian cups (Knickrandschalen) of South Ionian origin. Only a small portion (approximately 20%) of these vessels is less easy to assign but will not be discussed on this occasion. Similar classifications of the other shapes yield more or less the same results: most of the objects belong to well-known categories, underline the North Ionian predominance, and surprise only through their well preserved appearance. Much more interesting in this context is the fact that some of the objects exist in numerous and almost identical copies.34 Admittedly, North Ionian meander plates or meander rim kraters, for example, do not in general show a high level of creativity and can surely be classified as mass-produced ware, but nevertheless the unexpected and frequent repetition of certain products is somewhat remarkable for a rather small place like Berezan (Fig. 12). Yet the mere existence of several identical vessels holds true for other and more sophisticated wares as well, such as, for example, the Aiolian Dinoi of the so called London Dinos group. The number of sherds of these vessels found on Berezan makes the location one of the most important places of discovery for this group in the ancient world, next to Naukratis, which alone should be more than interesting.35 Taking a closer look at this material, one has to assume that most of these sherds can be assigned to a, naturally, smaller number of vessels; yet according to the rim pieces no less than five huge vessels of superior quality can be identified among the material kept in St. Petersburg (Fig. 13). Taking into account 164 | Naukratis: Greek Diversity in Egypt

Figure 13 Fragments of Aiolian dinoi (inv. nos B75–7; B83–8; B88–3; B89–20; B91–16) of the so-called ‘London Dinos group’ from the early 6th century BC found on Berezan

that this is most likely only a small percentage of all the vessels of this type once present in the settlement of Berezan, one is tempted to conclude that Berezan served as an important trading point for Ionian goods for a larger area during the 7th/ 6th centuries bc. If so, it would be interesting to find out whether certain preferences could be discerned concerning shape and, even more, iconography among the material stored at Berezan and waiting to be sent to other places. This task would certainly be difficult to accomplish with ornamentally decorated plates or simple drinking cups – but material such as Fikellura or Klazomenian black-figure style pottery could be investigated in this regard. Are there any indications for a local production of Archaic Ionian pottery on the island of Berezan? Even though already suggested by certain scholars,36 the idea of the production of vessels in, for example, Milesian or Klazomenian style on the island of Berezan itself is still almost entirely rejected by many archaeologists.37 If there were indications or even proof of such a production, the next question would concern the variety of copied material and, even more fascinating, whether the rules of a possibly existing local market were taken into consideration.38 Again, this complex of problems is strongly connected to one of the most interesting questions concerning Ionian pottery found in Naukratis. Pottery found on the island of Berezan consists, as already mentioned, not only of imported ceramics but also of handmade and locally produced coarse-ware – most likely used by the local population39 as well as by the Greek colonists. That ‘Greek’

Figure 14 Two stemmed dishes of banded ware (inv. nos B82–39; B84–32) from the 6th century BC found on Berezan

The Greeks in Berezan and Naukratis: A Similar Story?

Figure 15 South Ionian cup (Knickrandschale, inv. no. B66–76) of the first half of the 6th century BC found on Berezan

coarse-ware might have been produced by the colonists at the site should, in my opinion, be at least considered, but this question will not be addressed in the present investigation due to the more or less complete absence of such material among the objects kept in the State Hermitage in St. Petersburg. Also the possibility of a local production of painted pottery should be considered, at least theoretically.40 Even though it is not possible to discern peculiarities among the fine-ware pottery from Berezan at first glance, there are, in fact, some sherds that attract our attention in regard to this question. These fragments can be divided into three groups: first, copies of more or less well-known types of vessels originating from the southern or northern part of Ionia in style, shape and decoration. Only the poor quality of surface and slip, combined with minor irregularities, cast doubt on their Ionian origin. It is interesting to note in this context that Archaic lamps with central tube (Stocklampen) made of rough local clay, as well as unpublished imitations of Fikellura amphorae made of dark clay with black slip, are found on Berezan. The second group consists of vessels where – for whatever reason – shape or decoration do not fit into our known picture of Ionian vases. The third group consists of vessels which are apparently left unfinished – objects which would not normally have been exported one might think. In fact, a number of pieces belonging to these three groups – though faintly resembling Ionian products – must be classed as so peculiar that one feels forced to look for further explanations. A handful of examples should support this statement. For the first group we could exemplarily refer to two fragments of simple banded stemmed dishes (Fig 14) which find hundreds of parallels in Miletos or other Ionian cities: one (at the top) does not show any difference in shape, clay and surface to those found in Ionia, and recent clay analysis has in fact shown it to be Aiolian;41 yet the other (at the bottom) not only seems to be thicker and less carefully made, but most unusually bears a ropy and streaky coating, perfectly visible under soft raking light. Also the beige colour of this coating seems

Figure 16 Ionian cup (Knickrandschale, inv. no. B86–7) of the first half of the 6th century BC found on Berezan

Figure 17 Mixture between bowl and plate of banded ware (inv. no. B75–111) from the 6th century bc found on Berezan

Figure 18 Fragment (inv. no. B75–114) of a similar vessel to Fig. 17, but with spout on the inner rim from the 6th century bc found on Berezan

somehow unfamiliar, but one encounters this specific feature on several other vessels and fragments among the Berezan collection. Strikingly this kind of coating is used on other pieces in order to cover those parts of a vase which are normally left without such treatment: a so-called Ionian cup (Knickrandschale) of common type should be compared with a similar vessel, both found on Berezan (Figs 15, 16). While the first one bears no features that could attract our attention, the other again appears to be different, mainly because of the thickish beige layer below the rim. Furthermore a small detail is also different: while normally cups like these are decorated on their inside with thin bands of additionally applied red and white colour in the sequence white-red-white, this specific example most surprisingly shows a sequence red-beige-red. Even though these differences are minor, it should be clear that such peculiarities require an explanation – which is even more true for the following examples belonging to the second group. Among the numerous banded-ware plates or stemmed dishes of undoubtedly Ionian character (similar to those discussed above), one complete example stands out among the rest (Fig. 17): a peculiar mixture of plate and bowl, it has a shape which is, up to now, without parallels in the southern as well as the northern part of Ionia. Again, the thickish beige coating is remarkable, but even more puzzling is the fact that this vessel finds identical parallels only on Berezan itself – namely in at least four more pieces. Even though these other vessels are preserved only in fragments, small differences in execution – one has some kind of spout on the inner side of the rim, for example (Fig. 18) – make it perfectly obvious that the sherds in question belong to four different, if very similar, objects. This fact should be a clear warning not to assign strange fragments to one single vessel too easily, even when these fragments might be perfectly comparable to each other.

Naukratis: Greek Diversity in Egypt | 165

Posamentir

Figure 20 Profile drawing of South Ionian(?) krater Fig. 19

Figure 19 Fragments of South Ionian(?) krater (inv. nos B73–6.26.34; B78–14; B79–2) from the early(?) 6th century BC found on Berezan

Several sherds of a huge krater (Fig. 19), for example, bearing unfamiliar features and decorations besides the already well-known thickish beige layer, were found in different places on the island during eight different seasons and might belong to one or even to three or more objects – minor differences in the colour of the red-brown painted decoration do not offer any kind of solution to this question. At least the shape of the krater can be reconstructed (Fig. 20); it was a huge and coarse container with carelessly executed ornaments of doubtlessly Ionian character. These last two types of vessels certainly do not fit into the familiar spectrum of North or South Ionian pottery of the Archaic period – either they represent examples of local pottery production in the Black Sea area, or they derive from a hitherto unknown production centre which, for whatever reason, developed rather strange features. As I argue elsewhere, evidence from clay analysis (including the pieces illustrated in Figs 17–20) now suggests that the latter is in fact the more likely option, and that this production centre might be located in the Hellespont area.42 The third group is represented here by a huge body sherd of a small krater in black-figure or Corinthianising style (Fig. 21) and a small plate with floral decoration on the inside (Fig. 22). Both share a remarkable feature: they seem to have been left unfinished. The small krater shows a goat moving up towards

Figure 21 Fragment of North Ionian(?) krater (inv. no. B67–64) in black figure/corinthianising style from the first half of the 6th century BC found on Berezan

166 | Naukratis: Greek Diversity in Egypt

the left surrounded by some filling ornaments. Even though the animal’s body bears the usual application of added red colour, the otherwise obligatory incisions are, surprisingly, missing in the whole depiction. What makes the situation even more obscure are the again unusual (but in this case white) thickish coating of the image field and the generally awkward impression of the scene, which cannot be compared to any of the numerous North Ionian kraters of similar size and iconography. Normally, only deer but not goats carry dots on their belly in North or South Ionian depictions – on Aiolian dinoi of the London Dinos group, on the other hand, this detail is common. Of this latter group one also feels reminded as far as the rectangular application of the red colour on the back of the animals is concerned, but the difference in quality is still enormous.43 Again, a fragment like this could be an indication of local production – but also an origin in a less well-investigated geographic area seems conceivable. To the same group also belong, finally, two seemingly unfinished small plates (Fig. 22) – shape and surface are finished carefully here, but the customary central floral element on the inside is only laid down in its outline. Again, the fact that more than one object displays this phenomenon leaves the observer quite puzzled – but in this case not only the skillful execution, but also the existence of very similar pieces from other places, including Naukratis,44 seem to make local manufacture rather unlikely. In conclusion, it must be emphasized that certain indications for a local production of painted fine-ware pottery of the Ionian style definitely seem to be present among the ceramic material from Berezan kept in the State Hermitage of St. Petersburg. This does not necessarily mean that these vessels were produced on

Figure 22 Small North Ionian(?) plate (inv. no. B86–56) from the second half of the 6th century BC (?) with unfinished central floral element; found on Berezan

The Greeks in Berezan and Naukratis: A Similar Story? Berezan itself; it could also be the case that another production centre within the wider Black Sea area has not as yet been located. Again, this fact represents a strong parallel to the situation at the colony of Naukratis – but in this case a local production has already been proven.45 More archaeometric analysis must be undertaken on the material from Berezan in order to further verify the observations discussed above. Even if no proof of fine-ware pottery manufacture on Berezan or its surrounding region were to be revealed, one would still be forced to find explanations for some highly unusual ceramic products. Besides, as we have seen, the overall profile of the Archaic Ionian pottery found on Berezan also raises numerous questions that are of far-reaching importance for other colonies as well, such as Naukratis. It is, indeed, a similar story. Illustration credits

All photos S. Solovyov; Fig. 20 drawing R. Docsan; diagrams by the author.

Notes 1 2

Kerschner 2001, 72-7. I am greatly indebted to S.L. Solovyov, curator of the Berezan collection in the State Hermitage of St. Petersburg and former head of excavation in Berezan, for offering me the chance to publish large parts of the Archaic Ionian pottery from Berezan. Nevertheless this contribution is a preliminary report on my work, while the whole study will be printed within the framework of the Berezan publication project. For this reason most objects discussed here are presented in photography only – profile drawings and further information will be given on this later occasion; special thanks are due in this context to R. Docsan for producing thousands of profile drawings. For a detailed presentation of the first archaeometric analyses of the Berezan material, see also Posamentir and Solovyov 2006; Kerschner 2006; Mommsen, Kerschner and Posamentir 2006. 3 Cf. the comprehensive study of the site by Solovyov 1999, 19-27. 4 Russian archaeologists stopped working on the site in 1991 while a Ukrainian team under the leadership of V. Nazarov continued. Archaeologists from the State Hermitage St. Petersburg started with a special team again in 2003, now headed by D. Chistov. 5 Excluding Fikellura, Chian or Klazomenian pottery, which are the subject of a study by I.Y. Ilyina, material in St. Petersburg has been divided among various scholars in order to make possible the publication of a huge amount within a reasonable time; the first volume of this publication project containing, for example, the Chian pottery has already appeared in print; cf. Ilyina 2005, 70-173. 6 Solovyov 1999, 52 – but full drawing documentation of the pottery excavated in the 1980s is at hand. 7 A complete vessel of this ware has been found on Rhodes: Jacopi 1933, 54 fig. 52; similar objects (Walter-Karydi 1973, 19-20 fig. 23 pl. 36 no. 277) derive from Samos. Several fragments have been excavated in the Aiolian city of Larisa on Hermos: Boehlau and Schefold 1942, pl. 39. Some small jugs from Aiolis (see Ýren 2002, 179 fig. 11) display an odd mixture with Aiolian Wild Goat style and thus give a hint to look for a production centre within this geographical area. This supposition has now been confirmed by clay analysis of some of the Berezan pieces (samples Bere 105, 107, 108), which places them in the same group as the London Dinos group; see Kerschner this volume; Posamentir and Solovyov 2006. 8 Akurgal et al. 2002, 28-116, with full reference to other and earlier attempts towards the identification of production centres in this geographical area. 9 Kerschner 2000, 488-90. 10 For a critical review of each site once attributed to Miletos, see Ehrhardt 1983, 49-97; Tsetskhladze (1998, 36) goes one step further by considering the occasional participation of other Ionian centres. 11 Pottery decorated in Fikellura style is not included in this statistical analysis – but nor is pottery of Chian and Klazomenian origin; the two amounts almost neutralize each other. More interesting is the Aiolian share of the whole complex, which is surprisingly high, even

though only present in very specific products. 12 Kerschner 2000, 487; see also Tsetskhladze 1998, 51. 13 Ersoy 2000, 403-5. 14 This surprising but important fact has already been underlined by Ersoy 2000, 406. 15 Correctly described as a process of transition by Schlotzhauer (1999, 119-22; forthcoming b). Earlier contributions by Cook (1998, 63-6; 1999, 79-93) were focused on similar fragments but of Carian origin. 16 Özer 2004, 200-1; see also Ersoy 1993, 234-349; 2000, 403. The progressive parts of these transitional – or even bilingual – fragments or vessels show strongly corinthianising features and are still quite different from depictions of the developed Klazomenian black-figure pottery. Compare e.g. the vessels from Syracuse, Samos, Cyprus and Rhodes with several fragments from Naukratis and the huge amphora found on Berezan: Walter-Karydi 1973, pls 108, 109, 112, 115, 116, 119, nos 902, 907, 918, 941, 952, 975. 17 A new classification system of Archaic Ionian Pottery has just been presented by Kerschner and Schlotzhauer 2005, 1-56. 18 Tsetskhladze 1998, 51. 19 Solovyov 1999, 42-52. 20 Euseb., Chron. 95b Helm. 21 Kopeikina 1973, 241-3 figs 1-3. 22 Boardman 1998a, 201-2; Solovyov 1999, 29. 23 Schlotzhauer 2000; 2001; forthcoming b; Käufler 1999, 203-12; Ketterer 1999, 213-21; Posamentir 2002, 9-26; Villing 1999, 189-202. 24 Ersoy 1993, 291-419; 2000, 399-406; 2004, 51-66; Özer 2004, 199-219; Hürmüzlü 2004a, 82-7. 25 Compare Tsetskhladze 1998, 53. 26 Kopylov and Larinok 1994, 69. 27 Schlotzhauer 2000, 407-16; 2001. 28 See n. 22 – it is to my mind still hard to date these fragments accurately because of the lack of parallels – but certain details such as the applied red colour are certainly no indications for an early date of the cup. 29 Vakhtina 1996, 85-92; forthcoming. 30 Cook and 1998, 36 fig. 8.5. For a compilation of early Greek pottery in the Black Sea Area see Tsetskhladze 1998, 10-15. 31 The material at hand still proves the exceptional position of the settlement of Berezan; for outstanding but isolated fragments from Olbia see cf. Ilyina 2004, 76. 32 The author has been working extensively with Archaic Ionian pottery in Miletos, Ephesos and Samos. Frequent trips to many other important sites such as Klazomenai have enabled him to observe the differences as they are set out here. 33 The Milesian examples are not published yet, but the same type is also very common on Samos: see Eilmann 1933, 57-9. 34 Compare Kopeikina 1982, 10-30. 35 Extensively discussed by Kerschner, this volume. See also Kerschner 2001, 87-92; Ýren 2002, 165-207. For the clay analysis of the pieces from Berezan, see Posamentir and Solovyov 2006. 36 See Cook and Dupont (1998, 66-7, 90-1) for imitations found in Histria and Olbia. Cook already suggested that the imitations found in Histria might have been imports from the bigger place Olbia. 37 Shortly before his tragic death, V. Nazarov claimed to have found wasters of a pottery kiln in a rubbish pit. In fact, one can find a misfired table amphora of North Ionian style in the museum of Ochakiv, inv. no. Ab-021213. For a compilation of pottery kilns found in the Black Sea Area, see Tsetskhladze 1998, 42-3. 38 This question has already been most carefully raised by Tsetskhladze 1998, 13. 39 Solovyov 1999, 42-7, and comprehensively now Senatorov 2005, 174349. 40 Production of metalwork seems to find proof in several moulds found on the island; cf. Treister 1998, 182-8. Additionally it has to be said that the possible lack of suitable clay on Berezan directly is definitely no conclusive argument against local pottery production. 41 Posamentir and Solovyov 2006, sample no. Bere 138; Kerschner this volume. 42 Posamentir and Solovyov 2006. 43 Compare, for example, with Solovyov 1999, 61 fig. 45. 44 Naukratis: Cambridge, Museum of Archaeology and Anthropology NA 5; Tocra: Boardman and Hayes 1966, 50 no. 627 pl. 34. 45 Schlotzhauer and Weber 2005, 92-3; Mommsen et al. this volume; Schlotzhauer and Villing this volume.

Naukratis: Greek Diversity in Egypt | 167

Some Ceramic Inscriptions from Istrian Sanctuaries: The Naukratis Approach Iulian Bîrzescu Abstract Some inscriptions on pottery from the ‘Sacred Area’ in Istros are discussed in the wider context of similar discoveries from Ionian sanctuaries, especially in Naukratis. Although a great number of divinities are attested in this part of the city, the majority of the dedications do not form homogeneous groups. The Archaic inscriptions are written in the Ionian alphabet, albeit without a clear indication of the dedicators’ origins. In this context a ceramic inscription from the Archaic settlement is also considered which preserves the name of a certain Rhomis. The subject of this paper has presented itself as a result of the recent discoveries of a number of ceramic inscriptions in the Istrian sanctuaries (Figs 1-2). A catalogue of those unearthed until 1990 was published recently in a volume dedicated to the excavations of the ‘sacred area’.1 The small number of graffiti and dipinti discovered until now should be emphasized, about 35, among which 15 date from the Archaic period, less than 1% of the Naukratis total.2 In Istros it is, moreover, often difficult to relate the votive dedications to temples or other sacred monuments.

Figure 2 The ‘sacred area’ of Istros (2004)

The votive inscriptions presented here are arranged mainly by deity. Among the 12 epigraphically known gods from the sacred area, 7 of them are known from the graffiti. What distinguishes the ceramic inscriptions from Istros is first of all the lack of compact groups of dedications. Apart from the simple dedications to Zeus, there are various other inscriptions that can be understood only in the broader context of discoveries from the Istrian or other Ionian sanctuaries and places. Zeus and Apollo The first group relates to the family of Leto. Beginning with the Classical period there are five or six dedications to Zeus. Each of them bears the same simple Ionian inscription in the dative, Di/. Only one fragment on a West-Slope kantharos raises some questions concerning its restoration. All appear on black-glazed vases, four of them from the 5th century bc (Fig. 3), one on a Campanian black-glazed bowl from the second half of the 2nd century bc (Figs 4–5). The use of the same formula for such a long time is remarkable. Probably the Istrian graffiti are to be connected with the cult of Zeus Soter, widespread in Milesian colonies.5 Among the recent discoveries is also a well-preserved dedication to Apollo.6 The inscription,7 in the Ionic alphabet, was 3

4

Figure 3 Dedication to Zeus on an Attic black-glazed cup, third quarter of the 5th century BC

Figure 1 The ‘sacred area’ of Istros

Figures 4-5 Dedication to Zeus on a Campanian black-glazed bowl, second half of the 2nd century BC

Naukratis: Greek Diversity in Egypt | 169

Bîrzescu

Figure 6 Limestone base with dedication to Boreus, Hellenistic period

Figure 7 Limestone base with dedication to Orpheus (?), Hellenistic period

incised on the rim of an Attic eye-cup: )A[po]llwni/dhj m’ a)ne/qhken tw)po/llwni du/o e[?---] ‘Apollonides has dedicated me to Apollo, as two …’ Both the name of the dedicator and the name of the god represent the earliest epigraphic evidence of the Apollo cult in Istros. Furthermore, the inscription documents the worship of the god in the sacred area. The occurrence of the god without epiclesis is rather an exception for this time in the Pontic colonies, but almost the rule in the sanctuaries of Apollo in Didyma and Naukratis. Several hundreds of Archaic sherds from Apollo’s sanctuary of the Milesians in Naukratis bear votive inscriptions to Apollo. Among these there are some examples that preserve the same formula as our graffito. Despite the lack of an epiclesis, the dedication from Istros was perhaps made to Milesian Apollo, the god worshipped in Naukratis and Didyma. This is not the only inscription from the ‘sacred area’ related to Apollo, who probably had his main sanctuary here.8 In the 1970s a number of bases were uncovered, bearing inscriptions dated to the 3rd century bc. These were aligned on a via sacra oriented north-south.9 The inscription carved on the first base was restored by Pippidi as a dedication of the Boreis tribe, Borew/n (Fig. 6).10 But since there is no place on the stone for the letter ny, the inscription should rather be restored as Bore/w, an Ionic genitive dedication for Apollo Boreus. Such an epithet is attested five times in the Archaic and Classical period in Olbia.11 Furthermore, a 5th-century bc Olbian graffito provides evidence for the existence of an association of worshippers honouring Apollo Boreus, boreikoi thiasitai.12 As L. Dubois had already noticed, these were most probably related in some way to Orphic beliefs.13 The second limestone base, preserved only in fragments, was recently published in the above-mentioned volume concerning the excavations from the ‘sacred area’ (Fig. 7).14 On it the end of an inscription can be seen, – few (omega with iota subscriptum). At the beginning of this line, due to the limited space, not more than two or three letters could have been carved. In this case a suitable restoration would be 0Orfe/w. The slab that fits perfectly into this base is a marble stele dedication to Apollon Pholeuterios 170 | Naukratis: Greek Diversity in Egypt

Figure 8 Limestone base with dedication to Orpheus (?) and marble slab with dedication to Apollo Pholeuterios

(Fig. 8), according to Jurij Vinogradov a boundary stone.15 This latter stone was also discovered in the ‘sacred area’ and dates from the 3rd century bc. Another similar stone, this time with a dedication to Phorkys, preserved only in fragments, was discovered in 1949, on the south-eastern part of the acropolis (Fig. 9).16 I would suspect that this slab was aligned in the same way as the bases discussed above. The inscriptions on these bases for Orpheus and Boreus suggest that we can presume for Istros a situation similar to that in Olbia. Without wanting to insist here too much on the Orphic issue, I must, however, stress the relationship between Boreas and Orpheus, as it is portrayed by ancient authors. Homer in the Odyssey (13.109-112) describes the cave of the Nymphs, situated in Ithaka near the harbour of Phorkys. It is said that this cave had two entrances, a northern one, through which mortals would enter, and a southern one for the immortals. The cave myth, discussed at length by Porphyry, conveys the ancient belief in the transmigration of the souls, a common belief among the Orphics. As Aristotle (De anima 1.5) relates, the Orphics used to ascribe to the winds an important role in the metempsychosis process: ‘the

Figure 9 Marble slab with dedication to Phorkys, Hellenistic period

Some Ceramic Inscriptions from Istrian Sanctuaries: The Naukratis Approach

Figures 10–11 Fikellura eye-mug with dedication to Dionysos, second half of the 6th century BC

Figures 12–13 North Ionian cup with dedication to Hera, third quarter of the 6th century BC

souls enter from the Universe into their body when they are breathing, brought by the winds’. In this interpretation of the myth of the cave of the Nymphs, Boreas is the wind that brings life.

Dioskuroi in Naukratis, published for the first time by Ernest Gardner,21 and some dedications from Gravisca.22 This Istrian graffito is the first attestation of Hera’s worship in a Milesian colony of the Archaic period; although in Miletus itself, her cult had been known for some time.23 Hera’s cult could have been brought to Istros by the Samians, as in the case of Naukratis, where a great number of ceramic inscriptions has come to light in her sanctuary.24

Dionysos At the end of the archaeological campaign of 2004, fragments from the rim of a Fikellura trefoil mug17 (Fig. 10) came to light in the eastern part of the so called ‘great hollow’, in an Archaic layer directly on the base-rock. There is no difficulty in reading the five letters – onusw – on the inside surface of the rim. They belong to a dedication to Dionysos: [--- Di]onu/sw [---] (Fig. 11). The archaeological context, the ceramic typology and the letter-forms date the object into the second half of the 6th century bc. The four-stroked sigma and omega are indicative of the Ionic alphabet. The V-shaped ypsilon is common in the 6th century bc. Noteworthy is, of course, the deity to whom the dedication was offered. Until this find was made, the worship of Dionysos in Istros had been attested only from the Hellenistic period onwards, through theophoric names, illustration on coins and especially the reference to the Dionysia in a decree from the 2nd century bc.18 In the north-western colonies of the Black Sea, the cult of Dionysos is well-attested from earliest times onwards.19 At Olbia it is attested indirectly, by the Orphic tablets.20 Hera The vase, probably an East Greek cup, is only partially preserved, namely the lower part, yet the inscription written on the foot is complete (Figs 12–13). It is a votive inscription to Hera in the Ionic dialect, (/Hrhj. The archaeological context gives us as a terminus ante quem of 530 bc. The closest comparisons are a dedication discovered between the sanctuaries of Apollo and that of the

Aphrodite A roof-tile from the first half of 6th century bc, carrying a votive inscription to Aphrodite, has been meticulously published by K. Zimmermann.25 Its inscription, written in boustrophedon, has been discussed many times. The only difficulties in restoring the inscription are some letters from the end of the first and the beginning of the second line. In his study, Zimmermann repeats numerous reconstruction possibilities, using the few recognizable letters, EX...L?E... . One of the variants proposed by A.W. Johnston for the name of the dedicator was a rare Ionian name, Echeleon,26 which seems to me very likely. Zimmermann also admitted the possibility of a name with this patronymic.27 The reconstruction can now be supported by the reading of some letters discernible on the second line, which probably were the end of the patronymic name in the genitive case, – wnoj (Fig. 14). It is almost certain that the second line contains no ethnikon. )Afrodi/thi a)ne/qhken )Exe?l?e/[wn] | [---]wno?j? a)/pargma.

Figure 14 Dedication on a roof-tile to Aphrodite, first half of the 6th century BC (drawn after Pippidi 1983, 247, cat. 101)

Naukratis: Greek Diversity in Egypt | 171

Bîrzescu

Figures 15–16 Dedication to Hermes on an Attic band-cup, second half of the 6th century BC

Figure 18 Dedication of a Smyrnean on an East Greek krater, 2nd century BC

Hermes From the same archaeological context that contained the graffito for Apollo, and in the same year, 2003, a sherd was discovered with a dedication to Hermes (Figs 15–16): [---]hmeq[---|--]hj : (Erm[h=i ---]. Although the inscription is not well preserved, there are no difficulties in restoring the name of the deity. From the name of the worshipper only the two final letters survive, -hj, and from the name of the deity the first three. The separation between the two names has an accurate punctuation.28 The relationship between Hermes and Aphrodite is well-known from Olbia, where these two gods were worshipped together.29

Dedicators One of the important features of the Naukratian graffiti is the appearance of ethnika on some sherds. These are also to be found in the Istrian sanctuaries, but only in the Hellenistic period. In the 3rd century bc a Thasian who erected the temple of Theos Megas33 and a Smyrnaean (Fig. 18) appear in ceramic inscription discovered in the ‘great hollow’ four years ago.34 The few names that appear on the sherds from Istros are of Greek origin. Only four or five names date to the Archaic period, all masculine. As in Naukratis, one might tentatively suppose that the colonists brought with them women from different Mediterranean regions.35 The largest part of the visitors at this time came from Ionia. One of them engraved his name on a Milesian ‘Knickrandschale’:36 (Rw=mij. This cup (Figs 19–21) was discovered in the Archaic settlement, in area Z2.37 The name was presumed to have an Etruscan-Italic origin.38 But the reading of the sculptor’s name in an inscription on the monument of Theugenes the Potidean from Delphi made Alan Johnston presume that the name Rhomis was common in other regions as well in the Archaic period.39 Although different in quantity, the Archaic graffiti from Istros share several characteristic with many inscribed sherds of Naukratis, such as the related dedication formulae, the shape of the letters, the types of the vases – primarily of East Greek origin – and in part the same worshipped gods. Even if at this moment it is still too early to draw any conclusions regarding the origin of the dedicators in Istros, the lack of ethnika could indicate a single provenance, namely Miletos.

Phorkys The first graffito unearthed from the ‘sacred area’ is also one of the earliest Istrian inscription: [--- a)ne/qhk]e?n tw= Fo/r[kui --] (Fig. 17). Phorkys does not appear often in the literary sources, but his epithets seem favourable. The old god of the sea, as Homer called him (Odyssey 1.72), had a harbour named after him on Ithaka (Odyssey 13.96). In Greek art he is represented extremely rarely.30 Thought to be a pre-Greek deity, Phorkys was not treated as a ‘true’ god.31 Istros is the only place for which we can presume the existence of his cult. To the inscription from the ‘sacred area’ another two documents from the Hellenistic period can be added: a graffito on a roof-tile and the marble slab discussed above (Fig. 9).32

Figure 17 Dedication on a North Ionian krater (?) to Phorkys, first half of the 6th century BC

Figures 19–21 Graffito with the name of Rhomis on cup with everted rim

172 | Naukratis: Greek Diversity in Egypt

Some Ceramic Inscriptions from Istrian Sanctuaries: The Naukratis Approach List of deities from the ‘sacred area’ at Istros Deities Zeus

Epiclesis

Zeus

Polieus

Leto Artemis Apollo Apollo Apollo Dionysos Orpheus(?)

Phorkys Hera Aphrodite Aphrodite Aphrodite Hermes Theos Megas Moirai Kybele Apollo

Pythie Boreus Pholeuterios

Ietros

Object Six graffiti: four on stemless cup, delicate class, rim offset inside, and two on Hellenistic black glaze pottery Decree Dedication Dedication Dedication on an Attic eye-cup Limestone base Stele Fikellura trefoil mug Limestone base

Fig. 1 after Alexandrescu 2005; all other photographs and drawings are by the author.

Notes 1 2

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Alexandrescu 2005. The figure of over 1,500 ceramic inscriptions at Naukratis mentioned by Möller 2000a, 166, has been nearly doubled through the recent cataloguing of all the inscribed material in the British Museum by Alan Johnston; the great majority of these inscriptions date to the Archaic period. The only missing member of the family, Artemis Pythie, is now attested on a marble inscription, a dedication of an ex-priestess of the goddess honouring Leto: Avram et al. (forthcoming). Information from Dr. Vasilica Lungu. Ehrhardt 1988, 156. Another theophoric name, Ietrodoros, appears in Istros in the third quarter of the 6th century bc, Johnston 1996, 99-101. Zimmermann and Bîrzescu (forthcoming). Alexandrescu 2005, 62 and esp. 83-4. Zimmermann 1981, 463. Alexandrescu 2005, pl. 2.2. Pippidi 1983, cat. 97. Vinogradov and Rusiaeva, 2001, 134-40. Rusiaeva 1992, 18, fig. 4.1. Dubois 1996, 156. Alexandrescu 2005, 126. Vinogradov 2000, 139. Pippidi 1983, 250-1. For the shape, see Schlotzhauer 1999 234-6 figs 21-3. Pippidi 1983, 173, cat. 64. Herodotus 4.79, initiation of Skyles in Dionysos’ cult. For the cult in Olbia, Rusiaeva 1992, 96-9; Berezan, Tolstoýv, 1953, 55-6, graffito from

Date Four are dated to the 3rd quarter of the 5th c. BC, two from the Hellenistic period 3rd c. BC

Had an altar in the ‘sacred area’ Priestess

On the inside rim The ‘Apollo Pholeuterios’ stele probably fits into the Orpheus base

Graffito on an East Greek krater (?) Graffito on an East Greek high-foot cup Dedication on a ‘Sattelkalypter’ Dedication on a basalt perirrhanterion Dedication Dedication on an Attic cup Dedication on a temple architrave Dedication on a votive relief Statue Dedication on a marble architrave

Illustration credits

Observations Probably with the epiclesis Soter

Erected by a Thasian

Not discovered in the ‘sacred area’ but connected with the propylon (monument C)

4th c. BC 4th c. BC 3rd quarter 6th c. BC 3rd c. BC 3rd c. BC 2nd half 6th c. BC 3rd c. BC

1st half 6th c. BC 3rd quarter 6th c. BC 1st half 6th c. BC 2nd half 6th c. BC 3rd c. BC 2nd half 6th c. BC 3rd c. BC 3rd c. BC Hellenistic? End of the 5th c. BC

the 5th century bc, Dinu/so. 20 Rusiaeva 1978, 85-104. 21 Gardner 1886, 62, cat. 689; Bernand 1970, 673, cat. 318. Other dedications to Hera in Naukratis, Gardner 1888, 67, cat. 841-8. 22 Johnston and Pandolfini 2000, 17-9, cat. 4-46. 23 Kawerau and Rehm 1914, 162-3, cat. 31a. 24 The Samian provenance of Hera mugs from Naukratis is now confirmed by chemical analyses: Schlotzhauer 2006, 308-14, figs 124; Schlotzhauer and Villing, this volume. 25 SEG XXXIII 582; Zimmermann 2000, 239-51, with literature; Alexandrescu 2005, 69, 418 and 476. The beginning of the name of the dedicator is wrongly reconstructed in the last publication: the second letter is clearly a chi and not a xi. 26 SEG XXII 514; Fraser and Matthews 1987, 192. 27 Zimmermann 2000, 251. 28 The same punctuation in Olbia, Dubois 1996, 126, cat. 126, )Igdampaihj : (Ermh=i. 29 Rusiaeva 1992, 87, fig. 24.7. 30 LIMC VII.1 (1994), 398. Dipinto on a Corinthian vase in the Museum of Kavalla, Wachter 2001, 103-4, cat. 106a (Porkos). 31 See for example Heubeck and Hoeckstra 1989, 169-70. 32 Alexandrescu 2005, 418-20. 33 Alexandrescu 2001, 95. 34 Alexandrescu 2005, 422. 35 See also Oppermann 2004, 7. Until now only Metriche had been known in Istros, on a jug from the middle of the 6th century bc, Jeffery 1990, 479. 36 For the shape, see Schlotzhauer 2000, 410 (type 9). 37 Museum of Histria, Inv. no. V 26031. 38 Dubois 1989, 50, cat. 36 (Selinunt) and 120, cat. 114 (Camarina). The latter appears also at Arena 1992, 56, cat. 123; Fraser and Matthews 1997, 386. 39 Johnston 1980, 95-7.

Naukratis: Greek Diversity in Egypt | 173

Naukratis and Archaic Pottery Finds from Cyrene’s Extramural Sanctuary of Demeter Gerald Schaus Abstract It is well known that Cyrene was settled first by Greek islanders, with some later settlers from East Greece and the Mainland. Pottery and other finds from the extramural Sanctuary of Demeter and Persephone at Cyrene, help support historical sources about the settlers’ origins. Some evidence, however, especially from pottery, suggests links between Cyrene and Naukratis, whose Greek population largely came from East Greece.* The story in Herodotus (4.152) of Kolaios, the ship captain from Samos, who, in c. 639 bc, landed on Plataia, an island off the coast of Libya, is well known.1 His destination was Egypt, but his ship was blown off course, presumably having intended to sail directly to Egypt from the Aegean. Clearly the direct route south to Africa from the Aegean was not unusual.2 Korobios, the Cretan purple fisher or trader who guided the Theran colonists to the same island of Plataia, surely sailed this way after he learned about it from his own misadventure with contrary winds (Hdt. 4.151), and thereafter it must have been a common route for Greeks to sail between the Aegean and the coast of Cyrenaica. Though undoubtedly founded by colonists from the small island of Thera, literary evidence supports a mix of settlers at Cyrene by about 500 bc. 1) Therans were the original colonists (Hdt. 4.150–58), with a Cretan guide to Plataia. Later suggestions that Laconians had a hand somehow in early efforts, including mention of the Spartan three-time Olympic victor, Chionis, in the original expedition, cannot be trusted by themselves.3 Samian help at the earliest stage is remembered, but nothing confirms the presence of Samian settlers till later. 2) Herodotus (4.159) says that the population of Cyrene stayed the same during the reigns of the first two kings, but we know from archaeology that other settlements were founded along the coast during this time (Apollonia, Taucheira, ‘Ptolemais, Euhesperides), so there must have been an influx of Greeks.4 Did all come from Thera? This is not likely, given the size of the island.5 The Lindian Chronicle (17) mentions early settlers at Cyrene from Rhodes, for example,6 and it is possible that Laconians (Isokrates, Philip. 5) took an early interest.7 Common sense and a little archaeological material also suggest at least some settlers from Crete and the Cyclades.8 3) After a major second wave of settlers was encouraged by Battos II, c. 580 bc, causing serious conflicts with the native Libyans (Hdt. 4.159), we learn of a mid-6th century bc constitutional change by an arbitrator from Mantinea (Demonax) which divided the citizens into three tribes: a) Therans and Perioikoi (the latter are problematic. They clearly have some status to be joined with the Therans in this tribe. One suggestion is that they were native Libyans [Hdt. 4.159 – ïß ðåñßïéêïé Ëßâõåò] who revolted and won the battle

of Leucon, killing 7,000 [!] Cyreneans, says Herodotus [4.160]. Another possibility is that they were from districts on the island of Thera outside Thera town itself);9 b) Cretans and Peloponnesians (the latter again are problematic, but for lack of any literary evidence or clear archaeological evidence for Peloponnesians other than Lacedaemonians, surely the south Peloponnese was best represented, and rather than Spartiates sailing to Libya, it may be that members of the Perioikoi of Laconia are as likely candidates as any);10 c) Islanders (Rhodians and Samians have the best claim, but others from the Cyclades or Cythera seem likely). Quite separate from Demonax’s division of Cyrene’s citizens are the Samian-led mercenaries who won back the throne for Arkesilas III, probably in the third quarter of the 6th century bc, in return for a promise of land (Hdt. 4.162–4). These men may have come from Samos and elsewhere, including a wide region of Asia Minor opposite the island. Before examining East Greek wares excavated in the Demeter Sanctuary at Cyrene, which provide the best evidence for influence from Greek commercial activity in the direction of Naukratis, it is worth considering other fabrics found during the University of Pennsylvania excavations in the 1970s.11 It should be emphasized that almost all the Archaic material is scrappy. Besides 300 published fragments of Athenian12 and 400 of Corinthian wares,13 there were about 550 pieces of other imported fine wares published from the Sanctuary, divided into nine fabric categories. Although Cyrene was settled by c. 631 bc, the Demeter Sanctuary contains very little material from the 7th century bc. For example, of 5,000 pieces of Corinthian pottery, Kocybala was only able to identify six which could be ascribed with any confidence to Early Corinthian.14 Three Island gems, an amulet seal, two ivory objects, seven engraved tridacna shells and some terracotta figurines have also been dated to the 7th century bc.15 No Athenian, East Greek, Laconian or other fabric need be earlier than c. 600 bc. So this has become the accepted starting date for the Sanctuary. Of first interest then is the little bit of Theran, Cycladic and Cretan pottery which is so rarely found outside the Aegean.16 It is particularly clear in demonstrating an on-going link between the colonists and their south Aegean homeland for several generations after the initial settlement. What is notably absent is any of the so-called ‘Melian’, ‘Siphnian’ and ‘Parian’ pottery which occurred in the contemporary Demeter Sanctuary at Tocra, ancient Taucheira, 125km west of Cyrene.17 Over 50 examples of these wares appeared at Tocra. The difference, I suggested elsewhere, might be due to the arrival at Taucheira of a group of colonists from other islands in the Cyclades than went to Cyrene.18 These vases are also not commonly traded, except to Thasos, a colony of Paros.19 No pottery of these types has been Naukratis: Greek Diversity in Egypt | 175

Schaus reported from Naukratis. Laconian pottery has more of a story to tell. In the Cyrene Demeter Sanctuary, 223 vases of this fabric were published from a wide range of shapes, and dating from the early to the late 6th century bc.20 No Laconian vase painter is represented in significant numbers, although the workshops of the Hunt Painter, the Naukratis Painter, and the Rider Painter are best represented with 16, 13 and 8 pieces respectively.21 Regarding the range of shapes, there is a remarkable variety, greater than almost anywhere else, especially for the black-painted vases. The following table lists the most common types, with a comparable count of the same shapes among Tocra’s 110 published Laconian vases. Counts of Laconian vases by shape Shapes Cyrene Cups 94 Kraters 40 Lakainai 23 Dishes 12 Aryballoi 11 Chalices 4 Hydriai 4 Oinochoai 3 Flat-based bowls 3 Jugs and juglets 2

(% of 223) 42% 18 10 5.5 5 2 2 1.5 1.5 1

Tocra 20 5 5 3 6 0 0 7 27 11

(% of 110) 18% 4.5 4.5 3 5.5 0 0 6.3 25.022 10

Many rare or unique shapes also occur in the two assemblages, including a beaker, kothon, skyphos, stamnos, lekythos, pyxis and a double vase. The difference in the numbers of each shape between the two sites is quite striking and hard to account for, especially because both are Demeter sanctuaries. Only the workshops of the Naukratis Painter (10 vases) and the Rider Painter (4 vases) are at all well represented at Tocra.23 Despite the differences, the percentage of Laconian compared to the total Archaic pottery from both sites is similar, and they both have a full range of shapes and decorative types.

It is interesting to compare these finds with Naukratis, which Marjorie Venit has done in publishing the Laconian from Egypt.24 About 50 pieces are known, all are from Naukratis where provenience is recorded. Of these, there are only two shapes represented in numbers, kraters and black-figure cups. Of the 34 black-figure cups, about 27 can be attributed to painters, and of these, a remarkable 18 are by the Boreads Painter, and where datable, belong to the brief period c. 570–565 bc. The possibility that they arrived in a single shipment has been considered.25 The only other place where the Boreads Painter is represented in such impressive numbers is at Samos, where so much Laconian has been found. Venit suggested that there is no reason to believe that the Laconian pottery found at Naukratis came by way of Cyrene since the assemblages are so different, but instead came via Samos or at least in Samian ships.26 Maria Pipili has disconnected the Naukratis Painter’s name vase, found at Naukratis, from any link with Cyrene, identifying the goddess on this vase as Orthia (Artemis) holding two stylized boughs, not a silphium branch, the remarkable medicinal plant of Cyrenaica.27 Two fragments on Samos probably have the same subject. Also the only known vase with a dipinto by the Naukratis Painter was found at Cyrene, in a script which could be Cyrenean.28 It has properly been pointed out to me, however, that it may just as likely, if not more likely, be in the Corinthian script.29 If so, this would reduce the Egypt– Cyrene–Laconia connection to one vase, the name vase of the Arkesilas Painter (Fig. 1) who has clearly imitated an Egyptian theme of Osiris overseeing the weighing of the heart (soul) of the dead (Figs 2–3), but has substituted a named figure, generally accepted as King Arkesilas II of Cyrene, who here supervises the weighing of a bulky white material, likely wool or silphium, both of which were important exports of the city.30 Unusual features in the Laconian vase painting include the pointed hat worn by Arkesilas II, and the sceptre in his hand. These seem to imitate the figure of Osiris, with tall pointed Figure 1 Paris, Cab. des Méd. 189. Arkesilas Cup tondo

176 | Naukratis: Greek Diversity in Egypt

Naukratis and Archaic Pottery Finds from Cyrene’s Extramural Sanctuary of Demeter

Figure 2 The ‘weighing of the heart’ from the Book of the Dead of Hunefer, c. 1275 bc (British Museum EA 9901/3)

Figure 3 Detail from the ‘weighing of the heart’ with a monkey sitting on the scales from the Book of the Dead of Any, c. 1275 bc (British Museum EA 10470/3)

crown and sceptre on the Egyptian paintings. The workers around the scale pans as well as the writing over their heads also find similarities in the Osiris scenes, as do the birds, especially the flying stork, which recalls the composite, falcon-like winged eye of Horus in the Egyptian scene.31 The canopy over the group with Osiris is paralleled by the sail over the head of Arkesilas. The lizard crawling beside Arkesilas may be a symbol of foreboding or omen of ill-fortune, as Hurwit recently suggested.32 The most striking feature which has convinced scholars of the imitation of an Egyptian Osiris scene is the figure of the monkey on top of the balance beam, a common feature in Egyptian scenes of Osiris weighing the heart of the dead (Fig. 3). The monkey represents Astes, the associate of Thoth, and watches over the correct and just procedure of weighing.

The artist could have seen an Egyptian painting of the god Osiris as his model, without travelling to Egypt himself, but one must ask what made him depict an historical figure like the king of Cyrene, especially since historical figures of any kind are so rare as subjects in vase painting?33 Is this a joke, or political cartoon at the expense of a king nicknamed, ü ÷áëåðüò (the Harsh) (Plut. Mor. 260E) – in contrast to his father, Battos the Fortunate? Osiris has a very serious task in his weighing duties, but why should the king of Cyrene be depicted supervising the weighing and storage of a mass of material, presumably of some importance? One possibility is that the vase painter is here alluding to the close relationship between Arkesilas II and the Egyptian pharaoh, Amasis.34 According to Herodotus (2.161; Naukratis: Greek Diversity in Egypt | 177

Schaus 4.159) a Cyrenean army badly defeated an Egyptian one sent by Pharaoh Apries at Irasa in Cyrenaica (570 bc) just a few years before Arkesilas became king. This Egyptian army was not apparently supported by Apries’ Greek mercenaries and it was unused to fighting Greeks. The defeat led to the rebellion of the Egyptians against Apries whose Carian and Ionian mercenaries acquitted themselves well but were not enough to save Apries.35 Thereafter, Herodotus (2.181) says that Amasis concluded a treaty of friendship and alliance with Cyrene by which the two countries became close friends and allies, and he even married a woman named Ladike, who was either the daughter of the Cyrenean king (Battos or Arkesilas, accounts differ, notes Herodotus 2.181) or of a leading citizen of Cyrene, named Kritoboulos. Herodotus admits he does not know whether Amasis did it as a goodwill gesture or he just wanted to marry a Greek woman, but we now know that Herodotus, in fact, conflated events, that there were as many as three battles to settle the issue between Amasis and Apries, and that in the third, in 567 bc, Amasis was supported by soldiers from Cyrene in defeating Apries.36 It seems likely that Amasis’ marriage to Ladike of Cyrene occurred at this time for political purposes, cementing the alliance that Amasis made with the Greeks beyond his western border. It is just before he mentions the pact and the marriage that Herodotus discusses the favour with which Amasis treated the Greeks in Egypt, granting them various privileges, and in particular, establishing Naukratis as their commercial center, along with land for temples and altars. It is now recognized though that in the climate of anti-Greek feeling leading to the overthrow of Apries, and with a need to tighten the administration of foreigners in Egypt, Amasis intended to restrict all Greek commercial activities to a single location and to force Greeks in the country to be governed by him through their representatives, the ðñïóôÜôáé of Naukratis.37 The result of these restrictive measures, however, was that Naukratis enjoyed a period of considerable prosperity remembered by Herodotus a hundred years later, coinciding with the prosperity that Amasis’ long rule brought the country as a whole (Hdt. 2.177–9). With an alliance and friendship established between Cyrene and Amasis’ Egypt by 567 bc, and the port of Naukratis designated as the sole place for Greeks to do business in Egypt, one would expect that a close relationship between Cyrene and Naukratis was established. Herodotus (2.182) goes on to list the benefits that Cyrene received from Ladike and Amasis, including statues, one of Athena and the other perhaps of Aphrodite, and a painting of Amasis himself. It may be presumptuous to suggest that these gifts left Egypt on board ships that had docked at Naukratis, but surely there were many other items that made their way to Cyrenaica from this port. As a further sign of the alliance with Amasis, Arkesilas II was apparently supported during his reign by Egyptian soldiers, as Stibbe argued based on Plutarch (Mor. 261C).38 The reign was short (c. 566–560 bc), during which Arkesilas fought with his brothers, and was assassinated either by Learchos, one of his brothers, (Hdt. 4.160) or a friend named Laarchos (Plut. Mor. 260E). Arkesilas’ wife, Eryxo, says Herodotus, took revenge by killing Learchos. We return then to the vase by the Arkesilas Painter which seems also to reflect the close ties between Egypt and Cyrene at this time. It is certainly puzzling that he may have included 178 | Naukratis: Greek Diversity in Egypt

Eryxo’s name (here OPYXO retrograde), yet left her out of the scene,39 and this is all the more strange if the vase painter is copying from a panel painting as some have suggested. Although speculative, the artist might, for example, be making a pointed comment about an unpopular king who was being propped up on his throne by troops from Egypt. In any case, the painter was apparently aware of events in a distant Greek polis in Africa, and was interested enough to comment on them through his well-labelled vase. In the end though, the vase found its last use as grave furniture in an Etruscan tomb. Before discussing the East Greek fabrics from Cyrene’s extramural Demeter Sanctuary, one should note that certain Egyptian and Egyptianizing objects have been found both at Cyrene and Tocra. These include faience and alabaster vases, carnelian beads, seals, bronze figurines, as well as ostrich eggs and decorated tridacna shells.40 Some of these objects, however, might have been produced by Phoenicians, not Egyptians, and so are not necessarily confirmation of the ties with Egypt. On the other hand, one direct link between Cyrene and Naukratis has been argued by Steven Lowenstam in publishing the seals from Cyrene’s Sanctuary of Demeter. 18 of the 44 glyptic objects (or 41%) are Egyptianizing, and he believes that they all came from Naukratis itself where there was a workshop making scarabs of closely similar types. The alabaster alabastra from Cyrene and Tocra and a faience head scarab from Tocra have also been linked to Naukratis.41 In the other direction, the only certain objects from Cyrene found in Egypt are silver coins found in fair numbers in hoards.42 We arrive at the East Greek fabrics from the Demeter Sanctuary, and it is these which can be explained best by referring to Naukratis. The great majority of the Late Wild Goat pottery at Cyrene, including floral and banded wares, must have come from the North Ionian region.43 Add to this the large group of Chian vessels, and the black-figure vases related to Klazomenian and one can see that the representation of pottery from the area is substantial. Yet none of Cyrene’s known settlers came from North Ionia. At Naukratis, on the other hand, Herodotus (2.178) says that the largest, most famous and most frequently attended sanctuary, the Hellenion, was built by Greeks from nine states, of which four are North Ionian: Chios, Teos, Phokaia (once Aiolian) and Klazomenai. There was a large amount of Late Wild Goat and floral pottery at Naukratis, and these same types are reflected at Cyrene.44 Of particular note at Cyrene are the large hemispherical bowls with Wild Goat animal friezes, and in general the use of black-figure for the Wild Goat decoration.45 There is also great variety in the floral dishes, greater even than Tocra. Many of them are likely to come from North Ionia. Chian pottery in Cyrenaica is surpassed in quantity only on Chios itself, at Erythrai opposite the island, Berezan and Olbia in the Black Sea, Aigina, and, of course, Naukratis, to my knowledge.46 More than half of Lemos’s catalogue of decorated Chian pottery (887 out of 1659 pieces) comes from excavations at Naukratis where the fullest range of Chian styles and vase shapes was found. The assemblage at Cyrene is different from Tocra’s, particularly in its greater range of shapes. Most of the styles of decoration from the first half of the 6th century bc are represented at Cyrene, including the Animal Chalice, black-figure, Lion-and-Sphinx, Patterned Chalice, and perhaps even an example of the Grand Style.

Naukratis and Archaic Pottery Finds from Cyrene’s Extramural Sanctuary of Demeter Counts of Chian vases by shape Shape Chalices Lids Dishes Phialai Bowls 2-handled pots Fruitstands Large closed vases Plate

Cyrene 25 11 8 7 3 3 2 4 1

Tocra 56 1 0 1 1 0 0 0 0

Not only were the Chians among the founders of the Hellenion at Naukratis, but the Sanctuary of Aphrodite was also much used by them. It is hard not to recognize the influence of the Chian population at Naukratis on the finds of pottery from Cyrene’s extramural Sanctuary of Demeter particularly in the quantity and range of shapes and decoration of the Chian at Cyrene. It is especially noteworthy since the Chians were among the least likely of the Islanders to have come to Cyrene as colonists. It seems evident that Chian goods were being brought to North Africa in the first place because of the interest of the island in Egypt, but that ships bearing these goods also found their way to Cyrenaica.47 Some Chian transport amphorae have been found both at Cyrene and Tocra. Their numbers in both Demeter Sanctuaries of these cities are not unusual, but they are more common close to Cyrene’s agora.48 It is possible that Chians traded for grain, silphium and wool among other things in the region, leaving behind wine, and perhaps silver, given their contacts with the North Aegean mining area. Two fabrics remain, Fikellura and Ionian ‘bucchero’. John Hayes argued for a South Ionian origin for the latter, in the Ephesos – Meander valley area.49 A fair number of these ‘bucchero’ vessels were found on Samos. Many more have occurred in Magna Graecia and Etruria, and an Etruscan series is well known. The finds at Cyrene tell us very little at the moment, but the discovery of 23 examples is worth noting. As for Fikellura pottery, a product of Miletos for the most part, Strabo (801) says that it was a Milesian fleet of 30 ships that defeated the local Egyptians and founded Naukratis in the reign of Psammetichos I, while Herodotus (2.178) tells us that they established their own sanctuary of Apollo there. Much Fikellura has been found both at Naukratis and Tell Defenneh to suggest a continuing Milesian presence through the 6th century bc.50 The 19 pieces of Fikellura from Cyrene (just one was uncovered at Tocra), mostly amphorae and some cups, may well be an echo effect of the pottery being brought to Naukratis.51 There is no reason to think that Milesians came among the colonists in the second wave of settlement, especially when there were so many opportunities for them in their Black Sea colonies; however, there is also no reason to insist that Fikellura was brought in Milesian merchant ships rather than others after a stop in Egypt, though this is less efficient. Uhlenbrock has argued that the occasional Milesian perfume flask found in the Demeter Sanctuary at Cyrene arrived by a complicated route, having been sold and resold as the shipment was disbursed along the way.52 Nothing has been said about Athenian or Corinthian pottery, perhaps rightly in this context, but Corinthian has a remarkably long life at Cyrene, well down into the 5th century bc after other places had given up on this fine ware.53 Roebuck commented on the strength of the link between Corinth and Egypt, with

Corinthian silver coins being found in Egyptian hoards, while many Egyptian objects were found at Perachora.54 One wonders if this link has been overlooked too often, especially since both stone architecture and examples of polychrome wall painting occur as early in the Corinthia as anywhere, both owing some debt to Egypt. It is usually just assumed that Corinthian and Athenian wares were carried regularly in Aiginetan ships, perhaps to Cyrenaica as well as to Naukratis. To sum up, the origins of Cyrene’s settlers certainly help explain several special features of the Archaic imported fine wares in the city, but there are enough other features which cannot be explained this way to look for a second significant influence, and this second one seems to be the exchange of goods between Cyrenaica and the mainly East Greek settlement at Naukratis. Illustration credits

Fig. 1 photo © Bibliothèque Nationale de France; Figs 2, 3 the British Museum.

Notes *

I wish to thank Alexandra Villing for kindly inviting me to participate in the Naukratis Colloquium, and Udo Schlotzhauer for stimulating discussions and his efforts in assisting with my travel. I especially wish to thank Ivan D’Angelo for so generously sharing with me the results of his study of pottery from the Italian excavations at Cyrene. 1 Discussed again recently by Möller (2000a, 54-5 with refs). 2 Fulford (1989, 169-72) discusses navigational conditions, noting that prevailing winds made sailing in a north-south direction easier than in an east-west direction. Strabo (10.4.5 [475]) says that the trip between the western tip of Crete and Cyrene was two days and nights in length. He notes that the trip from the eastern tip of Crete to Egypt was four days and nights, ‘though some say three.’ 3 For Chionis, see Paus. 3.14.3, but note Jeffery’s (1961, 143 n. 10) suggestion of a later Chionis. 4 Boardman 1966, and 1994, 143. For Archaic pottery as early as the second quarter of the 6th century bc from Euhesperides, see Vickers and Gill 1986. Earlier material has now been uncovered, suggesting Greek settlement at Euhesperides at least by c. 600 bc, if not before. See Gill 2004; cf. also Zimi 2003, 212. The recent proposal by James (2005), suggesting the foundations of Taucheira, ‘Ptolemais’, Apollonia and Euhesperides occurred with the influx of new settlers under Battos II c. 580 bc, is dependent on a lowering of the chronology of Archaic pottery by about 35 years. This is less appealing than the assumption that Herodotus (4.159) only means Cyrene, not Cyrenaica in general, when he says that the population at Cyrene stayed the same as the original settlement during the reigns of Battos I and Arkesilas I, and that expansion did occur quite early along the coast. Even so, it is hard to take Herodotus at face value when he says that the population of Cyrene did not increase at all for its first 56 years. He (4.153, 156) indicates that it only had about 100 original settlers (‘two fifty-oared ships’), all men apparently, as the Foundation Degree seems to confirm. Pottery finds from the Casa del Propileo in the agora area of Cyrene, discussed by D’Angelo in this volume, tend to support the traditional dating of Archaic pottery. 5 Malkin 2003c, 160-3. 6 Shaya 2005 for a summary of the Chronicle and discussion of the context of the inscription, dated 99 bc. 7 For Rhodians, see Schaus 1985a, 102-3; and for Laconians, Schaus 1985a, 98-102; Schaus 1985b; and comments in Malkin 2003c, 160-1, 165. 8 For pottery from these areas at Cyrene, see below. For 7th century bc terracotta figurines likely of Cretan origin found in Cyrene’s extramural Demeter Sanctuary, see Uhlenbrock 1985, 297-9. An early house plan at Euhesperides with offset door is compared to houses on Crete, see Gill 2004, 399-402 fig. 3. 9 Jeffery 1961, 142-4. 10 On the Laconian Perioikoi, see Shipley 1992. 11 Archaic pottery from the extensive Italian excavations in the agora of Cyrene still awaits detailed publication. For references, see Naukratis: Greek Diversity in Egypt | 179

Schaus

12 13 14 15

16 17 18 19 20 21 22

23 24 25 26 27 28 29 30

31 32 33

D’Angelo’s paper (in this volume) nn. 9-10. D’Angelo is studying early material from one area of the agora. For a recent paper discussing pottery as an indicator by which to compare Greek activity in Cyprus and in Egypt, see Sørensen 2001. Moore 1987. Kocybala 1999. Kocybala 1999, 5. For the gems and seal, see Lowenstam 1987, 10-4, nos 20, 25-6, 28. The seal (no. 28) is an heirloom, dated to the first half of the 7th century bc. For the ivory objects, Warden 1990, 10-1 no. 26 (ivory ram, perhaps Laconian), and 24 no. 106 (2-faced head -– dated to the 7th century in Expedition 34, 1-2 [1992] 54 fig. 8), tridacna shells, Warden 1990, 61-2 nos 467-73. For the terracotta figurines, Uhlenbrock 1985, 297-9; 1992, 18. Schaus 1985a, 5-6 nos 1-11 (Theran), 7-9 nos 12-16 (Cycladic), 10-14 nos 17-43 (Cretan). Boardman and Hayes 1966, 73-8 nos 885-920; Boardman and Hayes 1973, 34-6 nos 2083-100. Schaus 1980, 24; 1985a, 107. Boardman and Hayes 1966, 73 with references. Schaus 1985a, 15-48 nos 44-266. Schaus 1985a, 121 Appendix II. The high number of black painted, flat-based bowls at Tocra compared to Cyrene can be explained to a certain degree by the poor preservation of the pottery at Cyrene, making it difficult to identify Laconian examples of this vase type. Based on Stibbe’s lists, Stibbe 1972, 269-74 (Naukratis P.), 285-7 (Rider P.). Venit 1985, 393-4. Ibid. Ibid., 394 n.28 for references to others who have likewise proposed Samos. Williams 1993a, 595 suggested that Aigina replaced Samos as transporter of Laconian pottery after c. 550 bc. She follows Droop, Lane and Shefton in this identification, see Pipili 1987, 41-2 with refs. Schaus 1979. Faustoferri 1985, 341 and pers. comm. Paris, Cabinet des Médailles 189 from Vulci. Stibbe 1972, 30, 115-7, 279-80 no. 194, pls 61-2 fig. 28. On the Arkesilas cup, Stucchi (1987) attempts to identify the white material as blocks of rock salt, but fails to explain how such a heavy friable material could be stacked in rounded ‘balls’ (skein-like) so high in the open weighing pans, or held easily in the hand of the far right worker. He omits from his discussion the dipinto, ÓËÉÖÏÌÁ×ÏÓ, for this worker (not his ÏÑÕÎÏ[Ó), most suggestive in identifying the material as silphium. An explanation for the lack of final sigmas in ÏÑÕÎÏ and ÌÁÅÍ (he suggests that ‘ìÜíåò’ was meant) in otherwise very careful labels is also not offered. For a recent opinion in favour of silphium being weighed on the vase, see Luni 2002, 359-62. I am grateful to A.J. Spencer and N. Spencer for information on the composite ‘bird’ of the Hunefer papyrus. Hurwit 2006, 129. Prof. U. Hoeckman kindly pointed out to me that Greeks may have been explicitly forbidden from entering Egyptian tombs, or at least learning of Egyptian burial customs, where such scenes with Osiris were commonly found. Prof. U. Verhoeven informs me that ‘the Book of the Dead from Herakleopolis Magna dated c. 600 bc (P. Colon. Aeg. 10207; cf. Verhoeven 1993, 304) mentions that no Greek (‘haunebut’) should know the spell (BD spell 148)’ and it should be performed within a cloth tent with yellow stars as decoration. ‘Du sollst ausführen im Innern eines Zeltes aus Stoff, der ganz mit Sternen von gelber Farbe besetzt ist. 117,13 Es ist ein wahres Geheimnis, nicht sollen die Nordvölker an irgendeinem kennen.’ On the other hand, instances of the scene of Osiris weighing hearts (souls) can be found on objects other than papyri, tombs, coffins, shrouds, and mummy wrappings, see Seeber 1976, 27-9, including an ostracon found in the tomb of Ramesses VI (Cairo CG 25057) as a sketch for a wall painting; a stele from the pyramid complex of Pepi II with a short version of the weighing scene; a pectoral (London, Univ. Coll. 7726) also from the Ramesside period; and two ushebti boxes of a man from the 21st dynasty with the balance on one side, adoration in front of Osiris on the other side (Louvre N 4124). Prof. Verhoeven adds, ‘I think there must have been model books for this kind of scene; the balance of Osiris is so widespread and of common knowledge that a Greek could anyway

180 | Naukratis: Greek Diversity in Egypt

34 35 36

37 38 39 40

41 42

43 44

45

46

47 48 49

50

51 52 53 54

get notice of it.’ I am indebted to Profs Hoeckman and Verhoeven for this information. For a discussion of the reign of Arkesilas II and the involvement of Egyptians, see Stibbe 1972, 195-201. Supposedly numbering 30,000 (Hdt. 2.163), but doubtfully so (Lloyd 1988, 41). On Carian mercenaries in Egypt see also Williams and Villing, this volume. Boardman 1994, 141 n. 10 (with references). Key to this is Edel 1978, 14-6 where a cuneiform tablet in the British Museum (ANE 33041) is understood to mention military help received by Amasis from ‘Putujaman’, identified as Cyrene, against a force sent by Nebuchadnezzar in his 37th regnal year (567 bc) to assist Apries against Amasis. Boardman 1994, 141; James 1981, 734-6. Stibbe 1972, 198-9. See Schaus 1983, 89. Besides the Egyptianizing faience objects believed to be made on Rhodes (Cyrene: Warden 1990, 11-3 nos 29-35 [human and animal figurines], 53-4 nos 382-6, 388-9 [vessels]; Tocra: Boardman and Hayes 1966, 165 nos 86 [horse figurine], 87-91 [aryballoi]), other Egyptian and Egyptianizing objects include from Cyrene: Warden 1990, 4, 7-9 nos 13, 16, pls 4-5 (bronze falcon and frog); 24-5 nos 113-18 pl. 18 (carnelian poppy-head pendants); 28 nos 154-60, pl. 21 (heart and face pendants of bronze and terracotta); 53-4 no. 387, pl. 39 (faience lentoid jar); 55-6 nos 403-13 fig. 8 pl. 41 (alabaster alabastra), 58 nos 436-8, 444-6 fig. 11, pls 43, 45; 60 no. 464 (ostrich egg cup?); 60-2 nos 465-73, pl. 49 (tridacna shells); Tocra: Boardman and Hayes 1966, 165 no. 92 (faience head scarab), no. 93 (faience disc beads), 166 nos 98-100 (alabaster and carnelian objects). Boardman and Hayes 1966, 166 no. 98 (alabastron); and 165 no. 92 (head scarab); see also Warden 1990, 56. At least 40 alabaster vases are in evidence at Cyrene. Roebuck 1950, 239 and n. 5. Thompson, Mørkholm, and Kraay 1973, hoard nos 1636-7, 1639, 1641-2, 1644-5, 1647, 1652 (a total of 50 silver coins from Cyrene, of which 35 occurred in the Asyut hoard, no. 1644). I owe this reference to Robert Weir. Schaus 1985a, 49-72 passim. For North Ionian Late Wild Goat pottery, see Cook 1998, 51-6. The material has been scattered widely among European and other museums. For a fair idea of its appearance, see Fairbanks 1928, pls 34-5. A small selection of LWG pieces is listed in Möller 2000a, 243-4 Appendix 1.d; for ones in Egyptian museums, Venit 1988, 6 (list of catalogued vases with incised Wild Goat ornament). A selection of the many Late Wild Goat and other pieces in the British Museum from Naukratis were made available for viewing to the Colloquium participants. Hemispherical bowls, Schaus 1985a, nos 299-308; and Late Wild Goat black-figure, nos 270 (amphora), 273 (oinochoe), 304-5 (hemispherical bowls), 327-30 (fruitstands), 350 (dish), 420-2 (plates). Cyrene, Demeter Sanctuary: Schaus 1985a, 77-85 nos 469-537; Tocra: Boardman and Hayes 1966, 57-63 nos 771-817; Boardman and Hayes 1973, 24-8 nos 2042-55. For the decorated styles of Chian pottery, Lemos 1991, especially Chapt. 6 ‘Distribution’ pp. 191-208; they are re-assessed in the article by Williams in this volume. For a few other remarks on its distribution, see Schaus 1996. For the strength of the Chian presence in Naukratis, see Roebuck 1950; and Chian trade, Möller 2000a, 79-81, 135. Schaus 1985a, 105 n.133. See D’Angelo Fig. 14, this volume, for Chian transport amphorae from the Casa del Propileo site near the agora. Boardman and Hayes 1966, 65-6 nos 828-37; Boardman and Hayes 1973, 28 nos 2057-8. For examples from Cyrene, mostly of narrow alabastra and ‘perfume pots’ with horizontal grooves, Schaus 1985a, 73-6 nos 446-68. Möller 2000a, 88 suggests a Milesian trading connection for Egyptian grain by the end of the 7th century bc, to help the city survive annual invasions by the Lydians. Pottery in the Apollo Sanctuary at Naukratis goes back to the earliest years of the settlement. For Fikellura pottery from Naukratis and Tell Defenneh, Cook 1933/4, passim; 1954, 1-13, pls 568-81. Schaus 1985a, 86-9 nos 538-56. Uhlenbrock 1992, 19. Kocybala 1999, 6. Roebuck 1950, 238.

Imported Greek Pottery in Archaic Cyrene: The Excavations in the Casa del Propileo Ivan D’Angelo Abstract Since the volumes on the agora by Stucchi and Bacchielli, very little has been published about imported Greek Archaic pottery from the Italian excavations in Cyrene. The finds in the Archaic levels of a domestic complex that were brought to light in the area of the late Hellenistic Casa del Propileo, and that are presented here for the first time, are therefore of particular importance and provide new elements for discussion. Together with the results from the extramural sanctuary of Demeter and Kore, the data from the Casa del Propileo seem to confirm that the overwhelming percentage of the imported pottery in Cyrene during the Archaic period came from and was produced in the East Greek region. This fact leads us also to speculate upon a possible trade link between Naukratis and Cyrene. Another relevant aspect is the remarkable number of Laconian kraters from the excavations, which suggests that elite symposia took place.* The aim of this paper is to present some preliminary results of the research into the pottery of Cyrene – the famous city which bears witness to the expansion of Greek colonies along the coast of North Africa. The still ongoing excavations in the Casa del Propileo1 started in the years 1966–69, and were reopened in 1999–2002, each time under the direction of I. Baldassarre2 of the University of Naples ‘L’Orientale’. The Casa del Propileo is located in the area measuring about 200m in width which remains between the acropolis and the agora (Fig. 1). The area is intersected by the Skyrotà, the street which Pindar sings about in the fifth Pythian Ode;3 starting at the acropolis, the street runs down towards the terrace of the sanctuary of Apollo. The Casa del Propileo, immediately to the west of the agora (Fig. 2), is a large uniform building complex, measuring 25 x 50m, which can be dated to the middle of the 1st century bc (Fig. 3). The entrance from the Skyrotà to the south is preceded by the tetrastyle propylon (Fig. 4) of which today only the foundations remain and which has given the whole complex its conventional name, Casa del Propileo (‘House of the Propylon’). Investigation has focused on this monument with the aim of understanding its functions, and above all because of the interest aroused by the Archaic structures brought to light underneath it (Fig. 5). All the trenches have actually revealed the same stratigraphy: an Early Hellenistic layer that seals the preceding remains from the Classical and Archaic period. In all the trenches, the structures from the Hellenistic period show remains of floors in opus signinum that preserve the homogeneity of the layers underneath. The excavations carried out in the first peristyle of the house have led to the identification of Archaic structures of a residential district probably laid with a different orientation with respect to the later urban grid. The entire complex is being studied by the Mission of the

University of Naples ‘L’Orientale’. Preliminary conclusions indicate that the Archaic pottery covers a stretch of time from the last quarter of the 7th to the end of the 6th century bc. The majority of these materials have been recovered in homogeneous strata, the rest from mixed contexts. The specific aim of this study is to examine Archaic materials from a domestic context rather than a sacred context, which is a substantially new approach compared to earlier publications on Cyrene4 and the Cyrenaica.5 Analysis of the Archaic material has led to the identification of local pottery and the clays used to make these vessels6 as well as to a better knowledge of Archaic pottery imports to Cyrene and the Cyrenaica, adding to the picture delineated in important publications7 by the American Archaeological Mission of the University of Pennsylvania, and the still fundamental volumes on Tocra8 (it should be emphasised that both are sacred sites, dedicated to Demeter and Kore). Furthermore we have partial data on materials recovered by the Italian Archaeological Mission at the terrace of the sanctuary of Apollo9 and in the agora,10 and initial reports from Euhesperides11 by the English Mission. As already mentioned, this is only a preliminary report and one needs to be aware of the very fragmentary state of the material. The pieces are often very small and complete vessels are rare. Looking at the classes of imported pottery found at the Casa del Propileo, 3% of the pottery can be traced to Thera and the Cycladic area in general. At first sight, this fact seems to add value to the observation that Theran pottery is very rarely found outside the island itself.12 In any case, important pieces like a Theran amphora (Fig. 6)13 and a Cycladic skyphos with the typical decoration of concentric circles were recovered. All these materials can be dated to between the end of the 7th century bc and the first half of the 6th century bc. As usual, Corinthian pottery is well represented at 19.2% of the total; most finds can be attributed to Middle Corinthian (Fig. 7). But there are also examples of Early Corinthian; in percentage terms there are more of them at the Casa del Propileo than at the extramural sanctuary of Demeter and Kore.14 This class provides chronological pegs for an early dating of the Archaic structures in the area of the Casa del Propileo, assigning them to the first phase of the colony.15 Although 8.1% of the pottery is Attic, this presents us with a particular problem: we have hardly any decorated fragments. This can be explained by the fact that at the Casa del Propileo, we are dealing with an Archaic domestic site. Decorated vessels seem to have ended up almost exclusively in necropoleis16 and sacred areas.17 Laconian pottery makes up a little less than 5%. The presence of this pottery in Cyrene has aroused the interest of scholars who have tried to single out possible Cyrenean subjects,18 mainly on cups, or to prove that these vessels were Naukratis: Greek Diversity in Egypt | 181

D’Angelo produced in Cyrene itself. Lane, in an article19 written more than 70 years ago, has already disproved this latter hypothesis, fittingly describing it as the ‘Cyrenaican heresy’. Sparta is certainly one of the most important factors in the reconstruction of Cyrene’s mythical and historical past.20 On one side, from the exegesis of literary sources like Herodotus and Pindar, Cyrene seems indebted to and thus strongly linked to the Lacedaemonian polis, for example institutionally. Also the legendary history of Chionis, mentioned21 as second mythical founder, is to be taken into consideration, but with appropriate caution. On the other hand, however, the vases would appear to be evidence of connections between aristoi rather than of regular and direct ‘trade’. Significantly, in fact, some 19 examples of Laconian kraters have been found at the Casa del Propileo (Fig. 8). The earliest krater (Fig. 9), from the beginning of the 6th century bc, belongs to the so-called group ‘with double-stepped rims’ in Stibbe’s classification,22 and finds a particularly close match, among others, in an example uncovered in the extramural sanctuary of Demeter and Kore.23 On the whole, however, with upwards of 40% of the total, East Greek pottery prevails. This fact accords with sources which reflect instances of close ties between the polis of the Battiades and the East Greek area.24 The reform of Demonax25 allows us to place the presence of settlers from this area around the middle of the 6th century bc. The conspicuous presence of East Greek pottery at the Casa del Propileo, some of which prior to this event, cannot be used sic et sempliciter to date the appearance of settlers or merchants; it does nevertheless testify to the existence of strong commercial links in which Naukratis could have played an important role.26 At present, in the absence of scientific analyses of the clays, it is not possible to determine exactly in which production centres the finds from the Casa del Propileo were made, something which is now possible for the East Greek area thanks to the results of analyses carried out by, among others, R. Jones27 and P. Dupont28 and most recently by M. Kerschner and H. Mommsen.29 There are fragments pertaining to nine bird-bowls, which can be dated to between the end of the 7th and the first quarter of the 6th century bc. One of them (Fig. 10)30 can be attributed to Coldstream’s31 Group III (c. 650–615 bc), while the rest are of Group IV (after 615 bc). The type with rosettes is less common. There are four lotus-bowls (Fig. 11), the diffusion of which is for the most part limited to the East Greek area.32 Significantly, examples are known from Naukratis.33 As in many other sites in the Mediterranean, a truly overwhelming percentage of ‘Ionian cups’ (cups with everted rim/Knickranschalen), a very common class34 of pottery, is found in the Casa del Propileo (Fig. 12). As is well known, it was widely exported from the second half of the 7th century bc through the 6th century bc. The main types are represented. Adopting the canonical typological classification, as proposed by Villard and Vallet,35 many examples of type A2 (with and without bands on the rim) and B2 have been found, but also some examples of A1 and B1. It is immediately possible to recognize imports from the East Greek area and also a series of local imitations, identifiable as such from clay and decoration.36 There are also numerous fragments of East Greek dishes (Fig. 13), some of which can be traced back to Cook’s North Ionian Late Wild Goat style type (NiA I).37 182 | Naukratis: Greek Diversity in Egypt

Finally, we should also point to the conspicuous presence of amphorae of different origin: for the East Greek area a remarkable number of Chian (Fig. 14) and some Samian amphorae have been found, but also examples of the Attic type, ‘SOS’ and ‘à la brosse’, as well as Corinthian and Laconian amphorae. Many of these came to light inside what would have been a storeroom of an Archaic house. All the Archaic classes of pottery from the Casa del Propileo presented here have also been found elsewhere in Cyrene and Cyrenaica (Tocra), at sites that appear to give an analogous picture.38 But we can claim without doubt that the excavations at the Casa del Propileo, together with the analysis of the materials and structures uncovered, have provided significant new information about Archaic Cyrene on issues of its history, topography and relations with other areas of the Greek world. At this point, it does not seem superfluous to underline once more the preliminary stage of the conclusions. The following brief considerations should therefore be understood as working hypotheses: Whether Cyrene is to be considered an anomalous colony within the scope of Greek colonial movements39 because of the supposed non-homogenous composition of its colonists and because of the nature of its relationships with local populations is open to question. More convincing is the particularity of its institutional, monarchical and hereditary regime.40 Right from its foundation in c. 631–630 bc, Cyrene became part of a network of established commercial naval routes which were well structured and well used. The hypothesis of A. Johnston, who suggests that goods from the East Greek area were passing through the ports of east and south Crete, is very credible.41 Evidence from Building Q at Kommos,42 dating from the last 30 years of the 7th century bc, fits well with the episode narrated by Herodotus43 in which Theran settlers were helped by Korobios, a Cretan from Itanos. Crete’s function as a passageway can also be imagined for the route that brought goods from the ports of Laconia to Cyrenaica, according to recent theories put forward by M. Gras.44 Furthermore, in line with opinions already expressed, by, among others, Roebuck45 and Schaus,46 it is possible to recognise the existence in the Archaic period of a second, coastal route, which connected Cyrene and its subcolonies to more eastern sites like Naukratis. Evidence recently published by D. Bailey47 on Marsa Matruh, where East Greek pottery is most in evidence among Archaic imports, backs this up. It is certainly possible to find points of contact between Naukratis and Cyrene,48 even as early as the Archaic period, but the exact nature of the relation between these two important Greek centres on the North African coast needs to be clarified. They were founded for different reasons and functions: while Naukratis was born initially as an emporion or port of trade,49 Cyrene is clearly a colony of essentially agricultural character.50 This emerges from several facts, such as the progressive acquisition of new land, the resulting deterioration of relations with local tribes, the foundation of subcolonies both in the territory and on the coast, the exploitation and trade of a natural resource – the famous sylphion plant.51 It is hoped that as the excavation and study of the structures and materials associated with them in the Casa del Propileo continue, more light can be shed on Archaic Cyrene and its relations with Naukratis and the Greek world in general.

Imported Greek Pottery in Archaic Cyrene: The Excavations in the Casa del Propileo Illustration credits

Fig. 1 after Stucchi 1967; Fig.2 after Baldassarre 2002; Fig.3 E. Mitchell, elaborated by H.J. Beste and C. Zieschang, Deutsches Archäologisches Institut Rom; Figs 4–14 the author.

21 22 23 24

Notes *

1 2 3

4 5

6 7 8 9 10 11 12 13 14 15

16

17 18 19 20

First of all I wish to thank Prof. I. Baldassarre (Naples), who oversaw my research with care and patience. Thanks are due to Prof. N. Bonacasa (Palermo) and all the members of the Italian Archaeological Mission at Cyrene, the Libyan Authorities in Cyrene (Shahat), the Supervisor of the Antiquities Mr. Abdulgader Mzeini, and to the colleagues and friends with whom I was able to discuss various aspects of this research: H.J. Beste, M. D’Acunto, I. D’Ambrosio, K. Iara, M. Kerschner, M. Mirold, R. Posamentir, A. Salerno, A. Santucci, G.P. Schaus, C. Zieschang. Finally, my gratitude to U. Schlotzhauer and A. Villing, who kindly invited me to the Colloquium, and to E. Manning, K.S. Powell and E.M. Steinby, for helping me with the translation of the text into English. The denomination of the complex is, as is mentioned below, conventional; see for instance Stucchi 1967, 95; 1975, 144, 308. For the excavations in the years 1966–69 see Baldassarre 1987, 17-24. A short summary of the results of the most recent excavations is given in Baldassarre 2002, 18-20. The building of the Skyrotà street is attributed to the mythical king and founder of Cyrene, Battos I (vv. 90-3: ‘he laid out a stretch of ground, level, cut straight to be a road of hoof-clattered cobblestones: Apollo’s martial parades pass there, by the edge of the market place,’ transl. F.J. Nisetich); on the scholia to these verses of Pindar, see Stucchi 1967, 21 with references. Pernier 1931, 1935; Stucchi 1967, 1984; Bacchielli 1981; Schaus 1985a; Kocybala 1999. For Tocra, see Boardman and Hayes 1966, 1973. For Apollonia, see Goodchild et al. 1976. For Euhesperides, see Vickers and Gill 1986. James (2005) has recently proposed a lower chronology of Archaic pottery which may not be impossible, but for the pottery from the Casa del Propileo we prefer to follow the traditional chronology; cf. also the doubts expressed by Schaus on James’ proposal: Schaus, this volume, n.4. The Archaic local pottery from the Casa del Propileo and the fabrics that have been identified will be published by D’Angelo (forthcoming). Such as, among others, the volumes of the series most quoted in the present paper: Schaus 1985a; Kocybala 1999. Boardman and Hayes 1966, 1973. Pernier 1931, 1935. Stucchi 1965; Bacchielli 1981; Stucchi and Bacchielli 1983; Santucci 1998; (forthcoming). Zimi 2001, 2002, 2003 and Gill 2004 for imported pottery; for coarse wares and local production, mostly post-Archaic, see most recently Swift 2003, 2005. As explained well in Schaus 1985a, 5. For the decorative details see Dugas 1928, nos 8 and 14, pl. 8A (belonging to the so-called ‘style insulaire orientalisant’); another very close match is illustrated in Dragendorff 1903, 17, fig. 12. Kocybala 1999, 5. The story which Herodotus (4.145-67) tells about the foundation of Cyrene seems to be confirmed by the chronology of the pottery recovered, thanks to survey activities, at the site identified as Aziris; see, for instance, Boardman 1966, 150-52. See, for instance, Beschi 1969/70; Thorn 2005, who has commendably reorganized the finds from the Cyrenaic tombs recovered by A. Rowe. I wish to thank here J.C. Thorn for giving me a copy of his work before publication. On Attic pottery from the necropoleis of Cyrene see also Maffre 1996, 1998; and finally Elrashedy 2002. Attic pottery from the extramural sanctuary of Demeter and Kore at Cyrene is published by Moore 1987 (black-figure and black-glazed) and McPhee 1997 (red-figure). An attempt by Faustoferri 1985, 337-48. See also Schaus 1985b, 395403, for the evidence for Laconians in Cyrenaica in the Archaic period. Lane 1933/4, 182-5. See, for instance, the in-depth analysis of: Nafissi 1980/1, 1985;

25 26

27 28 29

30 31 32

33 34

35 36 37 38 39 40 41 42 43

44 45 46 47 48 49 50 51

Calame 1996. Also very interesting is Malkin 2003c. Paus. 3.14.3. Stibbe 1989, 29-30 (no. C12). Schaus 1985a, 27 (nos 100 and 102). See the Lindian Chronicle (FGrHist 532, 17): this source is, however, a late document (early 1st century bc) and links the Therans with a group of Rhodians, who were led by a certain Pankis. See, for instance, Applebaum 1979, 12; Uhlenbrock 1992, 18. Schaus (1985a, 102-5) made a thorough and cautious analysis of the possible connections between the East Greek pottery from the extramural sanctuary of Demeter and Kore at Cyrene and the origin of Cyrene’s colonists. On this subject see also his article in this volume. Hdt. 4.161.7. See, for instance, Corcella 1993, with previous bibliography. Möller 2000a, 2001. And see also the important observations by Uhlenbrock 1985, 297-308; 1992, 16-23; (forthcoming), on the East Greek influence in the local production of Archaic and Classical votive terracottas found at the extramural sanctuary of Demeter and Kore at Cyrene. Jones 1986. Dupont 1983, 19-43. On East Greek pottery see Kerschner 1997b, 1999. On the results of chemical analysis, see Kerschner et al. 2002, 189-206; Akurgal et al. 2002; Mommsen et al., this volume; Mommsen and Kerschner, this volume. I wish to thank M. Kerschner for indicating typical details to me, present in the example from the Casa del Propileo, which date back to the third quarter of the 6th century bc. Coldstream 1968, 300. Examples of lotus bowls are also known from the Black Sea area, which is of course closely linked to the East Greek world. For this type in general, Cook and Dupont 1998, 26-8; for items from Histria see Lambrino 1938, 59-61, figs 28-30. See also here in this volume the very interesting article of R. Posamentir, who sheds new light on the East Greek pottery from this area. Price 1924, 187, fig. 9. See, among others, the very useful work by Boldrini 1994, with complete previous bibliography. A convincing suggestion to define these forms is made by Schlotzhauer 2000 (Knickrandschalen), cf. also Schlotzhauer 2001a, 122; 2001b. Villard and Vallet 1955, 7-34. The local examples of ‘Ionian Cups’ recovered at the Casa del Propileo will be published in D’Angelo (forthcoming). Cook and Dupont 1998, 51-6. For percentage terms and the use of statistics, see Stucchi 1984, 1648, and particularly the perspicacious observations of Boardman 1994, 144-7. Mitchell 2000 deals with this subject. On this aspect exhaustively De Vido 1998. Johnston 1993, 376-7. For which see Johnston 1993. Hdt. 4.151.8. Also Strabo, 10.4.5 (C475), reports details, though he refers to a later period, about the trip between Crete and Cyrene. For the navigational conditions between Crete and Cyrenaica see Fulford 1989, 169-72; Purcaro 1976, 285-95, indicates also other routes with many references. Gras 1997, 52-5; 2000a, 158. Roebuck 1950, 242, 247. Schaus 1980, 22. But see also Venit 1985, 393. Bailey 2002, 118. See among others Schaus 1985a, 104, and this volume, and Boardman 1999a, 123. As has been well demonstrated by Möller 2001. Chamoux 1953, 115-27, 230-4; Applebaum 1979, 74-82; Baldassarre 1999, 385; Boardman 1999a, 153-9. On sylphion (lat. laserpicium) see Chamoux 1953, 246-63 and Gras 1985, 165-72 with previous bibliography and information from the literary sources; see also most recently Roselli 2001, 11-20. On the botanical aspects see Manunta 2002; on the iconography see Luni 2002. The sylphion may appear also on the famous cup of Arkesilas (Schaus Fig. 1; Paris, Cabinet des Médailles 189, from Vulci): on this problem see Stibbe 1972, 115-17, 279, no. 194, pls 61-2, fig. 28; Stucchi 1987, 29-34; and Schaus, this volume.

Naukratis: Greek Diversity in Egypt | 183

D’Angelo

Figure 1 Cyrene: the acropolis, the terrace of the sanctuary of Apollo and the agora

Figure 2 Cyrene: the Casa del Propileo immediately to the west of the agora

Figure 3 Plan of the Casa del Propileo

184 | Naukratis: Greek Diversity in the East

Naukratis and Archaic Pottery Finds from Cyrene’s Extramural Sanctuary of Demeter

Figure 4 The Propylon on the Skyrotà street

Figure 5 Trench CP01B

Figure 7 Middle Corinthian closed vessel

Figure 6 ‘Theran’ amphora

Figure 8 Laconian krater

Naukratis: Greek Diversity in Egypt | 185

D’Angelo

Figure 9 Laconian krater of double-stepped rim type

Figure 10 East Greek pottery: bird bowl

Figure 11 East Greek pottery: Lotus bowl

Figure 12 East Greek pottery: Ionian cup

Figure 13 East Greek pottery: dish

Figure 14 Chian amphora

186 | Naukratis: Greek Diversity in the East

Etruscan and Italic Finds in North Africa, 7th–2nd Century BC Alessandro Naso Abstract North Africa has revealed some Etruscan and Italic finds dating to the 7th–6th century BC, and these are particularly concentrated in Carthage. The quantity and the nature of the pottery and bronzes found here show that the Punic city had direct and intense trade relationships with Etruscan partners such as Caere, that were both barbaroi and natural allies against the Greeks, as the literary tradition confirms. Etruscan artefacts have also been found in Greek colonies such as Cyrene and Naukratis, probably brought along the complex trade routes connecting the western and eastern Mediterranean in the Archaic period. The good relationships between Carthaginians and Etruscans continued for many centuries, since in Tunisia and Algeria there are isolated finds until the 2nd–1st century BC, sometimes inscriptions, revealing the presence of Etruscan people, who probably escaped from their homeland conquered by the Roman armies. Two finds lists are provided as appendixes to the paper, the first relating to Etruscan and Italic artefacts in North Africa, the second to a type of bronze tool, part of an Etruscan drinking wine set, that has been found all over the Mediterranean.* Introduction Etruscan finds in North Africa is a rather neglected field of research, and an overview of the relationship of Etruria and the Italian peninsula with North Africa that includes an exhaustive finds list is still lacking. The evidence for Etrusco-Punic relations was collected in the 1960s by M. Pallottino,1 in the 1970s by J. MacIntosh Turfa,2 in the 1980s by J. P. Morel and J. P. Thullier3 and in the 1990s by Fr. W. von Hase;4 however all these works are limited to finds from Carthage.5 This paper will extend the research to the whole of North Africa and compare the large amount of data from the Punic city with finds from elsewhere, including Greek colonies such as Cyrene and Naukratis. Two main phases can be distinguished, the first corresponding to the late Orientalizing to Archaic periods, and the second to the late Archaic to the Hellenistic periods. In accordance with the main theme of this conference, I will concentrate more on the earlier period, but as so many important finds belong to the later phase, this cannot be ignored completely.

The Iron Age One object that needs to be mentioned here is an antennae sword that the Marquess of Courtance bought in Egypt in the early 70s of the 19th century and gave as a gift to the king of Italy for the Royal Armoury Museum of Turin, where it is still preserved (Fig. 1). There is no record of the original provenance of the sword. The type has a wide distribution in Italy as well as in Central Europe in the 9th–8th centuries bc. R.C. de Marinis recently discussed the various typologies developed by scholars for these swords. The sword in Turin belongs to the oldest type, the so-called Tarquinia-Vetulonia type, dating to the 9th century bc.6 This chronology makes it highly improbable that the provenance of this Italic sword, a Prunkwaffe, could be North Africa, but of course the question is still open to debate. Late Orientalizing and Archaic periods From the second half of the 7th century bc the founding of Greek colonies, such as Cyrene7 (with its subcolonies at Taucheira, modern-day Tocra,8 and Apollonia9), Euhesperides10 in western North Africa and Naukratis in central-eastern North Africa, brought not only Greek colonists, but also new connections and new waves of trade.11 At that time, Carthage had not yet begun its expansion into the western zone of North Africa, but had already established trade relationships with the Etruscans, as the many finds indicate, and probably also with the Italic populations in Sicily, as reported by ancient authors.12 In its tombs, from both old and new excavations, more than 60 bucchero vases have been found, dating at least from the third quarter of the 7th century onwards (Fig. 2). Amongst these early finds, an oinochoe in thin bucchero (or bucchero sottile) is particularly notable; it is of a form quite typical for Caere and its district and dates to just after 650 bc (Fig. 2.7). Twenty-eight little amphorae, 12 oinochoai and jugs, found in several graves and probably connected to wine consumption, are further indications for contacts with southern Etruria in the second half of 7th century bc. In Carthage there are also bucchero drinking cups: 2 kotylai, 11 kylikes and above all 11 kantharoi. The bucchero kantharos, as is widely known, is the real marker of the Etruscans all over the Mediterranean from Spain to Turkey and from the South of France to North Africa,

Figure 1 Antennae sword

Naukratis: Eastern Greeks in Egypt | 187

Naso

Figure 2 Bucchero pottery from Carthage

188 | Naukratis: Greek Diversity in Egypt

Etruscan and Italic Finds in North Africa, 7th–2nd century BC

Figure 3 Etruscan bucchero kantharos from Naukratis

particularly the shape classified as 3e by T. Rasmussen, used from the end of the 7th to the first half of the 6th century bc. It is possible to add other finds to the distribution map compiled by Fr. W. von Hase, from Megalopolis and Paros in Greece, and Daskyleion, Miletos, Didyma and Datça on the Knidos peninsula in Turkey.13 In this map we can include Tocra and Naukratis, too: at Naukratis at least two kantharoi sherds were found, belonging to the type Rasmussen 3e. The first, once in the von Bissing collection, has been published by E. Prins de Jong (Fig. 3).14 The second, still unpublished, has been seen in the Museum of Fine Arts in Boston by U. Schlotzhauer, who kindly told me of its existence (Fig. 4).15 The first sherd is relevant, because, according to Prins de Jong, some traces of silvering are still preserved on it: this is a very fine coating that may have been achieved in two different ways. The first method, which is older, quite rare and more expensive, sees the application of a thin layer of metal (silver or gold) onto the pottery using mercury as adhesive material;16 according to research by K. Burkhardt, the second method, cheaper and more frequently used, was obtained by polishing the surface before firing the pot.17 Because both methods are exclusive to workshops in Caere, we can assume a provenance from that city for the bucchero kantharos found at Naukratis. The presence in Naukratis of bucchero with silvering decoration is all more significant, since in the whole of the Mediterranean I only know of one other sherd with such a decoration, and that is a sherd from the Heraion in Samos.18 According to B. Bouloumié, some bucchero kantharoi with silvering decoration were found in the wreck of Cap d’Antibes; but these materials are badly documented, because they are preserved in private collections.19 The provenance from Caere of bucchero vases found in Naukratis and Samos is not surprising: as we have seen, many bucchero vases found at Carthage probably come from Caere. Using other evidence, I am able to add that some bucchero vases from Miletos were also made in Caere. Because of the director of the Miletos excavations, V. von Graeve’s, interest in pottery analysis, it has been possible to analyse some bucchero samples found in the Aphrodite sanctuary on Zeytintepe in Miletos. The as yet unpublished results of the thin sections and the petrological analysis carried out in the laboratory of SOB University of Naples by G. Trojsi show values very close to those found with similar analyses by K. Burkhardt in his large research project on bucchero pottery from southern Etruria.20 It is noteworthy that in the older excavations in Carthage more than 20 Etrusco-Corinthian vases were found: they are Etruscan imitations of the Corinthian pottery and were very popular in Vulci and Tarquinia, but less so in Caere, the Etruscan city that imported the largest quantity of Corinthian pottery and

Figure 4 Etruscan bucchero kantharos from Naukratis

that therefore had less interest in the imitations (Fig. 5).21 The Etrusco-Corinthian vases found in Carthage were classified by J. Szilágyi as imports from Vulci and Tarquinia, all dating to the first half of the 6th century bc. These classifications have recently been confirmed by some new Etrusco-Corinthian fragments that were found in the two German excavations near the Decumanus Maximus of Roman Carthage, led respectively by Fr. Rakob and H.G. Niemeyer;22 the number of imports from Tarquinia, particularly for the vases of the ‘Pittore senza Graffiti’ has thus increased. In Carthage, Etruscan transport amphoras have yet to be found, but it would not be surprising if they were to be identified.23 Again these results are compatible with the Etruscan finds from Miletos, where some bucchero sherds may belong to vases from Tarquinia and perhaps Vulci: Miletos is the find spot of the only Etruscan transport amphora identified up to now in all of the eastern Mediterranean.24 How can we interpret the bucchero and Etrusco-Corinthian vases found in Carthage? I think they may be something more than simply objects of trade or exotic pieces for deposition as grave goods, especially if we connect these pots, whose numbers increase from the third quarter of the 7th century to 550 bc, with later events. We know, thanks to many historians, that the relationships between Carthage and the cities of southern Etruria were particularly good and intensive. The role of the Etruscans was not a secondary one in the middle of the 6th century bc, when the expansionist policy of Carthage, which was destined to become almost an empire in the following years, began with the famous expedition led by Malcus in Sicily and Sardinia.25 This is stated by Herodotus himself, who expressly mentions the alliance between Carthage and Caere against the Greeks of Phokaia in the battle of the Sardinian Sea in about 540 bc.26 Aristotle in his Politiká cites a deliberate, official alliance between Etruscans and the Punic empire: his references seem to indicate the existence of written documents (graphaí) concerning trade and military agreements.27 The existence of such treaties between Carthage and Caere can be supported by other historical traditions, such as the information from Polybios about the first treaty between Rome and Carthage, dated to about 509 bc.28 Scholars currently accept the existence of this first treaty, and only a few are convinced that this is an invention, a retrojection in the past of the treaty between Rome and Carthage dating to 348 bc. Some years ago C. Ampolo stressed the authenticity of this early treaty and dated it to the end of the 6th century bc.29 At that time the relations between Caere and Carthage were fruitful and included a military alliance: it is widely accepted that only after the battle of the Sardinian Sea, in the second half of the 6th century bc, the Carthaginian obtained control of Sardinia, while the Etruscans Naukratis: Eastern Greeks in Egypt | 189

Naso

1

2

3

Figure 5 Etrusco-Corinthian pottery from Carthage: round aryballos of the Poggio Buco group; cup of the Macchie Bianche group;Vulcian aryballos

began their domination of Corsica. An important find (Fig. 6) shows the existence of close personal relationships between Etruscans and Carthaginians, and probably also reflects the custom of both people visiting one another in their respective cities. In a tomb of the Santa Monica necropolis in Carthage, one of the few Archaic Etruscan inscriptions outside Etruria has been found. Inscribed on the reverse of an ivory tessera in the shape of a quadruped is: mi puinel karqazie elsf[—-]na, meaning ‘I belong to Puinel the Carthaginian … ’. The little tablet, dating to the last quarter of the 6th century bc, is a tessera hospitalis. It is almost an identity card, destined to match another similar piece belonging to an Etruscan. Only few other ivory tesserae hospitales are known; it is not by accident that one in the shape of a panther was found outside Etruria, in Rome.30 In the last quarter of the 6th century bc, then, after the battle of the Sardinian Sea, we can clearly see direct and personal contacts between southern Etruscans and Carthaginians, both barbaroi and therefore natural allies against the Greeks.31 From this perspective it is also possible to accept the proposal of D. Berges, who also included Etruscan people among the possible clients visiting the state archive in Carthage.32

Figure 6 Etruscan tesserae hospitales from Carthage and Rome

190 | Naukratis: Greek Diversity in Egypt

The end of the 6th century bc at Caere is the age of Thefarie Velianas, the king responsible for the construction of the socalled temple B in the sanctuary of Pyrgi, the main harbour of the city. Thefarie Velianas is expressly named in the three gold tablets from Pyrgi, two in the Etruscan language and one in Punic. According to many scholars, the honour of putting a Punic inscription in one of the main sanctuaries near Caere is closely connected to the alliance quoted by Herodotus.33 Therefore we may conclude that the presence of rich Archaic Etruscan finds in the western part of North Africa was due not only to trade, but in some cases also to direct relations with Etruria, particularly with Caere. In the Greek colonies of North Africa, on the contrary, we can ascribe the Etruscan finds to indirect contacts through trade. This may be the case with the bucchero kantharoi in Tocra and Naukratis and the bucchero oinochoe from Naukratis (Appendix 1, no. 22), published by A. Johnston (Fig. 7). I wish to stress that, in my opinion, the other bucchero sherds from Naukratis published as Etruscan are not Etruscan, both because of their shape and their clay; I have to add that P. Perkins found further bucchero sherds from Naukratis in the British Museum storerooms, which I have never seen (see Appendix 1, no. 23

Etruscan and Italic Finds in North Africa, 7th–2nd century BC

0

5 cm

Figure 7 Etruscan bucchero oinochoe from Naukratis (BM GR 1888.6–1.643a)

Figure 8 Etruscan bucchero kantharoi from Greek sanctuaries: top left, Perachora; top right, Ialysos; bottom, Leontinoi

Figure 9 Frog belonging to a bronze infundibulum from Cyrene

with note 57). The bucchero sherds from Karnak in Egypt and Tipasa in Algeria shall be mentioned only briefly: thanks to the kind information provided by P. Touillard we know that in Karnak only one sherd has been found, of a small amphora dated to 600 bc.34 In general we can consider the bucchero vases found in Greek sanctuaries as gifts from Greek merchants returning home, as can be seen from the Greek inscriptions on bucchero vases dating to the first half of the 6th century bc found in Perachora, Ialysos (on Rhodes) and in Sicily (in Selinous and now in Leontinoi): we know of one Néarchos in Perachora and probably one Leukios in Leontinoi (Fig. 8).35 What is quite surprising in North Africa is the absence of bucchero at Cyrene.36 This may, of course, be due to the scarcity of pottery published from the site, at least until this conference. What has been found in Cyrene, however, is a bronze fragment belonging to an infundibulum, a very elaborate Etruscan funnel that was part of a wine set.37 It is in the form of a little bronze frog with a cross hole, and a cut-away to fit a tang, that hinged the frog to a bronze handle (Fig. 9). The frog held a strainer, originally attached with rivets. Both frog and strainer could be raised backwards and the funnel be used alone. The infundibula usually have one handle in the form of a lyre; they end in a duck’s head with a long bill, or more rarely in a ram’s head. It is quite common for the hinge to have the shape of a T, or, if it is figured, of a couchant lion, or a frog or more rarely a sphinx. They are a typical Etruscan invention and were, of course, used to pour wine, for instance from a krater into an oinochoe or from an oinochoe into a kantharos. In the second half of the 6th century bc they were very popular all over the Mediterranean. From this perspective we can consider infundibula in the second half of the 6th century bc the counterpart in bronze of the bucchero kantharoi in the first half of the same century: a real Etruscan marker, one of the appreciated turrhnoi/ xalkoi celebrated in ancient Greek literature.38 Since the study of M. Zuffa in 1960 that listed 28 tools, many new finds have surfaced: I am now able to list more than 80 infundibula, belonging to at least four main types: 1. lyre-handled (the most numerous, with sub-types); II. San Martino in Gattara; III. Palmette-handled; IV. special forms, including tools that are not Etruscan. Although many have appeared on the art market without any provenance (Fig. 10), the find spots, when known, are significant (Fig. 11). In Italy they are quite widespread: the main source is in southern Etruria, but some tools have also been found in Campania, Umbria, ancient Picenum (corresponding to the modern-day southern Marche and northern Abruzzo) and in the Veneto. Outside Italy I know of three in Spain, one in Cyrene, three (or more) in Olympia (one with a Greek inscription), one in Argos, one in Ialysos on Rhodes. Another funnel was found in Switzerland, in the Arbedo hoard.39 Two bronze fragments representing a duck’s head from Carthage and from Didyma may belong to infundibula or to ladles, which have also been found in Greece.40 This wide distribution, including not only many Italic regions, but also the Mediterranean basin and central Europe, where Hallstatt imitations are also known, and the provenance of many pieces from illegal excavations of the 19th and 20th centuries, seem reason enough to localize the workshop in southern Etruria. Against current opinion, which presumes only one workshop in Volsinii, the different forms (or sub-types) of the lyre-handled tools are enough to postulate the existence in southern Etruria of more than one workshop. One Naukratis: Eastern Greeks in Egypt | 191

Naso of these may be located in Vulci,41 where the most famous Etruscan bronze workshops flourished, which were responsible both for masterpieces, such as the rod tripods found on the Athenian Acropolis and in a Celtic grave in Bad Dürkheim near Speyer in Germany, and everyday tools, such as the countless Schnabelkannen, found above all in the territories north to the Alps, but not yet in Greece, that were destined for long-distance trade, too.42 From the Late Archaic to the Hellenistic periods Finally, a few words on the later period, concerning EtruscanPunic relations only. Herodotus (6.17) reports that in the early 5th century bc Dionysios of Phokaia fought against Etruscans and Carthaginians, who, according to the same historian, also shared a common fate in the battles lost against the Greeks of Syracuse (the Carthaginians in Sicily at Himera in 480 bc, the Etruscans in the sea of Cumae in Campania in 474 bc). From Diodorus of Sicily (10.11.1) we learn that at the end of the 4th century bc Etruscan mercenaries fought for Agathokles of Syracuse against the Carthaginian army. The few, but relevant, archaeological finds may confirm the relationships that are behind these contacts. In a chamber tomb in Tunisia near Ksour es Saaf, not far from Mahdia, an impressive triple-disc cuirass of gilded bronze was found, perfectly preserved, in 1909. Similar cuirasses, datable to the end of the 4th century bc, are common in southern Italy among Samnites, Lucanians and other Italic populations. Initially a bronze belt was thought to be associated with the tomb group, too, but a recent restoration has excluded the presence of this belt, a typical south Italian product.43 So now the interpretation of the cuirass without the belt is less clear: is it war booty? Or is it the panoply of an Italic soldier, or

better of an officer of Agathokles? Both are possibilities. In Carthage, in the so-called Salammbô tophet, a 50cm high marble cippus was found. Such cippi are typical markers for male tombs in Caere from the 4th century bc onwards. It would seem very probable that this cippus was the gravestone of an Etruscan who died in Carthage, perhaps in the early 3rd century bc.44 In Carthage and in Cyrene there are also some red-figured Etruscan plates of the Genucilia class, dating to the end of the 4th–early 3rd century bc.45 During the 4th and 3rd centuries bc, when the Roman armies were conquering Etruria city by city, some north Etruscans probably fled their land and tried to find a new homeland in Africa. This could explain how the longest Etruscan inscription, the so called Liber Linteus Zagrabiensis, now in Zagreb, written on a mummy linen cloth and dated by radiocarbon to 390–25 bc, came to be found in Egypt.46 There is no record of the circumstamces of the transport of the Etruscan book to Egypt: it has been presumed to have happened quite late, perhaps after the bellum Perusinum (41–40 bc), because its characteristics suggest that the book was written in the Perugia area.47 Finally, eight Etruscan inscriptions on three boundary stones, found in Tunisia in the hinterland of Carthage, record the same person, the Etruscan Marce Unata Zutas. They probably relate to the escape from Clusium in 82 bc of the Roman consul Cn. Papirius Carbo and his Etruscan friends, quoted by Appianus, because Unata is a typical name of Clusium and its district.48 We can conclude that the presence of Etruscans in North Africa was a persistent phenomenon, a feature of the histoire de longue durée of the region.

Figure 10a & b Etruscan bronze infundibula of the lyre handled type, nos 26 and 34

192 | Naukratis: Greek Diversity in Egypt

Figure 11 Distribution of infundibula in the Mediterranean

I

■

■▲

Cancho Roano

II

●

Xàbia

■

III

▲

IV

▼

Carthage ★

■ ▼ ■ ■ ■ ▼■ ■ Campovalano ■ Cales ▼▼ Cuma ■ ▲ Nola ■ ■Sala Consilina

★

?

M.Bubbonia ● Gela ●

Volsinii ■ ■

■

San Martino in Gattara ■ Numana Populonia ■ ●

■■ ●

■

Ceregnano Marzabotto

■

Arbedo

▼

■

Cyrene

■Argos■ Olympia

Trebenishte

▼

Novi Pazar

■ Lindos

★ Didyma

● Pantikapaion

Naso

49

Appendix 1 Etruscan and Italic Artefacts from North Africa Algeria Gouraya 1. Small bronze disc (diam. 7.7cm) with incised decoration and an inscription, dated to the 3rd century bc. Found in a Punic grave near Gouraya, approximately 130km west of Algiers. Liébert 1996. About the inscription: Briquel 2004, with previous bibliography.50

Tipasa 2. Bucchero pottery is mentioned, but is still unpublished. Hase 1989, 327-8 note 2.

Tunisia Carthage 3. Bucchero pottery: 28 small amphorae, 12 oinochoai, two kotylai, 11 cups, 11 kantharoi. Fig. 2 Hase 1989, 383-92; for the unpublished sherds from the excavations led by H.G. Niemeyer: ibid., 330-2, note 15; Docter 1993, 229-30 nos 234: bucchero and impasto (?) 4. Etrusco-Corinthian pottery. Fig. 5 MacIntosh Turfa 1982; Hase 1989, 377-8; Docter 1993, 229-30 nos 25-6 (Etrusco-Corinthian sherds); Trias 1999, nos 26-7 (two non-joining Etrusco-Corinthian sherds probably belonging to the same plate of the ‘Pittore Senza Graffito’); Szilágyi 1998: 375 no. 61 (cup of the ‘Pittore delle Code Annodate’), 414 no. 15 (Vulcian aryballos), 444 no. 19 (plate with foot of the ‘Pittore senza Graffito’) 448 nos 132-3 (plates with foot of the ‘Pittore senza Graffito’), 526 no. 34 (cup of the ‘Macchie Bianche’ Group), 532 no. 22 (cup of the Poggio Buco Group), 533 no. 42 (round aryballos of the Poggio Buco Group) 601 no. 72 (alabastron of the ‘Galli Affrontati’ Cycle, Michigan group, standardized), 684 no. 98 (unattributed sherd), 694 (general considerations). 5. Bronze handle ending in a duck’s head, belonging to a ladle or infundibulum. Mackensen 1999, 540-1, no. 35, figure 1.1, pl. 44.1. 6. Etruscan bronze statuette. Hase 1989, 378. 7. Seven bronze Schnabelkannen, probably Vulcian. Hase 1989, 378. 8. Ivory tessera hospitalis with Etruscan inscription. Fig. 6 Petersen 1903, 23; Martelli 1985a, 237 fig. 91; Martelli 1985b; Hase 1989, 374. 9. Etruscan marble cippus, probably Caeretan. Pallottino 1966, 12, pl. I.2 (= Pallottino 1979, 393, pl. 8.1); Hase 1996.

194 | Naukratis: Greek Diversity in Egypt

10. Etruscan red-figured plates of the Genucilia class (at least six examples) von Hase 1996, 188 with previous literature; Morel 1990, 85-6, fig. 22 (3 further sherds of Genucilia plates); Poulsen 2002, 90.

Uadi Milian 11. Three cippi with Etruscan inscriptions. Heurgon 1969a, 1969b; Carruba 1976; Colonna 1980b, 1-5; Sordi 1995, 115-20.

Utica 12. Bucchero cup. Morel 1981, 484-5, note 100 with previous bibliography.

Ksour es Saaf 13. South Italian triple-disc cuirass. Tunis, National Museum. Colonna 1981, 177-8, pl. 8; Tagliamonte 1994, 153-4; Carthage 1995, 147-9; Ben Younès 1997, 2001; Tagliamonte 2004, 161 note 103.

Libya Cyrene 14. Bronze ladle handle from the second Artemision in Cyrene, which is dated (p. 226) to 450–400 bc. Pernier 1931, 214, fig. 40.51 15. Infundibulum handle. Warden 1990, 8-9, no. 17, pl. 5. Fig. 9 16. Bronze ladle handle with incised decoration. Warden 1990, 55, no. 402, pl. 40.

21. Boston, Museum of Fine Arts, unpublished (Inv. no. 88.959). Fig. 4 Sherd of kantharos with three horizontal grooves beneath the rim and knotching on the sharp carination. Johnston 1982, 38, pl. 4, published five bucchero sherds (four are preserved in the British Museum, one in University College, London). 22. Johnston 1982, sherd no. 1 (Inv. no. BM GR 1888.6-1.643a) is an oinochoe rim formed by three joining sherds. H 6.8cm. It may belong to an oinochoe Rasmussen 3a (Rasmussen 1979, 161), which is documented at the end of 7th–6th century bc (Rasmussen 1979, 78-9, pls 7-8) or Rasmussen 7a, very common in the first half of the 6th century bc (Rasmussen 1979, 84-5, pl. 16).56 Fig. 7 23. Johnston 1982, sherd no. 2 (UCL-357) was never reproduced nor photographed. I have never seen it, in spite of the kind efforts of Dr. A. Johnston.57 24. Two sherds of skyphoi of the Gnathia class 2nd century bc. Prins de Jong 1925, 70, nos 1-2.

Karnak 25. Karnak, storeroom, Inv. no. A 960. One sherd belonging to a small amphora (kind information of P. Rouillard). Rouillard 1985; Hase 1989, 327-8 note 2.

Tell Defenneh (?)

17. Etruscan red-figured plate. Bacchielli 1976; Colonna 1981, 183 note 107; Bacchielli 1986, 375 note 15.

26. Cairo, National Museum, Inv. no. 27963. Bronze handle with two plastic horse heads on the top, belonging probably to an Etruscan podanipter. Edgar 1904, 81, no. 27963, pl. 10.58

18. Two cups of the Gnathia class. Kenrick 1987, 3-4, nos 20-1.52

(?)

Leptis Magna53 Tocra 19. Sherd belonging to the handle of a bucchero kantharos. Boardman and Hayes 1973, 58 no. 2246, pl. 31 from Deposit II, dated (p. 3) to 590–565 bc, a votive deposit of the sanctuary of Demeter and Kore.

Egypt Naukratis54 20. Once in the Fr. W. v. Bissing collection, whereabouts unknown.55 Fig. 3 Rim of a bucchero kantharos with remains of one handle. Silvering and arches on the shoulder. Dimensions: 55mm (height), 60mm (width). Prins de Jong 1925, 55-6 no. V.2, pl. 3 (top right).

27. Zagreb, National Museum, Inv. no. 1 (bought in Egypt). Liber linteus. Roncalli 1980b; Roncalli 1985; Mirnik 1986. For the chronology: Srdocv et al. 1990.

(?) 59 28. Turin, National Museum, Inv. no. A’ 43 (bought in Egypt). Fig. 1 Antennae sword. Angelucci 1876, 25; Bianco Peroni 1970, 113 no. 305, pl. 45; Venturoli 2002, 36-7, no. A’ 43. 29. From Alexandria (presumably bought in Alexandria). Cairo, National Museum, Inv. no. 27902. Etruscan mirror with Dioscuri and two shields. Edgar 1904, 68, no. 27902, pl. 18.60

Etruscan and Italic Finds in North Africa, 7th–2nd Century BC

Appendix 2 Etruscan Bronze Infundibula I LYRE-HANDLED TYPE Populonia (Livorno) 1. Grave ‘dei Flabelli di Bronzo’. Florence, Museo Archeologico, Inv. no. 89332. Zuffa 1960, 178-9, no. 1, pl. 21; Schindler 1998, 276 (Typ I). Duck’s head and hinge in the form of T. Grave ‘dei Colatoi’. Florence, Museo Archeologico, Inv. no. 92589-92590. 2. De Agostino 1961, 86, no. 4, fig. 24.1; Terrosi Zanco 1974, 163; Schindler 1998, 276 (Typ I). Duck’s head and hinge in the form of a quadruped. 3. De Agostino 1961, 86, no. 5, fig. 24.2; Terrosi Zanco 1974, 163; Schindler 1998, 276 (Typ I). Duck’s head and hinge in the form of a couchant lion.

Bisenzio (Viterbo) 4. Grave 74 (540–520 bc). Rome, Museo Nazionale Etrusco di Villa Giulia, Inv. no. 57165/3. Colonna 1980a, 45 note 9, figs 3-4; Schindler 1998, 275 (Typ I). Duck’s head and hinge in the form of a frog.

Volsinii or Todi 5. Florence, Museo Archeologico Nazionale, Antiquarium. Zuffa 1960, 186-7 no. 13, pl. 29.a-b. Handle with the hinge in the form of a couchant lion.

Castelgiorgio (Terni) 6. Florence, Museo Archeologico Nazionale, Inv. no. 82.892. Zuffa 1960, 190-1 no. 18, pl. 32.c-d. Duck’s head and hinge in the form of a quadruped.

Volsinii (Terni) 7. Crocefisso del Tufo, grave 17. Bizzarri 1962, 89-90, 333, 34061 fig. 30; Schindler 1998, 276 (Typ I). Bottom of the funnel.

Falerii Veteres, grave 34 (LIII) (Viterbo) 8. Rome, Museo Nazionale Etrusco di Villa Giulia, Inv. no. 371. Cozza and Pasqui 1887, 175d;62 Cozza and Pasqui 1981, 170 no. 8 (grave 48).63 Duck’s head and hinge in the form of a couchant lion.

Castro, grave della Biga (530–520 bc) female deposition (Viterbo) 9. Moretti Sgubini and De Lucia Brolli 2003, 382, fig. 37. Duck’s head and hinge in the form of a couchant lion.

Castel San Mariano, grave del Carro (Perugia) 10. Perugia, National Museum, Inv. no. 1433. Zuffa 1960, 192-3 no. 21, fig. 7: Höckmann 1982, 159; Schindler 1998, 275 (Typ I). Broken handle.

Todi (Perugia) 11. Rome, Museo Nazionale Etrusco di Villa Giulia, Antiquarium, Inv. no. 24594. Zuffa 1960, 185 no. 8, pl. 25. Duck’s head and hinge in the form of a couchant lion.

22. Poggio Sommavilla (Rome) Whereabouts unknown. Funnel Bellelli 2006, 94.

Cuma (Naples)

12. Rome, Museo Nazionale Etrusco di Villa Giulia, Antiquarium, Inv. no. 24595. Zuffa 1960, 193 no. 22, pl. 33.b-c. Duck’s head; hinge not preserved.

23. Naples, Museo Nazionale, Inv. no. 86069. Zuffa 1960, 186, no. 11, pl. 27; Albore Livadie 1985, 137 note 49; Grassi 2003, 502, note 70. Duck’s head and hinge in the form of a couchant lion.

Vetulonia (?) (Grosseto)

Castellammare di Stabia (?)

13. Collection Stefani. Zuffa 1960, 185 no. 9, pl. XXVI.a. Only the hinge in the form of a couchant lion remains.

Ceregnano near Adria (Rovigo) 14. Whereabouts unknown (perhaps to be identified with no. 43 in Manchester?). Zerbinati 1994, 148-9, fig. 1; Schindler 1998, 276 (Typ I). Ram’s head and hinge in the form of a couchant lion.

Marzabotto (Bologna) 15. Marzabotto, Museo P. Aria, Inv. no. B 9. Zuffa 1960, 197 no. 27, pl. 35.d; Muffatti 1968, 155, no. 32, pl. 21.b 3; Schindler 1998, 275 (Typ I). Only the funnel remains.

Casalfiumanese (Bologna) 16. Bologna. Museo Civico. Zuffa 1960, 193-4 no. 23, pl. 34. Ram’s head (hinge not preserved).

Belmonte Piceno (Ascoli Piceno) 17. Ancona, Museo Archeologico Nazionale, Inv. nos 12563 (funnel), 12581 (handle). Zuffa 1960, 187-9 no. 15, pl. 30. Duck’s head and hinge in the form of a frog. 18. Once Ancona, Museo Archeologico (destroyed during the Second World War). Zuffa 1960, 194-5 no. 24, pl. 35.a-c. Lyre handled-type with a peculiar funnel (pastiche?).

Tolentino, grave near Porta del Ponte (Macerata) 19. Tolentino, Museo Civico, Inv. no. 1854/1. Zuffa 1960, 186 no. 12, pl. 28; Massi Secondari 1982, 38-9, note 1, fig. 2. The hinge is in the form of a couchant lion.

Numana (Ancona) 20. Ancona, Museo Archeologico Nazionale, Inv. no. 50769. Landolfi 1997, 237, no. s.2. Double lyre handled, duck’s head and hinge in the form of a couchant lion.

Campovalano, grave 2 (Teramo) 21. Chieti, Museo Nazionale, Inv. no. 5146. Zanco 1974, 51-2, no. 18; Schindler 1998, 275 (Typ I); Grassi 2003, 502, note 70. Duck’s head and hinge in the form of a T.

24. Albore Livadie 1985, 137 note 49; Grassi 2003, 502, note 70.

Sala Consilina (Salerno) 25. Paris, Petit Palais, Inv. no. 235. Zuffa 1960, 195-6, no. 25, pls 36-37. Double lyre handled with ram’s head and hinge in the form of a sphinx.

Provenance and whereabouts unknown 26. Zuffa 1960, 180 no. 2, pl. 22.2; Terrosi Zanco 1974, 163; Schindler 1998, 276 (Typ I) DAI Rome, Inst. Neg. 29.441, 29.442, 29.443. Fig. 6 Duck’s head and hinge in the form of a T.

Provenance unknown 27. Florence, Museo Nazionale, Antiquarium, Inv. no. 1537. Zuffa 1960, 183-4 no. 6, pl. 24. Duck’s head and hinge in the form of a couchant lion. 28. Florence, Museo Nazionale, Antiquarium, Inv. no. 1538. Zuffa 1960, 189-90 no. 17, pl. 32.a-b. Duck’s head and hinge in the form of a frog. 29. Florence, Museo Nazionale, Antiquarium. Only the hinge in the form of a couchant lion remains. (May be no. 13?) 30. Milan, Museo Civico Archeologico, Inv. no. 1055. Zuffa 1960, 185 no. 10, pl. 26.b-c. Duck’s head and hinge in the form of a couchant lion. 31. Raccolta Benedetto Guglielmi. Città del Vaticano, Museo Gregoriano Etrusco, Inv. no. 34864. Magi 1941, 230-1, no. 117 pl. 68; Zuffa 1960, 187 no. 14, pl. XXIX.c. Only the hinge in the form of a couchant lion remains. 32. Turin, Museo Archeologico Nazionale, Inv. no. 933. Zuffa 1960, 189 no. 16, pl. 31. Duck’s head and hinge in the form of a frog. 33. Perugia, Museo Archeologico Nazionale, Inv. no. 600. Saioni 2003, 56. Duck’s head and hinge in the form of a frog. 34. London, British Museum, Inv. no. GR 1937.10-21.1 (Bronze 2469). Fig. 10b Walters 1899, 322, note 2469. Duck’s head and hinge in the form of a quadruped.

Naukratis: Greek Diversity in Egypt | 195

Naso 35. Leipzig, Antikenmuseum der Universität, Inv. no. MB 4-M 53a. Bought in Vienna in 1917 from L. Pollak. Zuffa 1960, 184-5 no. 7, pl. 23.c-d; Paul 1988. Rest of the handle with the hinge in the form of a couchant lion. 36. Vienna, Kunsthistorisches Museum, Inv. no. VI, 2962. Zuffa 1960, 182-3 no. 4, pl. 23.a. Duck’s head and hinge in the form of a T. 37. Vienna, Kunsthistorisches Museum, Inv. no. VI, 4637. Zuffa 1960, 183 no. 5, pl. 23.b. Duck’s head and hinge in the form of a T.

Olympia, sanctuary of Zeus (Olympia, Museum)64 49. Inv. no. Br 12844. Zuffa 1960, 180-2 no. 3, pl. 22.b-c. Duck’s head without any rest of a hinge. 50. Inv. No. B 286. Zuffa 1960, 191 no. 19, pl. 33.a. Partly preserved, without any rest of a hinge. 51. Inv. No. B 4574. Siewert 1991, 82 no. 7, pl. 9.2/3. Handle partly preserved, with a duck’s head, without any rest of a hinge.

Lindos (Rhodes), sanctuary of Athena

38. New York, Metropolitan Museum, Inv. no. 34.11.8. Zuffa 1960, 196-7 no. 26, pls 38-39. Ram’s head and hinge in the form of two couchant lions

52. Istanbul, Archaeological Museum, Inv. no. 3495 m, 3503 m. Zuffa 1960, 191-2 no. 20, fig. 6. Duck’s head without any rest of hinge

39. Newcastle upon Tyne, Shefton Museum of the University, Inv. no. 139. Shefton 1970, 55-6, figs 5-6. Duck’s head and hinge in the form of a couchant lion.

Libya Cyrene, sanctuary of Demeter and Persephone

40. Newcastle upon Tyne, Shefton Museum of the University, Inv. no. 667. Unpublished. 41. Collection H. Cahn, Basel. Das Tier in der Antike 1974, 52 no. 311, pl. 52. Only the hinge in the form of a sphinx remains. 42. Whereabouts unknown. Kunstwerke der Antike. Auktion 51. Münzen und Medaillen AG, Basel 1975, 102 no. 228; Treister 1990, 165; Schindler 1998, 276. Duck’s head and hinge in the form of two couchant lions. Two further lions are on the lid (pastiche?). 43. Manchester Museum, Inv. no. 29973. Perhaps to be identified with no. 14 from Ceregnano. MacIntosh Turfa 1982, 175 no. 33, pl. 14.d; Schindler 1998, 276 (Typ I). Ram’s head and hinge in the form of a couchant lion. 44. Private collection (CH). Reusser 1986, 27, no. 6.2. Duck’s head and hinge in the form of a couchant lion. 45. From the Gorga collection. Rome, Museo Nazionale Romano. Lodovici 1999, 49, fig. 12. Duck’s head without any hinge (the funnel is a strainer, too). 46. Stuttgart, Württembergisches Landesmuseum (Inv. no. 3. 190) Hinge in the form of a couchant lion. 47. Formerly in the collection E. Berman (now Museo Archeologico, Civita Castellana?) De Lucia Brolli 2004, fig. 2 (top left). Handle, probably complete.

Greece Argos, Heraion 48. Fletcher De Cou 1905, 203-4, no. 31, pl. LXXVI. Only the hinge in the form of a frog remains.

196 | Naukratis: Greek Diversity in Egypt

53. Warden 1990, 8-9, no. 17, pl. 5. Only the hinge in the form of a frog remains.

Spain Cancho Roano (Estremadura) 54. Badajoz, Museo Arqueológico Provincial. Celestino Pérez 1991, 78, fig. 12a; Pallottino 1992, 179, 260, no. 304; Schindler 1998, 275 (Typ I); Celestino Pérez and de Zulueta 2003, 56-8, 92, n. 213. Ram’s head without any rest of a hinge.

From the sea near Jávea (Alicante) 55. Museu Arqueològic i Etnogràfic ‘Soler Blasco’, Xàbia.65 Vives-Ferrándiz Sánchez (forthcoming). Duck’s head without any rest of the hinge.

Switzerland Arbedo, hoard 56. Schindler 1998, 80-2, 275, 321 no. 153 [154], 397 pl. 7. A funnel and a duck’s head, belonging not necessarily to the same tool.

II SAN MARTINO IN GATTARA TYPE San Martino in Gattara, male grave 15 (530–520 bc) 57. Ravenna, Museo Nazionale. Bermond Montanari 1975, 74, fig. 4; Colonna 1980a, 45-6; Bermond Montanari 1982, 172-4, no. 20, pl. 93; Treister 1990; Schindler 1998, 276 (Typ IIIa).

Pantikapaion 58. Moscow, Pushkin State Museum of Fine Arts, Inv. no. GMII. M 410. Treister 1988; Treister 1990; Treister 1991, 73-4; Schindler 1998, 276 (Typ IIIa); Treister 1999, 815; Naso 2000, 180-1, pl. IV.2; Naso 2001, 179, fig. 8.

Populonia 59. Collection Gasparri. Populonia, Museo Archeologico, Inv. no. 1237. Romualdi 2001, S 2.

Provenance unknown 60. Geneva, Musée d’Art et d’Histoire, Inv. no. MF. 1170. Treister 1990, 166; Schindler 1998, 276 (Typ IIIa).66 61. Gela (Caltanissetta), Archaic wreck Gela, Museo Archeologico, Inv. no. 38303 Panvini 2001, 31,62.

Monte Bubbonìa (Mazzarino, CL), grave 13/1971 (550–500 bc) 62. Caltanissetta, Museo Archeologico, Inv. no. 34981. Panvini 2003, 194.

Close to San Martino in Gattara type Monte Bubbonìa (Mazzarino, CL), grave 10/1955 63. Gela, Archaeological Museum, Inv. no. 9302. Pancucci and Naro 1992, 126 no. 397 pl. 31.3. The tomb group is dated to the early 5th century bc by an Attic olpe by the Painter of Berlin 2268 (ARV2, 156 no. 63).

III PALMETTE-HANDLED TYPE Nola (?) 64. Brussels, Musée Royale, Inv. no. R 1127. Meester de Ravestein 1884, 329-30, no. 1127. Frog on the lid (pastiche?).

Provenance unknown 65. Vienna, Kunsthistorisches Museum, Inv. no. VI-932. Zuffa 1960, 197-8 no. 28, fig. 8, pl. XL.

Spain Cancho Roano (Estremadura) 66. Badajoz, Museo Arqueológico Provincial. Celestino Pérez 1991, 78, fig. 12b; Celestino Pérez and de Zulueta 2003, 56-8, 92, n. 233. Hinge in form of a lion.

IV OTHER TYPES Bisenzio 67. Olmo Bello, grave 80 (excavations Benedetti 1927-31). Rome, Museo Nazionale Etrusco di Villa Giulia. Schindler 1998, 276 (Typ III). 68. Bazzano (L’Aquila), grave 1566 (excavation V. d’Ercole). Celano, Museo di Preistoria Unpublished; kind information of J. Weidig (Mainz-Marburg)

Trevignano Romano, grave AnnesiPiacentini. 69. Trevignano Romano, Museo Civico. Moretti 1967, 65 no. 47, pl. (bottom right); Colonna 1980, 45 note 8; Schindler 1998, 276 (Typ III); Bellelli 2006, 41-54.

Etruscan and Italic Finds in North Africa, 7th–2nd Century BC Provenance unknown 70. Rome, Museo Nazionale Etrusco di Villa Giulia, Antiquarium, Inv. no. 51370. Zuffa 1960, 204 no. 31, pl. XLV; Schindler 1998, 276 (Typ III).

74. Museo di Santa Maria Capua Vetere (?) Terrosi Zanco 1974, 162-3 (included in the exhibition ‘Gli Etruschi in Campania’, Teano 1963); Schindler 1998, 276 (Typ I); Grassi 2003, 502, note 70.

Trestina

Provenance unknown

71. Florence, Museo Archeologico, Inv. no. 77813. Tarchi 1936, pl. C (bottom, middle); Colonna 1980, 45 note 9; Romualdi 1991, 629; Schindler 1998, 276 (Typ I); Naso (forthcoming).

75. Rome, Museo Nazionale Etrusco di Villa Giulia, Antiquarium, Inv. no. 24689. Zuffa 1960, 203-4 no. 30, pl. XLIV.

Cales, grave 89 72. Passaro and Ciaccia 2000, 21; Grassi 2003, 502, note 70.

Santa Maria Capua Vetere 73. Berlin, SMPK, Antiquarium, Inv. no. 6332 and Copenhagen, Kunstmuseet, Inv. no. 3284. Brown 1960, 111-112; Zuffa 1960, 198-203 no. 29, pls 41-43; Bellelli 2006, 41-54.

Provenance unknown 76. Warsaw, National Museum, Inv. no. 147078. Dobrowolski 1966, 377-8, figs 1, 3.

The following infundibula are not Etruscan:67

1 2 3 4 5 6 7 8 9 10 11

12 13

Novi Pazar 79. Beograd, National Museum. Mano-Zissi and Popovic 1969, 80-1, pl.. 8, 39; Mano-Zissi and Popovic 1971, 195, pl. 56-60; Popovic 1975, 89, fig. 18; Treister 1990, 166; Schindler 1998, 276 (Typ III).

Olympia, sanctuary of Zeus (Olympia, Museum) 80. Inv. no. Br 14030 Furtwängler 1890, 147, no. 924-924a. 81. Inv. no. Br 12866 Furtwängler 1890, 147, unnumbered, between 924-5).

77. Rhodes, Museum. Zuffa 1960, 207 no. 33, fig. 10.

Fig. 1 after Bianco Peroni 1970, 113 no. 305, pl. 45; Fig. 2 after Hase 1989, fig. 29; Fig. 3 after Prins de Jong 1925 ; Fig. 4 photo MFA, Boston; Fig. 5 after Hase 1989, pl. 28.II; Fig. 6 after Pugliese Carratelli 1986, figs 55–6; Fig. 7 Drawing Kate Morton; Fig. 8 after Hase 1997 and Rizza 2003; Fig. 9 after Warden 1990; Fig. 10a DAI Rome, Inst. Neg. 29.442; Fig. 10b the British Museum; Fig. 11 the author.

*

78. Sofia, Archaeological Museum. Zuffa 1960, 204-7 no. 32, pl. 46.

Rhodes, sanctuary of Apollo Erethimios

Illustration credits

Notes

Trebenishte, grave VII

I wish to thank the organizers, particularly U. Schlotzhauer and A. Villing, for the kind invitation to the Naukratis conference. Thanks are due to the directors of excavations who permitted me to examine unpublished material from many sites, namely Prof. V. von Graeve (Miletos), Prof. A. Furtwängler and Dr. H. Bumke (Didyma), and Dr. H. Kienast (Samos). I have discussed various aspects of this research with many friends and colleagues, receiving useful information: I would like to mention in particular Rosa Maria Albanese Procelli, Regina Attula, Vincenzo Bellelli, Gebhard Bieg, Massimo Botto, Brenda Breed, Dominique Briquel, Alan Johnston, Pierre Rouillard, Brian Shefton, Stephane Verger and J. Vives-Ferrándiz Sánchez. Finally, Phil Perkins improved my English and made some helpful comments on an earlier draft of the paper. Pallottino 1963. MacIntosh Turfa 1977 (369-70 for a finds list). Morel 1981; Thuillier 1985. Hase 1989, 1993. See for instance the in-depth, documented analysis of Gras 1985 and the synthesis of Gras 1997, 48-55, where Naukratis and Cyrene are quoted in relation with the Greek expansion. See Appendix 1, no. 28. For the type: de Marinis 1999, 542-7. On the founding of Cyrene: Parisi Presicce 2003. Boardman and Hayes 1966, 1973. Humphrey 1976. Vickers and Gill 1986; Gill 2004. Excavations have shown that tumuli were used in North Africa for a long time (Camps 1961, 65-91). It would be interesting to collect the few scientifically explored grave mounds and compare them with the monuments of Greece and Etruria. The list of the tumuli in Cyrenaica includes Rowe 1956, 6-7, fig. 1 (five stone tumuli containing wooden chamber tombs, dating to the 6th century bc; Stucchi 1964, 127-31 (tumulus near Messa); Stucchi 1975, 12-13; Bacchielli 1985, 10-12, fig. 1.4 (tumulus in the agora in Cyrene, probably erected in honour of Battos, the mythical founder of Cyrene). The role of Phoenicians and Carthaginians is analyzed by many authors in Pisano 1999. The literary tradition is collected and commented on by Hans 1983. The bucchero vases from Carthage, studied by Hase (1989, 1993), are listed infra (Appendix 1, no. 3). The distribution maps are Hase 1992,

14 15 16 17 18 19 20 21 22 23

24 25

26 27 28 29 30

fig. 1 (general map of bucchero finds) and 27 (kantharoi 3e). Further finds of kantharoi: Megalopolis: Tripoli, Archaeological Museum, unpublished. Paros: Buschor 1929, fig. 8. Daskyleion: Izmir, Archaeological Museum, unpublished; Miletos: Pfisterer-Haas 1999, 265, 267; Naso 2001, 175-6, fig. 4. Datça: Berges and Tuna 2000, 198201, fig. 15b; Berges and Tuna 2001, 162, fig. 13 (three kantharoi are actually identified). Didyma: sherd belonging to the handle of a giant kantharos, unpublished (from Taxiarchis, Inv. no. Ke 01-264). See Appendix 1, no. 20. See Appendix 1, no. 21. For the identification of the sherd thanks are due to Brenda Breed (MFA, Boston). Hirschland Ramage 1970, 17 note 45 for the leaf-silver: Naso 2005a, for the leaf-gold. Burkhardt 1991, 114-15. Unpublished (Inv. no. So 91-208), but already mentioned in Naso 2001, 175. Boulomié 1982, 14-16, pl. 2. A selection of the bucchero vases from Miletos and the results of the clay analysis will be published by G. Trojsi and myself in AA. They are listed infra (Appendix 1, no. 4). They are listed infra (Appendix 1, no. 4). Cristofani 1985. The so-called ZitA amphores, which R. Docter considered to be also of Central Italic origins (Docter 1998), are now definitively attributed solely to Sardinian production (Oggiano 2000, 241-2). Naso 2001, 180, fig. 9. Concerning Malcus, depicted in the literacy tradition as a general from Carthage, an interesting opinion has recently been expressed by M. Gras, who identifies him with the leader of the Carthaginian fleet, who fought with the Etruscan ships against the Greeks in the Sardinian Sea battle (Gras 2000b, 38-9). This hypothesis, although very stimulating, is far from certain. Therefore, I still prefer to follow Bondì 2000, 63-5, Fantar 2000, 77-8 and Krings 2000: according to them, the more probable enemies of Malcus were the Phoenician cities in Sardinia or, less convincingly, the Sardinian people. Hdt. 1.166.1-2. On this battle see now Bernardini et al. 2000 (comprising many articles devoted to the various people involved in the battle); Bernardini 2001. Arist. Pol., 3.5.10-11. MacIntosh Turfa 1977 (the chronology that puts the early treaty between Carthage and Rome in 580 bc is now completely obsolete). Polyb. 3.22.4-13. Ampolo 1987, 80-4; Scardigli 1991, 47-87. See Appendix 1, no. 8. The identification as tesserae hospitales is due to Messineo 1983. The animal of the tablet from Carthage is usually considered to be a wild boar; the inscription is ET, Af 3.1. For the tablet from Rome see Torelli 2000, 554, no. 38 (with previous bibliography; more has been written since); the inscription is ET, La 2.3. For further tesserae hospitales from Murlo see Maggiani (forthcoming). Naukratis: Greek Diversity in Egypt | 197

Naso 31 Krings 1998, 2. 32 Berges 1997, 52. 33 The literature on Caere in this period is quoted in Colonna 2000, which concerns the relevant finds from Pyrgi. Further data about the new excavations in the urban area of Caere are presented in Cristofani 2003. 34 It has been found in the excavation of the treasury of Thoutmosis I, carried out by the Institut Français d’Archéologie Orientale (IFAO) under the direction of H. Jacquet-Gordon: Rouillard 1985, 24. A Klazomenian amphora from Karnak has been published by Boardman 1958. 35 The inscribed kantharoi from Perachora and Ialisos are discussed by Hase 1997, 317-18, fig. 24. The bucchero vases with Greek inscriptions from Sicily (Selinous and Girgenti) are gathered in Gras 1985, 498; the new finds from Leontinoi are illustrated by Rizza 2003, 546-8, figs 7-8, pl. 6. 36 Only Ionian bucchero is known from the site (Schaus 1985a, 73-6 nos 446-68). 37 Warden 1990, 8-9, no. 17, pl. 5. 38 The ancient literacy tradition on the Etruscan artefacts is collected and discussed by Mansuelli 1984. 39 The ‘classic’ work on infundibula is Zuffa 1960, added to by Colonna 1980, 45-6, who proposed to localize the workshop in Volsinii; M. Schindler has furnished a list (Schindler 1998, 275-6) and a distribution map (Schindler 1998, 81, fig. 20) that are not always congruous; further bibliography is listed in Naso (forthcoming). Camporeale 2003, 164 quotes the infundibula among the products of craftsmen of Volsinii. I give here a comprehensive list (Appendix 2). I hope to devote a proper study to these objects soon, to support my hypothesis. 40 Carthage: Mackensen 1999, 540-1, no. 35, fig. 1.1, pl. 44.1. Didyma: Bumke and Röver 2002, 97-9, fig. 20. About Etruscan ladles: Donati 1998, 163-6; Jurgeit 1999, 439-47, nos 740-56; Naso 2003a, 105-6, no. 159-61. Recently two Etruscan ladles from Macedonia and Nemea have been published (Blackman 2001/2, 21 fig. 38). 41 The localization in Vulci of the infundibula workshop has already been hypothesized (Martelli 1988, 23-5; Paul 1988). Even if only one infundibulum was presumably found in that centre (Appendix 2, no. 30) and one in Castro, a minor centre of the Vulcian territory (Appendix 2, no. 9), one can presume the provenance from Vulci of many tools of unknown provenance (Appendix 2, nos 27-47), because the necropoleis around this city have been often disturbed by illegal excavations. 42 On the Vulcian bronzes the old article of Neugebauer 1943 is still useful, although the most comprehensive and recent study is now Riis 1998; M. Martelli provides a comprehensive study of the Vulcian stone workshops in the 6th century bc (Martelli 1988, 2001, 2004, forthcoming). For the bronze Schnabelkannen see now Vorlauf 1997. 43 See Appendix 1, no. 13. A similar cuirass is preserved in Naples, Archaeological National Museum, Inv. no. 5735: Acquaro and Ferrari 2004, 114-15, no. 154. The Italic bronze belts have been collected by Romito 1995. 44 See Appendix 1, no. 9. On these cippi see Blumhofer 1993 and my review (Naso 1994). 45 See Appendix 1, no. 10. For the distribution of Genucilia plates in central Italy: Naso 1996, 175, note 265; Poulsen 2002. Dr. L. Vuono (Rome/Mannheim) is publishing her thesis including new data about the Genucilia plates from the Palatine Hill. 46 See Appendix 1, no. 27. Concerning Etruscan linen books: Roncalli 1980a. 47 Colonna 1988, 16, note 8. 48 See Appendix 1. no. 11. The inscriptions are ET, Af 3.2. 49 The following bibliography has been accessible to me: AfrIt (1, 19278, 1941); LibAnt (1, 1964-16, 1979, n.s. 1, 1995- 4, 1998); LibSt (1, 197033, 2002); Monografie di archeologia libica (1-19); QAL (1,1950-17, 2002); Karthago I-III. Die deutschen Ausgrabungen in Karthago, edited by F. Rakob, 1991-1997. Mainz am Rhein: Philipp von Zabern. 50 For the Italic bronze discs: Tomedi 2000, reviewed by Naso 2003b. 51 These objects, as parts of Etruscan wine sets, are concentrated in Etruria, but they are occasionally also found in southern Italy (Jurgeit 1999, 462, nos 778-9). 52 These vases are not a certain import from Italy, since a production similar to Gnathia has been suggested for Alexandria (Piekarski 2001a, 107-8; Alexandropoulou 2002, 196-7).

198 | Naukratis: Greek Diversity in Egypt

53 In my opinion the kotyle from the necropolis of Leptis Magna published as Etrusco-Corinthian (De Miro and Fiorentini 1977, 31, fig. 38) is not Etruscan, as J.P. Morel correctly hypothesized (Morel 1981, 484-5, note 100). 54 The presence of Etruscan bucchero in Naukratis is stressed by many authors (Boardman 1958, 12 note 47; Morel 1981, 468 note 15; Boardman 1999a, 124; Möller 2000a, 144; Kerschner 2001, 75). 55 The collection v. Bissing was divided among the museums of Amsterdam, Berlin, Bonn, Den Haag and Munich (Kerschner 2001, 72). The sherds in the museums of Berlin and Munich were destroyed during the Second World War; in Bonn there is no Etruscan bucchero (Piekarski 2001a). 56 Johnston 1982, nos 3-5, Inv. no. BM GR 1888.6-1.643b-d (d refers to two sherds), are four sherds not all joining, but belonging to the same vase, with a form similar to an Attic black-figure olpe. Similar forms are not represented among the Etruscan bucchero repertoire; the clay, quite fine, with mica and a slim slip, is also untypical for Etruscan bucchero. 57 Dr. Phil Perkins kindly informed me that in the British Museum’s storerooms also the following bucchero sherds from Naukratis are preserved: 22 sherds of an oinochoe Rasmussen 3a or 3d (GR 1924.121.76 b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,v,w); one sherd incised with a palmette (GR 1924.12-1.75); two unidentified sherds (GR 1924.12-1.77 and 78); one oinochoe, nearly complete but restored, of type Rasmussen 3a, the provenance of which is, however, not certain (GR 1977.10-11.89); one unidentified sherd possibly of bucchero (GR 1888.6-1.633). This material will be published by Perkins (forthcoming). 58 For this type of vase, probably from Volsinii (Colonna 1985, 45, note 2): Höckmann 1982, 100, nos 57, 66-7, pl. 55.1-5 (without horse heads), dated to 550–500 bc. A further Etruscan podanipter with lion feet has been found in a grave group in Vaste dating to 430 bc (Semeraro 1990, 89-90 no. 102; Tarditi 1996, 39 no. 53). 59 In the Cairo Museum also fakes of Etrusco-Italic antiquities are preserved, such as a bronze Herakles (Edgar 1904, 71, no. 27918, pl. I), belonging to a group of similar statuettes (Franzoni 1966, 50-1, fig. 11; Naso 2003a, 280, no. 531, pl. 106). One should also mention a female head-oinochoe from el Kantara (Bissing 1903, 146 fig. 3f; Edgar 1904, 29, no. 27743, pl. VII) close to (but not Etruscan: the eyes have been made of glasspaste!) a group of Etruscan female headvases of the Hellenistic period (Menzel 1959; Haynes 1959; Naso 2003a, 77-9, nos 119-21). 60 This subject is very popular (Naso 2003a, 122, note 172). 61 No. 332 had usually also been attributed to an infundibulum, but it probaly belonged to a torch-holder (like Hostetter 2001, 142-3, no. 353-5, pl. 63-4, from Spina). 62 ‘Manico fuso di bronzo appartenente a un simpulum. Nella parte piana, sopra ad un ornamento traforato, incastra in una cerniera girante un piccolo leone a tutto rilievo, le cui zampe anteriori erano inchiodate nella lamina che serviva da coperchio’. 63 ‘Manico elegante di colum. Si può distinguere in due porzioni, l’una vicina all’alto del vaso, è in forma di …, che parte, sulla linea mediana della quale un leoncino si accovaccia, l’altra semplice, ricurvandosi in basso, termina in una testa d’oca. Lungh. mm 220’. This is probably the same tool that A. Furtwängler saw in the Villa Giulia Museum (Furtwängler 1890, 196, ad no. 1267, 1267a: grave 38) and that could not be found by H. Sauer (1937, 296), F. Magi (1941, 230) and M. Zuffa (1960, 181 note 37). 64 Furtwängler 1890, 147, no. 924a, is a small fragment of a handle, perhaps belonging to a non-Etruscan infundibulum. Brown’s (1960, 111, note 2) careful attribution of Furtwängler 1890, 152 no. 966, pl. 57, to an infundibulum is very probably wrong, because the plomb rests under the paws and the hinge under the hindlegs are never documented among infundibula. 65 Xàbia is the Catalan name of Jávea. Thanks to Vincenzo Bellelli I have heard from Dr. J. Vives-Ferrándiz Sánchez (Valencia) of the existence of this infundibulum, which will be published in a German periodical. 66 Contrary to the brief published description (Fol 1874, 252, note 1169 from Vulci), another bronze sherd preserved in Geneva (Musée d’Art et d’Histoire, Inv. no. MF. 1169) does not belong to an infundibulum. 67 B.B. Shefton classified as ‘Hallstatt imitation’ the tool from Novi Pazar (Shefton 1970, 55-6).

Identity in the Making: Greeks in the Eastern Mediterranean during the Iron Age Alexander Fantalkin Abstract Although Greek contacts with the Southern Levant during the Iron Age have been studied at length, the matter remains controversial in many aspects. The present study provides an overview of EastWest contacts during the first half of the 1st millennium bc, suggesting to divide it into five major periods of contact. These periods, involving a different chronological setting, are characterized by different ‘total contexts’, heavily shaped by geopolitical dynamics. It is suggested that every period of contacts (or their absence) requires a different explanation. Introduction For scholars interested in Greek contacts with the Southern Levant during the Iron Age two developments in the late 7th century bc are truly remarkable: the establishment of Naukratis in Egypt and the massive appearance of East Greek pottery on the eastern shores of the Mediterranean. It is not surprising therefore that these themes were chosen, inter alia, for the 28th British Museum Classical Colloquium.1 However, any attempt at discerning and decoding patterns in the dispersion of East Greek pottery in the Levant, as well as explaining the Naukratis phenomenon, requires an understanding of East–West contacts during the first half of the 1st millennium bc. Such an overview is undertaken here. However, since I could not hope in the present format to do justice to the whole range of issues that preoccupy scholars dealing with Greeks in the East, I offer instead an extremely brief synopsis of Greeks in the East during the Iron Age, with special emphasis on a few thorny issues. Since I shall concentrate on a number of broad historical/archaeological issues, it is perhaps prudent to acknowledge that every generation writes its own history and that every scholar has a view of the past coloured by his/her education, experience and environment. I have no pretensions therefore that my interpretations of East–West contacts will be taken as the only possible scenario. On the other hand, I hope that among the pool of potential explanations for the changing nature of East-West contacts, the model I offer best accounts for the available evidence.2 From an epistemological point of view, I am on the side of many who argue that among the three main poles – realism, positivism and idealism3 – it is usually realism that offers the most useful point of departure for any archaeological reconstruction, especially when this realism is combined with a healthy dose of scepticism and a pinch of imagination.4 And although I can accept, at least to a certain extent, that in too many cases ‘there are no facts, only interpretations’, archaeology does often supply facts. Some facts, such as the presence or absence of Greek pottery on the eastern shores of the Mediterranean, matter a great deal. The question remains: what we are going to do with these facts? But before I embark on the

‘pots and people’ question, I would like to emphasize the significance of the historical/chronological context – the backbone of any historical interpretation. The accumulation of data, an essential beginning, should lead to contextualization involving the understanding that different chronological settings may represent different geopolitical dynamics. Ian Morris rightly observes that one of the major shortcomings of the post-modern trend of emphasising connectivity and mobility is its timelessness.5 He points out that many of what he calls first wave studies ‘showing links between Greek and Near Eastern cultures, often threw together evidence scattered across centuries, disregarding traditional chronologies’.6 The recent contribution of Horden and Purcell takes this approach even further,7 arguing ‘against interpretations that emphasize radical change and violent discontinuity in the Mediterranean past’.8 What is offered instead is a vision of a permanently integrated Mediterranean, wherein change is constant and ubiquitous, but generally local in its effects. Such a reconstruction, with its emphasis on microregions, leaves little room for pivotal turning points in Mediterranean history, since the assumed connectivity stretches across extremes of time, by-passing geo-political boundaries and empires, together with symbolically expressed ideologies of economic exchange and political domination.9 With mobility as the norm and a permanent feature of human activity around the Mediterranean shores, we are forced to ask questions differently. Or, as Emma Blake recently put it, ‘rather than ask, why did people move, one may ask, why did people stay put in some cases?’10 Heavily affected by current globalization,11 Horden and Purcell’s vision of the Mediterranean is already considered by some, and not without reason, as ‘one of those manifest watersheds in the study of antiquity’, which will take a generation of historians to digest.12 Indeed, taking into consideration a number of earlier studies in favour of a permanently connected Mediterranean, one is tempted to suppose that we are witnessing a paradigm shift.13 What is missing in the portrait of a permanently connected Mediterranean, however, is the notion of historical/ chronological context. In this regard, Bakhtin’s concept of the total context of an utterance provides an applicable insight. The total context relates to the ways in which voices circulate in both spoken and written dialogues and, according to Bakhtin, is unrepeatable.14 Even if one repeats the words employed in the same order, the total context would be always different, if for no other reason than because the words have already been uttered once.15 And when Horden and Purcell insert the distribution of Late Bronze Age ox-hide ingots into the model of a permanently connected Mediterranean, for instance, comparing it simplistically with the whole spectrum of later metallurgical distributive systems,16 the ‘total unrepeatable context’ of Naukratis: Greek Diversity in Egypt | 199

Fantalkin particular periods is lost. The problem is not one of comparing some chronologically distant metallurgical distributive systems. After all, the merits of the comparative approach are undeniable.17 Likewise, analogies are appropriate tools and salient features of any historical/archaeological investigation. The problem is a deliberate unwillingness to recognize that the distribution of Late Bronze Age ox-hide ingots should be understood on its own terms and against the background of Late Bronze Age geo-political dynamics,18 which are a world apart from the distributive systems of the Greeks and Romans, let alone those of medieval Genoa. Or, as Mario Liverani observes, ‘the “Bronze Age”, invented as a classificatory device for tools and weapons, can still be used as a large historical label, encompassing similarly structured socioeconomic systems and quite sharply opposed to the (differently labelled) preceding and succeeding periods’; (emphasis added – A.F.).19 Although it might be relevant, I am not concerned here with the long-running debate involving polarising tendencies ‘to see the past as Same (a primitive version of our present, which teleologically evolves into it) or as Other (as a remote, alien, fundamentally different world)’.20 My main concerns are socially embedded cultural contexts21 and their chronological settings. Therefore, with regard to metallurgical distributive systems, the only reliable conclusion that may be deduced from the analogies scattered across the centuries is, in my view, an acknowledgment that different distributive systems have existed in the Mediterranean at different times. However, in order to understand the forces driving these and other exchange activities, they must be viewed in their proper chronological/ historical contexts. It is not helpful to gather all the cases of connectedness and mobility under the same rubric of a permanently interconnected Mediterranean without distinguishing between different historical periods. Indeed, the presence or absence of Greeks in the Eastern Mediterranean during the Iron Age suggests that there is no single model that would explain these contacts (or their absence) through different time periods. Quite the opposite: judging from the facts on the ground (and there are some), every subsequent historical period requires a different explanation, a different narrative. Greek contact with the eastern Mediterranean during the Iron Age: stressing the context The area under discussion runs from the coast east of Cilicia down to the Sinai Peninsula. The contacts in question may be divided roughly into five major periods, each involving a different chronological setting. These settings are characterized by different ‘total contexts’ heavily shaped by geo-political dynamics. First period: a renewal of contact The first period is characterized by the presence of mainly Euboean pottery (but also Attic and Atticizing) found in northern Syria, Phoenicia and northern Israel in the late 10th, the 9th and the better part of the 8th centuries bc.22 The assumed Phoenician superiority in virtually everything leaves, according to many modern scholars, no room for independent Euboean ventures at such an early date, especially to the East. When even pure Cypriot ventures are labelled CyproPhoenician,23 it is quite obvious that Euboeans could not 200 | Naukratis: Greek Diversity in Egypt

compete with the advanced Phoenicians, let alone establish a trading post at Al Mina toward the end of the 9th century bc. The dominant view among Aegean specialists, although with notable exceptions, is that the Phoenicians brought Euboean pottery with them to the East.24 However, the trend during the last decades of pinpointing the beginning of Phoenician expansion to as early as the 11th/10th centuries bc,25 if not earlier, is based almost entirely on a handful of presumably historical sources: to a lesser extent on the so-called ‘Report of Wenamun’26 and to a larger extent on the biblical accounts regarding the cooperation between Kings Solomon and Hiram I.27 These sources can no longer be treated as reliable.28 Furthermore, the low Iron Age chronology, advanced in Israel nearly a decade ago,29 has enormous implications for the Aegean world. First, it leaves no room for Phoenician colonial expansion before the late 9th–early 8th centuries bc.30 The presence of imported Phoenician vases in the assemblages at Palaepaphos Skales31 should not imply the beginning of Phoenician colonisation of Cyprus before their establishment in Kition at the late 9th century bc.32 Indeed, judging from available archaeological evidence, the initial Phoenician expansion overseas, accompanied by settlements abroad, took place only in the second half of the 9th century bc; and I refer to the wellknown Phoenician establishment at Kition,33 but also to evidence from new radiocarbon dating from Carthage34 and Southern Spain.35 In my view, this expansion may be explained as a result of pressure from Hazael, the king of Aram Damascus.36 A plethora of archaeological data accumulated in Israel, such as Hazael’s inscriptions37 and possible destruction layers, mostly in northern Israel,38 but also to the south in biblical Gath,39 suggests that Hazael’s kingdom was one of the most serious players in the Southern Levant during the second half of the 9th century bc.40 I believe that Susan Frankenstein’s theory,41 that the Phoenician specialization in trade, accompanied by their settlements abroad, should not be seen entirely as free-trade activity, but rather in the context of their functioning as commercial agents for the Neo-Assyrian Empire, is basically correct. However, judging from the archaeological data regarding the beginning of Phoenician expansion overseas, this delicate arrangement, which eventually transformed the Phoenicians into pan-Mediterranean traders, started in the days of Hazael, with Phoenicians serving the trade ambitions of Aram Damascus.42 Second, and even more important, the low Syro-Palestinian chronology provides, finally, an anchor for Aegean ProtoGeometric and Geometric chronologies.43 A minimalist approach to the beginning of Phoenician affairs in the Mediterranean44 leads, in conjunction with a low chronology,45 to an emphasis on the principal role played by the Euboeans in the renewal of contact between East and West,46 culminating in the establishment of Al Mina sometime around 800 bc.47 This, of course, occurred on behalf of local rulers.48 The same pattern will be observed almost 200 years later, with the establishment of Naukratis in Egypt. In this regard, Boardman’s notion that we should consider a trading port at Al Mina as a modest precursor of Naukratis is rather attractive.49 The Greek presence in the Eastern Mediterranean at this early period seems always to be restricted and controlled by local

Identity in the Making: Greeks in the Eastern Mediterranean during the Iron Age authorities.50 Therefore, I strongly disagree with the idea that accepting a prominent Euboean role in Early Iron Age journeys to the East makes one Helleno-centrist.51 The Euboeans were conducting these journeys because they were interested in reestablishing lost contacts with the East.52 It would give to the ruler of Lefkandi, for example, an enormous advantage compared to other contemporary Greek rulers.53 For the Greek side it meant a great deal. For the East, it does not seem to mean much at all. But for the Greeks it meant the beginning of the Orientalizing movement, with a minor Phoenician contribution, but mainly, through the Syrians, as was already suggested long ago and on many occasions by John Boardman. To this, one should add the adoption of the Greek alphabet, sometime around the middle of the 8th century bc.54 All in all, although the renewal of contact may be attested during the 10th/9th centuries bc, it certainly intensified during the better part of the 8th century bc at least until the beginning of the Neo-Assyrian domination over the Southern Levant. Second period: the Neo-Assyrian domination Greek contacts with the East were halted by Assyrian expansion; here we arrive at a second period, the period of Assyrian domination. The recent understanding of the processes that took place in the Southern Levant near the end of the 8th and during the main part of the 7th centuries bc shows unprecedented involvement of the Assyrian administration in local affairs. This involvement may be seen in a variety of fields, such as the annexation of many Levantine kingdoms accompanied by the transformation of some of them into Assyrian provinces; population exchanges; re-arrangement of the borders and intensive construction activity. The latter is particularly visible in the coastal area, which is dotted with Assyrian emporia and fortresses.55 One of the most important Assyrian goals was the supervision of Phoenician trading activity. In this regard, as I have already stated, Susan Frankenstein’s theory viewing the Phoenicians as commercial agents for the Neo-Assyrian Empire seems to be basically correct.56 Concerning the Eastern Mediterranean, it is quite clear that every aspect of Phoenician commerce was closely overseen and taxed by Assyrian officials. What we are witnessing here is a delicate balancing act. On the one hand, the Phoenicians enjoyed the stability produced by the pax Assyriaca and the exclusive access to the network of trade-routes and trade-centres across the Eastern Mediterranean. On the other hand, their commerce was strictly regulated and taxed.57 The Phoenicians involved in commercial and colonial activities in the Western Mediterranean, far from their Assyrian masters, doubtless enjoyed a higher degree of flexibility than their counterparts in the Eastern Mediterranean. From the point of view of the present colloquium, however, the most important conclusion is that, with regard to the southern Levant, this new world-order left most of the mainland Greeks quite effectively out of the game. The single limited point of contact that was left was again Al Mina, which became a port of trade toward the end of the 8th and during the 7th centuries bc. But after c. 700 bc, Euboean imports to the Southern Levant almost disappear. Starting from Al Mina’s Level 6, it is mainly East Greek pottery that shows up during the period of Assyrian domination, not Euboean. Besides it is not yet entirely clear who was responsible for carrying this

pottery to Al Mina. Did it arrive directly from Eastern Greece or was the Cypriot connection involved? What appears to be quite clear, however, is that mainland Greece seems to be without direct connections with the East, starting from the period of the Neo-Assyrian domination. In fact, excluding Al Mina, while even at this site there is a clear structural break between Levels 7 and 6, Greek pottery (except for a few insignificant cases) is almost non-existent in the Neo-Assyrian contexts.58 This contrasts with a much broader distribution prior to the Neo-Assyrian domination and, especially, immediately after its collapse. Lanfranchi’s recent speculations regarding Greek contact with the Neo-Assyrian Empire,59 which are based, archaeologically, almost exclusively on Haider’s earlier study,60 will find no echo in the archaeological realities of the Southern Levant. Dependent as they are on mistaken representations and understandings of the archaeological data involved,61 Lanfranchi’s historical implications, according to which Assyrians favoured Greeks over Phoenicians in commercial and settlement activities in the southern Levant,62 can confidently be rejected. Similar confusion regarding the Greek pottery in the Southern Levant appears in Rollinger’s recent attempt to draw a picture of Greek contacts with the East during Neo-Assyrian period.63 Likewise, his suggestion that we consider the individuals mentioned in the Near-Eastern texts as Iaman + suffixes other than ¯aya as possible Greeks acting in the midst of the Neo-Assyrian Empire, seems to reside on rather shaky ground. Both archaeological and historical data suggest that during the Neo-Assyrian regime the Greeks occupied a marginal space in the Mesopotamian understanding of the universe. Bearing in mind the Neo-Assyrian imperial ideology, with its pretensions of ruling a universal domain,64 such a role for Greeks is understandable. Located in the ‘midst of the sea’,65 where the Neo-Assyrian regime was not able to insert them physically into the ‘correct relationship’ with the imperial new-world order, Greeks were reduced to the status of ‘disparate, remote people living on the edge of the world’66 in the Neo-Assyrian mappa mundi. The Phoenicians apparently were chosen to serve as commercial agents for the Neo-Assyrian empire not because they were natural-born traders,67 although their expertise should not be underestimated, but because the Neo-Assyrian regime was able to control their trade, which was not without benefits for both sides. Given this state of affairs, I tend to agree with Helm’s suggestion that, for the Greek side, ‘the imperial obligations imposed on permanent residents in Assyrian provinces made life in the Levant unattractive’.68 Indeed, as Helm pointed out more than 25 years ago: Even in the few nominally independent port cities such as Arvad, Tyre, Ashkelon and Gaza it is likely that Greek traders would have encountered Assyrian administrators, commercial regulations and economic institutions. It was doubtless these contacts, and the contacts with other representatives of Assyrian provincial government, which gave visiting Greeks the not inaccurate impression that the entire east Mediterranean coast comprised ’Assuri&h.69

The unprecedented involvement of the Neo-Assyrian administration in the local affairs of the Southern Levant (see above), attested both historically and archaeologically, is certainly in accord with Helm’s suggestions. In this regard, Amélie Kuhrt’s rather sceptical look at the evidence for direct Naukratis: Greek Diversity in Egypt | 201

Fantalkin contact between Greece and the Mesopotamian empires is particularly revealing.70 Although, as in the earlier periods, the Greeks definitely continued to meet Easterners, this time these were mostly Phoenician competitors. And these are indeed the Homeric Phoenicians.71 The nature of direct contact between the Greeks and the Near East during the second period in my provisional scheme suggests therefore the beginning of a ‘Great Divide’ rather than Burkert’s Orientalizing revolution.72 It should be explicitly stated, however, that the concept of a Great Divide does not imply an immediate break in contacts. It is better described as a gradual process, starting with Tiglathpileser III’s annexation of the kingdom of Unqi/Patina in 738/737 bc. If Zadok’s identification of Al Mina as A∆tâ in Tiglath-pileser’s inscription on the Iran stele is correct,73 this might indicate that right after the annexation of Unqi, an Assyrian emporium was installed at Al Mina,74 in order to regulate and incorporate the existing Greek enclave into the sphere of the Neo-Assyrian realm. Already at that time, a letter from Calah (Nimrud)(ND 2370), sent most probably to Tiglathpileser III by Qurdi-Aššur-la ¯mur, points to a possible Ionian raid on the Phoenician coast.75 To this one may add a reference to the town of Yauna, mentioned in a Neo-Assyrian letter (ND 2737) published a few years ago by Saggs.76 The letter contains no firmly dateable details. However, the themes discussed and the arenas of operation seem to be echoed in the letters of QurdiAššur-la ¯mur, who was probably the governor of S.imirra in the time of Tiglath-pileser III.77 In this regard, Na’aman’s suggestion that we identify the town of Yauna with Ras el-Bassit,78 would, if accepted, point to a possible Greek presence at this site at that time. Hereafter, however, the handful of Neo-Assyrian sources that mention Ionians, mostly in hostile contexts,79 when combined with an almost total lack of Greek pottery in the NeoAssyrian assemblages (see above), leave little doubt about an intensification of the Great Divide. Third period: stressing the significance of the late 7th-century BC contact, during a brief period of Egyptian domination The next period, although chronologically brief, is the most important for the purposes of the present colloquium. I refer to some 20–25 years of Egyptian rule in the Southern Levant, following the Assyrian withdrawal. When the Assyrians pulled out from the Levant sometime in the twenties of the 7th century bc,80 the Egyptians took over their territories and ruled until the Babylonian invasion. This period, the third in my provisional schema of the Greek presence in the Levant, lasted until the Babylonian destructions at the end of the 7th and in the early 6th centuries bc. The sudden and massive appearance of East Greek pottery on the coastal plain of Israel toward the end of the 7th century bc 81 and its subsequent disappearance after only a few years fit the time-span during which the area fell under Egyptian rule.82 Following Nadav Na’aman’s insightful observations, I have elsewhere discussed at length the East Greek pottery assemblages found in places such as Ashkelon, and the fortresses of Mez.ad H.ashavyahu and Kabri, arguing that these represent Greek mercenaries in the employ of the Egyptians.83 In this reconstruction, the placement of these garrisons along the coast together with the employment of Kittim along the southern fringe of the kingdom of Judah, conformed to two 202 | Naukratis: Greek Diversity in Egypt

Egyptian goals: first, to protect the coastal plain – the main route to the North; and second, to protect the Arabian trade networks, which the Egyptians inherited from the Assyrians.84 The modest finds of East Greek pottery in the vicinity of major military bases85 probably reflect Greek mercenary activities in these areas rather than pottery trade. Many scholars, however, have claimed that the abundance of East Greek pottery should be taken as evidence of East Greek trade.86 In these reconstructions even the coarse East Greek cooking pots are considered a tradable commodity to the East.87 In my view, most of these reconstructions are untenable. The attested distribution and the nature of East Greek finds in the region of Palestine are insufficient to prove either the existence of a developed pottery trade88 or the existence of a directional exchange of other goods that may be less visible in the archaeological record.89 An additional point that argues in favour of East Greek mercenary garrisons rather than trading emporia is the restriction of East Greek trade to Naukratis in Egypt.90 It must be remembered that the establishment of Naukratis toward the end of the 7th century bc overlaps with the appearance of East Greek pottery on the Israeli coast. There is hardly any doubt that the entire coastal plain up to Phoenicia should be considered Egyptian domain.91 In these circumstances it is reasonable to assume that Egyptians would not have allowed the uncontrolled establishment of East Greek emporia on the Southern Levantine coast, just as they did not allow it in Egypt itself. While Phoenicia proper and the areas to the north might have enjoyed East Greek trade during the Egyptian interlude,92 the evidence collected so far from the southern part of the Eastern Mediterranean points mainly to East Greek mercenary activity.93 The sudden appearance of Greek mercenaries in the East and their employment by the different Near Eastern Powers continues to be a subject of debate.94 In my opinion, both historical and archaeological evidence suggests that the presence of Greek mercenaries in the region should be explained as an organized movement orchestrated by a central Egyptian authority. These Greeks were not individual mercenary adventurers but were formally garrisoned.95 I cannot accept the ideas expressed by several scholars that East Greek assemblages point to individual adventurers or small groups of Greek mercenaries96 pursuing Homeric honour and glory.97 I dealt with this issue in detail a few years ago,98 and I intend to expand the discussion elsewhere. Likewise, today I am even more convinced that attempts to attribute the employment of Greek mercenaries to Egyptian vassals, be it the kingdom of Judah or the kingdom of Tyre, should be abandoned. Most recently, however, Wenning99 defended his date for the establishment of Mez.ad H.ashavyahu between 600 and 598 bc, under the reign of King Jehoiakim.100 This is in contrast to Na’aman’s suggestion that the fortress of Mez.ad H.ashavyahu was abandoned in 604 bc, the year in which Nebuchadnezzar II launched a campaign to the Philistine Coast and destroyed Ashkelon.101 In my opinion, however, Na’aman’s scenario remains the most plausible option. Moreover, I hope I was able to demonstrate that since the abandonment pattern attested at Mez.ad H.ashavyahu points to a ‘planned abandonment without anticipated return’,102 it fits nicely with the assumption that this Egyptian fortress was intentionally abandoned in face of the approaching Babylonian army.103

Identity in the Making: Greeks in the Eastern Mediterranean during the Iron Age The historical improbability of Wenning’s scenario, on the other hand, which attributes the employment of Greek mercenaries to Jehoiakim, who was an Egyptian vassal, has already been demonstrated104 and there is no need to revisit it here. Likewise, from a strictly archaeological point of view, Wenning’s entire case rests on the presence of a single pottery sherd he attributes to the North Ionian Late Wild Goat style. Even if we assume that the sherd has been identified correctly, Wenning’s belief that it cannot be earlier than 600 bc is untenable. The East Greek pottery chronology for this period, with its approximate dates, rests on synchronisms with Palestinian destruction levels and on synchronisms with Corinthian and Attic pottery.105 It is simply impossible to assume such precision (+/– 4 years, which is the difference between Wenning and myself!) in dating this North Ionian East Greek sherd. In terms of absolute chronology, both the East Greek pottery and the local pottery from Mez.ad H.ashavyahu may be placed either in the late 7th