Late Iron Age Gold Hoards from the Low Countries and the Caesarian Conquest of Northern Gaul, AAS 18 / Atuatuca 3, Publications of the Gallo-Roman Museum of Tongeren, Amsterdam, 2012. COMPLETE EDITION

GRM Gallo-Romeins Museum Tongeren (BE); Simone Scheers

COMPLETE EDITION Roymans N., Creemers G. & Scheers S. (Eds), Late Iron Age Gold Hoards from the Low Countries and the Caesarian Conquest of Northern Gaul, Amsterdam Archaeological Studies 18 / Atuatuca 3, Publications of the Gallo-Roman Museum of Tongeren, Amsterdam, 2012. This volume presents eight new Iron Age gold hoards from the southern Netherlands and Belgium, consisting of gold coinages and in several cases also gold ornaments. The study of these gold hoards provides a wealth of new information on the archaeological contexts in which they were found, on the dating of many coin types and jewellery, and on the social role of gold in pre-Roman society. All these hoards seem to have been buried in the 50s BC, thus making a direct association with the historical context of Caesar's war campaigns in northern Gaul very plausible. This makes the volume important for archaeologists as well as numismatists and historians.

N. Roymans, G. Creemers and S. Scheers to in u bo A m s t e r d a m Late Iron Age Gold Hoards from the Low Countries and the Caesarian Conquest of Northern Gaul U n i v e r s i t y P r e s s Late Iron Age Gold Hoards from the Low Countries and the Caesarian Conquest of Northern Gaul A msterdam A A tuatuca hi rchaeological S t u d i e s 18 Editorial Board: Prof. dr. E.M. Moormann Prof. dr. P.A.J. Attema Prof. dr. N. Roymans Prof. dr. E Theuws Other tides in the AAS series: 1. N. Roymans (ed.): From the Sword to the Plough. Three Studies on 13. T. Derks & N. Roymans: Ethnic Constructs in Antiquity. T he role the Earliest Romanisation o f Northern Gaul o f power and tradition Open Access edition: http://dare.uva.nl/record/19675 Open Access edition: http://dare.uva.nl/aup/en/record/301890 2. T. Derks: Gods, Temples and R itual Practices. The Transformation o f Religious Ideas and Values in Rom an Gaul Open Access edition: http://dare.uva.nl/aup/en/record/172370 14. T. D. Stek: C ult places and cultural change in Republican Italy. A contextual approach to religious aspects o f rural society after the Rom an conquest ISBN 978 90 8964 177 9 3. A.Verhoeven: Middeleeuws gebruiksaardewerk in Nederland (8e — 13e eeuw) Open Access edition: http://dare.uva.nl/aup/en/record/172373 4. F. Theuws / N. Roymans (eds): Land and Ancestors. Cultural Dynamics in the Urrfield Period and the Middle Ages in the Southern Netherlands Open Access edition: http://dare.uva.nl/aup/en/record/172372 15. P. A.J. Attema, G.-J. L.M. Burgers & P. M. van Leusen: Regional Pathways to Complexity. Settlement and land-use dynamics in early italy from the bronze age to the republican period ISBN 978 90 8964 276 9 16. E.M. Moormann: D ivine Interiors. M ural paintings in Greek and Rom an sanctuaries ISBN 978 90 8964 261 5 5. J. Bazelmans: B y Weapons made Worthy. Lords, Retainers and Their Relationship in Beowulf Open Access edition: http://dare.uva.nl/aup/en/record/172337 17. N. Roymans / T. Derks: Villa Landscapes in the Rom an North. Economy, Culture and Lifestyles ISBN 978 90 8964 348 3 6. R . Corbey / W. Roebroeks (eds): Studying H um an Origins. Disci plinary History and Epistemology Open Access edition: http://dare.uva.nl/aup/en/record/172272 7. M. Diepeveen-Jansen: People, Ideas and Goods. N ew Perspectives on ‘Celtic barbarians’ in Western and Central Europe (5 0 0 -2 5 0 B C ) Open Access edition: http://dare.uva.nl/aup/en/record/172273 8. G.J. van Wijngaarden: Use and Appreciation o f Mycenean Pottery in the Levant, Cyprus and Italy (ca. 1 6 0 0 -1 2 0 0 B C ).T h e Significance o f Context Open Access edition: http://dare.uva.nl/aup/en/record/172274 9. F.A. Gerritsen: Local Identities. Landscape and community in the late prehistoric Meuse-Demer- Scheldt region Open Access edition: http://dare.uva.nl/aup/en/record/172820 10. N. Roymans: Ethnic Identity and Imperial Power. The Batavians in the Early Roman Empire Open Access edition: http://dare.uva.nl/aup/en/record/172930 11. J.A.W. Nicolay: Arm ed Batavians. Use and significance o f weaponry and horse gearfrom non-military contexts in the R hine delta (50 be to ad 450) Open Access edition: h ttp ://dare.uva.nl/aup/nl/record/397232 12. M. Groot: Animals in ritual and economy in a Rom an frontier com munity. Excavations in Tiel-Passewaaij Open Access edition: http://dare.uva.nl/aup/en/record/301888 Other titles in the Atuatuca series: Creemers, G. /B. Demarsin/ P. Cosyns (eds), 2006: Rom an Glass in Germania Inferior. Interregional Comparisons and recent Results. Proceedings o f the International Conference held in the Gallo-Roman M useum o f Tongeren (May 15th 2 0 0 5 ), Tongeren (Atuatuca 1). Demarsin B./ M. Otte, 2006: Neanderthals in Europe. Proceedings o f the International Conference, held in the Gallo-Roman M useum in Tongeren (September 17 - 19th 200 4 ), Liege/Tongeren (ERAUL 117 /Atuatuca 2). Late Iron Age Gold Hoards from the Low Countries and the Caesarian Conquest of Northern Gaul E D I T O R S Nico R o y m a n s , G A G u i d o m ste r d a m a l l o - R o m a n C & S reem ers U n iv er sity M u seu m T P i m o n e ress o n g er en S cheers This book meets the requirements of ISO 9706: 1994, Information and documentation - Paper iso 9706 for documents —Requirements for permanence. This volume appears in a collaboration between Amsterdam University Press and the Gallo-Roman Museum Tongeren. Cover illustration: Eburonean gold staters from the Amby hoard Cover design: Kok Korpershoek, Amsterdam Lay-out: Bert Brouwenstijn, ACVU Amsterdam ISBN ISBN e-ISBN NUR 978 90 8964 349 0 (The Netherlands) 978 90 7460 550 2 (Belgium) 978 90 4851 484 7 682 ©Amsterdam University Press, Amsterdam/ Gallo-Roman Museum Tongeren, 2012 All rights reserved. Without limiting the rights under copyright reserved above, no part o f this book may e reproduced, stored in or introduced into a retrieval system, or transmitted, in any form or by any means (electronic mechanical, photocopying, recording, or otherwise), without the written permission o f both the copyright owner and the editors o f this book. CONTENTS Preface Eight gold hoards from the Low Countries. A synthesis Nico Roymans / Simone Scheers v 11 1 The gold hoard of Fraire Simone Scheers / Guido Creemers 47 Three gold hoards from Thuin Simone Scheers / Guido Creemers / Nico Roymans / Luc Van Impe 71 A coin hoard with an animal-headed gold bracelet from the vicinity of Philippeville Nico Roymans / Simone Scheers 109 Remains of a disturbed gold hoard at Orp-le-Grand? Luc Van Impe / Simone Scheers 117 The gold hoard of Heers Simone Scheers / Guido Creemers 125 The gold and silver hoard of Maastricht-Amby Nico Roymans / Wim Dijkman 171 Analytical investigation of the Late Iron Age gold hoards from the Low Countries Helena Wouters 215 List of contributors 239 PREFACE The plan to publish this book was first mooted in 2008 when the newly excavated Late Iron Age coin hoard of Maastricht-Amby was presented at a press conference. Two of the present editors, Guido Creemers and Nico Roymans, felt it would be a good idea to bring together the gold hoards of Amby and Heers in a publication in English. After all, the hoards come from the same region, there is a consid erable overlap in terms of their composition, and they date from the same period. A short time later, we decided to also include the unpublished, or provisionally published, gold hoards from Fraire,Thuin and Orp-le-Grand in Belgium. Simone Scheers, who preserved the major part of the documentation, imme diately offered her cooperation. Then Luc Van Impe was invited to appear as co-author. In December 2010, when we learned of a new hoard find near Philippeville in Belgium, we decided to include that as well. In this way we have ended up with an ambitious project to bring together about eight Late Iron Age gold hoards from the Low Countries. The publication of these hoards is important for archaeolo gists numismatists as well as historians. Because the hoard finds are clearly almost contemporaneous, an integrated analysis will make a substantial contribution to the study of Late Iron Age gold circulation in the northern and central zones of Belgic Gaul, in particular the territories of the Eburones, the Nervii and the Aduatuci.We worked enthusiastically on the project from 2009 until early 2011 and the volume before you is the concrete result of our endeavours. The book is simply organised. We open with a synthesis article summarising the most important archaeological and numismatic results and placing them in a broader historical and geographical context. This is followed by individual presentations on the hoards of Fraire,Thuin-l, -2 and -3, Philippeville, Orp-le-Grand, Heers and Amby. The book closes with Helena Wouters’ contribution on the metal com position of the coins from the different hoards. The importance of the hoards presented in this volume also appears from the prominent place they occupy in museums and other public presentations: the hoards of Thuin-1, Heers and Amby are among the top items in the collections of the IConinklijhe Museu v o o y Kunst en Geschiedenis (Brussels), the GdlloRomeins Museum (Tongres) and the Centre Céramique (Maastricht). We are grateful to the administration and management of the Koninklijke Mused voor Kunst en Geschiedenis (Claire Massart and Anne Cahen), the Koning Boudewijn Stichting (Anne De Breuck), the representatives of the Municipality of Maastricht (Wim Dijkman, Eric Wetzels) and in particular also the Bestendige Deputatie of Limburg (Belgium) for their support in the realisation of this study. We would like to thank all those who have helped to bring this book to its final form: Jaap Fokkema and Bert Brouwenstijn (Archaeological Centre VU University Amsterdam) for making and adapting the illustrations and the many coin plates; Linda Bogaert, Guido Schalenbourg (Gallo-Roman Museum Tongeren), Eddy Daniëls, André Dedoff (VIOE) and Marijke Willaert for photography and illustrations; Roland Fauville and Florimond Wouters for their find reports, and Annette Visser (Wellington, New Zealand) for translating most o f the texts into English. Finally, we also thank two anonymous reviewers for their critical comments on an earlier version of the text. The editors VII Eight gold hoards horn the Low Countries. A synthesis Nico Roymans/Simone Scheers 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Introduction Composition of the hoards The seriation of coin dies: method and results The dating of the hoards O n the dating of the gold coinages ‘de type beige’ The distribution of the Scheers 31 staters and their ascription to the Eburones O n the northeastern diffusion of the Nervian epsilon staters The silver rainbow staters and the Rhineland connection The volume of emissions of Scheers 29 and 31 coinages, The presence of gold ornaments in the hoards The archaeological contexts of the hoards Possible reasons for the deposition of the hoards The hoards ffomThuin and their possible link with the Caesarian campaign of 57 BC Potential setdements with centre functions in the territory of the Eburones Power relations and coin use Conclusions References Appendix 1. Seriation table for coin dies of the gold staters of the Eburones, class I. Appendix 2. Seriation table for coin dies of the gold epsilon staters of the Nervii. Appendix 3. List of sites of Scheers 31 coins ascribed to the Eburones. Appendix 4. List of sites of Scheers 29 coins ascribed to the Nervii. Appendix 5. List o f sites of Scheers 152-V quarter staters ascribed to the Nervii.I I I N T R O D U C T I O N This volume publishes eight new gold hoards from the Late Iron Age that have been found in central Bel gium (Fraire, Thuin-1,-2,-3, Philippeville, Orp-le-Grand, Heers) and the Southern Netherlands (Amby) (see fig. 1). While brief find reports have been published on the hoards of Heers, Fraire andThuin-1, the remaining five have not previously appeared in the literature. The study of these eight hoards has yielded a wealth of new information on the archaeological contexts in which they were found, on the dating of the coinages they contain and on the social role of the coins in the societies of the time. An interesting feature is the mixed nature of five of the eight hoards, which tells us which coinages were in circulation at the same time. The hoard finds have also proved a treasure trove for coin die research. We have been able to identify large numbers of die connections between coins from the different hoards, which has laid the foundation for a comprehensive quantification and seriation of the dies used for several of the coinages.This has enabled us to comment on the relative dating of the hoards and the volume of the coin émissions. I We already know o f a considerable number of coin hoards from Belgic Gaul from the Late Iron Age,1 but m the majority of cases these are old, poorly documented finds, which seriously reduces their scientific relevance. There is a pressing need for new high-calibre publications of hoard finds. Against this background, it is not surprising that the emphasis in this book is on the presentation of primary data Our aim has been to reproduce all the coins from the hoards under discussion in colour and on a scale of 2:1 so that the photos can be used to analyse the coin images. We also explore the archaeological and landscape contexts of the com hoards. Further, we devote considerable attention to questions o f dating and the ascription of coinages to specific tribes or subtribal groups as there is still uncertainty and debate surrounding these issues.They are critical when it comes to deciding whether we can associate the hoards to the historical context outlined in Caesar’s De Bello Gallico (fig. 2). 2 C O M P O S I T I O N OF T HE H O A R D S With the exception of the Fraire, Philippeville and Orp-le-Grand hoards, which also contained gold ornaments, the hoards discussed here comprised only coins. Table 1 gives an overview o f their composi tion. Apart from one quarter stater of Scheers type 152, the three hoard finds fromThuin contain only See especially the overview in Scheers 1977, 872-903. 2 Fig. 2. Tribal m ap o f n o rth e rn G aul at the tim e o f Caesar’s conquest. 4 116 “ 31-lb, Eburones - - silver rainbow stater 4 - - - 152-V, quarter stater 10 - 1 - 4 - " 1 30-V, Treveri - - 25-II, Bellovaci (?) - 30-I, Remi - total - - - - - - “ 105 73 ' " 20 78 1 " “ “ 1 - - 29-11, Nervii 31 1 " 21 5 42 39 gold ornaments - 1 13 47 gold coin blank - 7 22 29-1, Nervii quarter stater, Remi 8 30 30 Thuin-2 4 Orp Philippeville Thuin-1 31-la, Eburones Amby Heers Thuin-3 Fraire Scheers type 1 1 1 1 53 20 16 158 119 2 Table l .T h e com position o f th e hoards in the Fraire/A m by horizon. a single type of coin - class I and II Scheers 29 staters - whereas the hoards of Fraire, Heers and Amby contain a combination of coinages. A salient feature is the considerable degree of overlap in the composi tion of the hoards: Scheers 29 staters of the Nervii (classes I and II; fig. 3) are present in six of the eight 3 SCHEERS 9 ("AMBIANI") \ Fig. 3. lÿ p o ch ro n o lo g y o f the Scheers 29 staters o f the N ervii and their prototype. H H H Sch"”31■ »f g B I (class fig. 4); and silver'rainbow' staters and Scheers 152-V quarter staters (fig. 5) are present in two and three hoards respectively. The com spectrum is dominated by late gold series which Scheers defines as ‘type beige’, struck by groups in the central and north eastern zone of Belgic Gaul, most notably the Nervii (Scheers 29) and Eburones (Scheers 31). Also conspicuous is the presence of Scheers 152-V quarter staters, which - given eir distribution - can perhaps be ascribed to the Nervii (fig. 6)3 Finally and somewhat out of place T hese small coins o f Scheers type 152, class V, form three subclasses (see fig. 5), w hich correspond to vari ants D oyen 1987, types X IV and X V I. We do n o t find 4 this breakdow n into subclasses in Scheers’ 1977 Traité, as coins o f subclass II w ere still unkn o w n at that time. Fig. 4. Typochronology o f th e Scheers 31 staters o f the Eburones and th eir prototypes. in the coin spectrum, are the silver triquetrum rainbow staters, which are representatives of a Rhineland coin tradition. 5 Fig. 6. D istribution o f the Scheers 152-V quarter staters o f th e N e rv ii.T h e num bering refers to A ppendix 5. a hoard; b single coin; c > 5 coins 3 T HE S E R I A T I O N OF C O I N D IE S: M E T H O D S A N D RESULTS Because of the large numbers of coins and the closely parallel composition o f the hoards, we decided to make the analysis of coin dies a particular focus, chiefly in an attempt to then finetune the relative chronologies of the hoards. The methodology for coin die research was o rig in a lly developed for ancient Greek coinages and was introduced into Celtic numismatics by Colbert de Beaulieu. It is used to establish die connections, particularly in studies of hoard finds.3 In the present study the method is applied for the comparative analysis of the dies represented in the hoards o f the Fraire-Amby horizon in order to arrive at a comprehensive die seriation for several coinages,4 most notably the gold staters of the Nervii and Eburones (for which the seriation tables are presented in Appendices 1 and 2). The procedure for studying the dies of the Eburonean and Nervian staters actually involved a com bination of methods. An initial rough chronological order was obtained for the dies on the basis of a coinage s stylistic development. This led to a subdivision into classes and subclasses, with the criterion of weight also being a relevant factor. The seriation was then refined by reconstructing detailed die gene- For examples o f die links betw een gold hoards, see Zieghaus 1995. Earlier attem pts to use the m eth o d o f die linking in this 6 way w ere m ade by Jo h n Sills for several Late Iron Age coinages from B ritain and Gaul. Sills 2003. alogies. This is possible because reverse dies tend to wear more quickly than obverse dies during the minting process, and therefore have to be replaced sooner. As a result, we can frequently establish short ‘genealogies’ for dies, which constitute a sound basis for seriation. This still leaves us, however, with the problem o f ‘floating’ dies. We have included these in the seriation table using stylistic criteria, but their exact position has not been established. Future coin finds, and in particular hoard finds, will allow us to further refine and modify the die seriation. 4 T H E D A T I N G OF T H E H O A R D S We can see at a glance from the composition of the hoard finds that the hoards should be located close together in time. They belong to the same horizon, which is dominated by class I/II epsilon staters of the Nervii and class I staters of the Eburones. O f significance are the numerous die links between the coins from the different hoards (see Appendices 1 and 2).This applies in particular to the class I and II Nervian staters, with six of the eight class I dies from Heers also represented in Fraire. Die links have also been established between two of the four Eburonean staters from Fraire and coins from Heers, Amby and Philippeville, and there is also a die link between a rainbow cup from Fraire and several coins from Amby. In view of their corresponding composition and the many die finks, our first conclusion is that all eight hoards belong to the same chronological horizon, referred to here as the Fraire/Amby horizon. Key coins for the absolute dating of the Fraire/Amby horizon are a POTTINA stater of the Treveri from Heers and Scheers 152-V quarter staters from Fraire,Thuin-2 and Philippeville. Both coinages were also encountered at Alesia at the site where Caesar inflicted a decisive defeat on the collective Gallic army ofVercingetorix in 52 BC.5We can view this fairly reliable historical date of 52 BC as a terminus ante quem for the production of these coinages and probably also for the burial of the hoard finds under discussion.6 Gold staters inscribed with legends in Latin script, such as POTTINA, are generally dated after 60 BC.7The Heers hoard shows that there are no grounds for placing coin hoards containing only class I/II epsilon staters (like those fromThuin) in the period before about 60 BC.They are more likely to belong in the 50s BC. gold hoards proposed dating Fraire, Thuin-1, Thuin-2, Thuin-3, Philippeville early 50s BC Heers, Orp-le-Grand, Amby late 50s BC Table 2. Proposed absolute dating o f the coin hoards o f the Fraire/Amby horizon. The relative and absolute dating of the individual hoard finds can be refined still further. Table 2 offers an overview o f the datings proposed in this volume for the individual coin hoards of the Fraire/Amby horizon.8The oldest are the Fraire, Philippeville and Thuin hoards, which we have dated to the early 50s BC. They contain class I and II Nervian staters (but not the youngest ‘Heers variant’ with crosses; 5 and 655 (Camp C). 6 under discussion. See the discussion below. Cf. respectively Fischer/Gruel 2001, nos. 678 (Camp C) 7 grove 1999,150 (stage 4). We consider it improbable that the hoards were buried after 52 BC. O f significance here is the absence o f class Loscheider 1998,151 (advanced stage o f LT D2); Hasel- 8 For detailed arguments regarding the proposed datings, II staters o f the Eburones in the Heers and Amby hoards, we refer the reader to the separate presentations on the which are regarded as the youngest o f the series o f hoards gold hoards in this volume. 7 see fig. 3), class la Eburonean staters (but not the younger die variants that predominate in Heers and Amby), as well as Scheers 152 quarter staters and silver rainbow staters of the Rhineland type. Next we propose that the hoard of Heers be placed in the second half o f the 50s BC. O f crucial significance here is the POTTINA stater of theTreveri, class lib Nervian staters of the youngest variant with crosses, and Eburonean staters of the youngest variant of class la.9We have only a single coin from the hoard (?) of Orp-le-Grand, an Eburonean stater with a late die o f class la, which argues for the same dating as the Heers hoard. Lastly, the youngest hoard is that o f Amby. The presence o f Eburonean staters of class lb tells us that the date is somewhat later than that of Heers, yet the absence of Eburonean class II staters shows that the hoard will have been buried before the end of the 50s BC.10 We can thus identify an older (Fraire, Thuin 1-3, Philippeville) and a younger group (Heers, O rple-Grand, Amby) in the gold hoards of the Fraire/Amby horizon. Nevertheless, there are so many die finks between the coins in the two groups that there cannot have been too great a time gap between when the hoards were buried. The hoards are likely to have been deposited within the space of a single decade —the 50s BC.11 The hoards o f the Fraire/Amby horizon display a wide variation in used dies among the different coinages, which allows us to say something about differences in the level of wear of the coins.The general picture is that differences in wear are fairly small, with the exception o f the hoard of Heers. The Eburo nean staters comprise 92 mint-condition coins (nos. 25-116) struck with the same, heavily worn pair of dies. In terms of wear, however, they do not differ much from Eburonean staters struck with older dies (nos. 1-24). It is a different story with the Nervian epsilon staters in the Heers hoard. Here we find a big difference in wear between the youngest group of mint-condition coins ‘with crosses’ from class lib (nos. 126-154) and a group of eight older staters from class I (nos. 117-124). The hoards ofThuin, Fraire and Philippeville do not show such differences between epsilon staters in classes I and II. All this confirms our impression that the coins from the Fraire/Amby horizon were only in circulation for a short time which largely corresponded to the 50s BC. Only the emission of Nervian staters from class la may have’ begun slighdy earlier. However, the search for traces of wear is hampered by the regular use of worn dies (such as the die pair O18-R30 for the Eburonean staters from Heers). We should also bear in mind that the wear of gold coins was very much determined by intensity of circulation and by the extent to which the coins’ owners carried them about each day in a purse. This latter practice will have become much more common in times of crisis, such as the years o f Caesar’s Gallic Wars. 5 ON THE BELGE’ DATING OF THE GOLD COINAGES ‘d E TYPE The coin hoards of the Fraire/Amby horizon give us reason to reopen the debate about the absolute dat ing for the series of North Gallic gold emissions ‘de type beige’, which are related both typologically and metrologically. In her synthesis published in 1977, Simone Scheers originally placed them in the period of Caesars Gallic conquests.'2 She finked the start of the emissions to the formation o f an anti-Roman coalition among the Belgic tribes in 58 BC, which sought to resist the impending Rom an invasion. In The numbers Heers 126-154 (Nervii) and Heers 25-116 (Eburones) respectively. The Scheers 31 staters o f class II may have been struck in 52-51 BC (see also note 6). Post-conquest gold emissions are rare in Belgic Gaul and often concern inscribed coin ages. Cf. Loscheider 1998,156; Haselgrove 1999, 149. This is not to say, however, that the minting o f some coin series may have begun somewhat earlier. According to Haselgrove, the Scheers 152 quarter staters first appeared in the earlier part o f his phase 3 (c. 125-60 BC), but younger variants were still in circulation in the 50s BC. Cf. Haselgrove 1999,139^140; Doyen 2005. 12 Scheers 1977, 60-83. the years that followed, according to Scheers, coin production continued with varying intensity, but by the time Caesar’s troops withdrew in 51 BC there was virtually no native gold production and circula tion to speak of. Although Scheers observes a certain typological and metrological evolution within the coin series, she believed that these occurred in a period of less than ten years. The period of the Gallic Wars corresponded with a sudden rise in the need for coins to fund warfare and the intra- and intertribal coalitions related to it. Scheers clearly used a historical dating model for the gold emissions that she studied. She could hardly have done otherwise in 1977, when there was an almost complete absence of independent archaeologi cal dating information for Late Iron Age coinages.13This is now starting to change. Although researchers do not harbour any doubts about the relative dating that Scheers proposes for the different types and classes, there is considerable debate about the absolute dating. In 1999 Haselgrove presented a revised archaeological dating model for the evolution of Celtic coins in Northern Gaul.14 He argued for an extended chronology, placing the gold series de type beige, which Scheers had earlier associated with the Gallic Wars, in the latter part of his stage 3, which corresponds to LaTene D2a, or the period between 90-60 BC.This includes among others the Scheers 29 and 31 coinages, which have been ascribed to the Nervii and Eburones respectively. The problem, however, is that precise archaeological datings continue to be difficult because of the limited chronological resolution of the material.There is no ‘hard’ archaeological evidence for dating the Scheers 29 and 31 staters to the period before 60 BC. Based on the material from the eight hoard finds, we propose qualifying Haselgrove’s early dating of the coin series de type beige and reconsidering the link with the period of the Gallic Wars. The seriation tables for coin dies (Appendix 2) reveals that class lib Nervian staters were still in full production at the time when the Heers hoard was buried and that the production o f class III had not yet begun. This does not mean that we should go back to Scheers’ historical terminus post quem o f the winter of 58/57 BC, when the pan-Belgic coalition against Caesar was being put together. The emission of some gold series de type beige may have begun towards the end of Haselgrove’s stage 3, but production certainly continued, not reaching a peak until the 50s BC, or the beginning of Haselgrove’s stage 4. Johan van Heesch has given us a further argument for a late dating of the Scheers 29 staters of the Nervii. He notes that class II and III Scheers 24 uniface staters exhibit the densest distribution in Nervian territory between the Scheldt and Sambre rivers (fig. 7) and that, like the epsilon staters, they can there fore be regarded as a Nervian emission. Given the considerable differences in gold content and weight, however, it is unlikely that the two coinages were produced contemporaneously.15Van Heesch suggests that the class II/III uniface staters are somewhat older than the Nervian staters. Bearing this in mind, it is perhaps no coincidence that uniface staters of class II/III are absent from the hoards of the Fraire/Amby group, although stray finds do show that they circulated in this region (fig. 7).This also gives us a better understanding of the coin spectrum from the sanctuary of Empel.16The earliest stage in the spectrum corresponds to the Fraire/Amby horizon or the 50s BC, characterised by the presence of Scheers 31 staters (24 coins), silver rainbow staters (12 coins) and Scheers 29 staters (4 coins). However, there is a conspicuous absence of uniface staters, which have been found elsewhere in the Southern Netherlands region (fig. 7). It would appear that uniface staters o f class II/III were no longer in circulation when coin deposition began at the cult place of Empel in the 50s BC. 13 Scheers (1977, 23) foresaw the future value o f archaeo 14 Haselgrove 1999. logical datings for Celtic numismatics: ‘il est certain, que, 15 Van Heesch 1998,32,35.The gold content o f the uniface d’un jo u r à l’autre, the nouvelles monnaies peuvent sur staters averages between 61 and 65%, as against about 50 gir du sol, et que les fouilles ou les trouvailles pourront to 52% for the epsilon staters.The weights come to 6.20 bouleverser l’image que nous allons essayer d’esquisser for uniface coins and 5.9 to 6 g for epsilon coins. dans ce livre.’ 16 Roymans/Aarts 2005, especially Table 1 9 Fig. 7. Distribution o f Scheers 24 uniface staters, class II/III, probably produced by the Nervii. After Huysecom et al. 1987,112, with additions (Roymans 2004, Appendix 4.1, and documentation S. Scheers). a hoard; b single coin; c several coins 1 Waardenburg; 2 Lieshout; 3 Bladel; 4 Kessel; 5 Baexem; 6 Cadier en Keer; 7 Leuze-en-Hainaut; 8 Fontaine-Valmont; 9 Frasneslez-Buissenal; 10 Anvaing; 11 Hensies; 12Thulin; 13 Mons, vicinity; 14 Ciney; 15 Evrehailles; 16 Grand-Leez; 17 Tielt-Winge; 18 Aalter; 19 Tielrode; 20 Braives; 21 Ghlin, Long Coron; 22 Waudrez, near Binche; 23 Oupeye; 24 Pommeroeil; 25 Denain; 26 Flaumont-Waudrechies; 27 Flines-les Riches, M er de FKnes; 28 Sarcus; 29 Calais, vicinity; 30 Crepy; 31 Frencq; 32 Hamblain-lesp res; 33 Henin-Beaumont; 34 Neuville-sous-Montreuil; 35 Sangatte; 36 Tardinghen; 37 W arlencourt-Eaucourt; 38 FaUencourt; 39 Fesques; 40 Abbeville, vicinity; 41 Fontaine-sur-Somme; 42 Marchelepot; 43 Morvillers-St-Saturnin, Digeon; 44 Mirecourt, vicinity; 45 Ziilpich There is every indication that the coin series defined by Scheers as de type beige, and which Haselgrove dates to his phase 3, included both older and younger emissions. The series from the Fraire/Amby horizon, especially the Scheers 29 epsilon staters and Scheers 31 staters, represent a younger group whose circulation and production were concentrated in the 50s BC.This dating of the hoard horizon renders a link with the Caesarian conquests highly plausible. Although it remains conjecture, we can perhaps asso ciate the burial of the Frame, Thuin and Phihppeville hoards with Roman campaigns against the Nervii and Aduatuci m 57 BC (fig. 8), and the Heers, Orp-le-Grand and Amby hoards with the Eburonean Revolt of 54 BC and the subsequent Roman punitive campaigns against the Eburones. The new dating evidence offered by the hoards o f the Fraire/Amby horizon can be used for mak ing cross-links with other hoards from Northern Gaul, especially those containing Scheers 29 epsilon staters. O f particular interest in this regard is the Ledringhem (F) hoard, discovered in the 19th century 10 ADUA BIANI ATREBATESi BELLOVAC Fig. 8. Caesar’s military campaigns o f 57 BC and the approximate locations o f major battlefields. Brown: tribes united in the anti-Rom an coalition led by the Nervii. 1 along the river Aisne; 2 along the river Sabis/Sambie; 3 siege o f oppidum o f Aduatuci in a ceramic container.1611“ The hoard had a mixed composition o f epsilon staters of the classes III and IV (with the VIROS legend) and Scheers 24 uniface staters of the classes IV and VI. According to our chronological model the presence of class III and IV epsilon staters (both absent in the Fraire/Amby hoards) means that the Ledringhem hoard postdates the deposition of the Heers hoard, and should be dated at the very end of the 50s BC.The class IV epsilon staters with the VIROS legend must have been contemporary with the class II Eburonian staters. The association of late epsilon staters with late uniface 16b“ Scheers 1977, hoard no. 49.T he hoard from Frasnes-lez- —and early Scheers 24 uniface staters o f the classes I— III Buissenal (B) also consisted (apart from the famous gold (Scheers 1977, hoard no. 36). However, the find reports ornaments) o f a mixture o f Scheers 29 epsilon staters — are very ambiguous which makes the association o f the including several late specimen with the legend VIROS two coin types highly unreliable. See Scheers 1990b, 159. staters in Ledringhem offers possibilities for making further cross-links with other hoards in Northern Gaul and Britain, but that falls outside the scope o f this study. 6 TH E D I S T R I B U T I O N A S C R I P T I O N TO T HE OF S C H E E R S 3 I STATERS A N D T H E I R E B U R O N E S Figure 9 shows the distribution of Scheers 31 staters, a coin series with a limited distribution area that largely corresponds to the southeast Netherlands and central and northeastern Belgium. The coins are generally ascribed to the Eburones, but for German archaeologists in particular this remains a matter for discussion. We will attempt below to arrive at a delineation of Eburonean territory based on historical information, and will then return to the distribution pattern for Scheers 31 staters. The Eburonean area is only referred to in vague terms in Caesar’s Commentaries, a fact which has led to differences o f opinion among Dutch, Belgian and German scholars as to its precise boundaries. These differences have arisen in part because the Eburonean territory did not survive as an administrative entity into the early imperial era, rendering later sources virtually useless. Caesar informs us that the main area of the Eburones lay between the Rhine and Meuse rivers.17 Their neighbours to the west were the Nervii and the Menapii. The border with the Nervii probably followed the Rupel and Dijle rivers towards the Sambre,18 while the border with the Menapii probably traced the Lower Scheldt through western North Brabant towards the Rhine delta. Caesar reports that Menapian territory at that time extended to just above the Rhine.19 However, De Laet and Van Es have argued convincingly that Caesar, when referring to the Rhine, in fact meant the Waal river, the R hine’s southern main branch.20A tradition has also existed since the 19th century of extending the Menapian territory eastward across the entire Rhine/Meuse delta to Kleef and beyond, an interpretation that we still encounter among some German archaeologists.21 However, this view has been conclusively refuted by De Laet and Van Es.22 Eastern North Brabant and Gelderland as far as the Waal (i.e. the greater part of the later Batavian region) must have belonged to Eburonean territory. This is also evident from the site of the batde between the Romans and the Germanic Tencteri and Usipetes at Kessel/Lith, near the confluence of the Meuse and Waal. We can infer from Caesar’s account that this battle was fought on Eburonean territory.23The Eburones’ neighbours to the southwest were the Aduatuci, but we are unable to pinpoint their territory more precisely. They are usually placed on both sides o f the Sambre, and/or in the area between the Sambre and Meuse. In the south, the Eburonean area bordered on that o f some smaller Germamc groups - the Condrusi in the Condroz and the Segni, Paemani and Caerosi in the Ardennes and Eifel. Furnished with this information, we can form a general picture of the Eburonean territory (fig. 2). If we compare this with the distribution of Scheers 31 staters (fig. 9), it is clear that this coinage must have been an Eburonean emission. However, there are some question marks concerning this point. Firsdy, we note that a number of find spots are also known from the area between the Sambre and the Meuse, which probably belonged to the Aduatuci. However, the composition of the contemporaneous gold hoards of Fraire,Thuin and Philippeville show that this region lay outside the core circulation area for Eburonean coins; the hoards are clearly dominated by Scheers 29 epsilon staters of the Nervii and only in Fraire 17 Caesar, BG 5.24. Raepsaet-Charlier 1994. She bases herself here on the reconstructed boundaries o f the later civitas Nerviorum. 19 Caesar, BG 4.4. C f.D e Laet 1961. 20 D e Laet 1961;Van Es 1981,25-27. .<> 12 21 Cf. Bechert 1982, 53-54 and fig. 50; Sicherl 2009, 55; Heinrichs 2009. 22 De Laet 1961, 22-24;Van Es 1981,25-26. 23 Caesar, BG 4.14. Cf. D e Laet 1961, 23. Fig. 9. Distribution o f the Scheers 31 gold staters ascribed to the Eburones.The numbering refers to Appendix 3. a = 1-5 coins; b = >20 coins; c = hoard find and Philippeville were four Scheers 31 coins present. Secondly, almost no Eburonean coins are known from the German Lower Rhine region west of Cologne. Going by Caesar’s comment that the heart of Eburonean territory lay between the Meuse and the Rhine (inter Mosam ac Rhenutri), we would expect more find sites there.24 It is hard to explain their absence. The lack of finds may partly be due to a less intensive detecting and reporting of coins. Scheers has suggested that Caesar’s comment should perhaps be understood from a north-south rather than an east-west perspective.25 Another interpretation is that the Eburones were a segmented polity involving a considerable degree of autonomy for the subtribes.26 The distribution map shows which subgroups used the coins and which ones used them not. Finally, German archaeologists have come up with a completely different interpretation: the Lower Rhine region 24 German researchers see this as grounds for interpreting 26 We can point here to the dual kingship o f the Ebu the Eburonean coins found west o f the Meuse as having rones, which may have had a territorial basis. Caesar been brought and lost by Eburonean refugees outside (BG 6.31.5) refers to Catovolcus as rex dimidiae partis their core territory (Joachim 2007, 54; Heinrichs 2009, Eburonum. Cf. Roymans 1990, 33, 37. The only subtribe 212-213), or as emissions o f the Aduatuci, Menapii or o f the Eburones that Caesar (BG 4.9.3) mentions are Nervii (Sicherl 2009, 57, note 3). the Ambivariti, who lived along the Meuse in northeast 25 Scheers 1996, 32. Brabant and north Limburg. 13 Fig. 10. Distribution o f Scheers 29 staters, classes I-III, ascribed to the N ervii.The numbering refers to Appendix 4. a. hoard; b. sanctuary; c. isolated find west of Cologne appears to have been virtually uninhabited from the early 1st century BC to the m id-lst century AD.27This would support Scheers’ proposal o f giving a different reading to the relevant passage from Caesar. Whatever the case, we have no reason to doubt the ascription of the triskeles staters to the Eburones as there is no plausible alternative attribution. 7 ON TH E E P S I L O N N O R T H E A S T E R N D I F F U S I O N OF N E R V I A N STATERS The staters of Scheers type 29 belong to the dominant coin series in the central zone of Belgic Gaul. Based on their distribution, this series is fairly generally ascribed to the Nervii (fig. 10).28They constitute an important element in the hoards under discussion. It is perhaps not surprising to see their relative proportions decline in the hoard finds as we move northeastwards - from almost 100% (Thuin-1, -2 and -3) and 83% (Fraire), to 25% (Heers) and finally to 0% (Amby). However, the presence of Nervian staters in the Heers hoard in the Eburonean area is not an isolated case. There was a significant northeastward diffusion of this coinage, with 17 find sites of stray coins now known from the northeastern peripheral zone of Belgic Gaul (fig. 10). Interestingly, they almost always involved coins from classes I and II. C£ the discussion (with further references) in Andrikopoulou-Strack 2001. 14 28 Cf. the recent analyses by D eru 2009. The question is how we should interpret this flow of epsilon staters to the Eburonean territory in particular. The dating of the coin flow is critical here. The hoard of Heers shows that class I and II epsilon staters were in full circulation in the 50s BC, suggesting a connection with the Roman conquest. The diffusion of the coins may be linked to the historically documented alliance of the Nervii and Aduatuci with the Eburones during the revolt of 54 BC, or perhaps with Nervian refugees fleeing northwards after suffering a heavy defeat at the hands of Caesar in 57 BC.29 In any event, this coinage was a well-known means of payment in the northeastern peripheral zone of Belgic Gaul in the 50s BC, akin to the probably somewhat older uniface staters of Scheers type 24-11/111. 8 TH E SILVER R A I N B O W STATERS A N D T HE R H I N E L A N D C O N N E C T I O N In the Fraire and Amby hoards, the gold staters de type beige of the Nervii and/or the Eburones are shown to be associated with triquetrum-type Rhineland silver rainbow cups. Both must therefore have circulated at the same time in the 50s BC. A surprising discovery is that this dating also applies to silver rainbow cups bearing the characteristic additional marks of what we call the Lith group;30 these are pre sent in both Amby and Fraire. These rainbow cups represent an advanced stage of a Rhineland coin series; they were the immediate successors in silver of identical gold coins of the Mardorf group, which were struck between about 80 and 60 BC in the oppidum on the Dunsberg in Hessen, Germany. Subsequently, a major shift occurred in the circulation of triquetrum coinages from the eastern Middle Rhine to the Lower Rhine region. The distribution patterns for silver triquetrum coins point to close relations between tribal groups on both sides o f the Rhine. Just how we should interpret these relationships is the subject of debate. Based on distribution patterns for variants with additional marks (fig. 11), Roymans and Dijkman suspect that both the circulation and production of silver coins from the Lith group shifted westwards to the regions around Kessel/Lith and Maastricht in the Southern Netherlands. The silver rainbow cups have so far been placed in the period between 50 and 30 BC and associated with Tacitus’ account of Batavian migration from the area east of the Middle Rhine to the Dutch river delta.31 Now that we know that the silver rainbow cups were produced during the period of Caesar’s conquest rather than in the first few decades thereafter, this disposes of the previously supposed Batavian connection. We should now primarily understand the coins within an Eburonean context. Roymans and Dijkman postulate that we are dealing here with a coinage struck by a northern subgroup of the Ebu rones. This group maintained close connections with groups east of the Middle Rhine and was probably absorbed later into the new Batavian polity that evolved during the post-Caesarian period. It is there in particular that the circulation, and in part also the production, of the more recent bronze variants of the rainbow cups (‘Bochum type’) continued.32 These trans-Rhenine connections of the Eburones with Germanic groups are also historically docu mented. The Eburones themselves are described by Caesar as Germani cisrhenani, an implicit reference to close cultural and political finks with groups across the Rhine. We see this in the course of the Eburonean revolt o f 54 BC, when their leader Ambiorix received military backing from Germanic allies from across the Rhine.33 29 Caesar, BG 5.38 and 2.28 respectively. The Nervii had chapter 6. struck an alliance with the Aduatuci earlier in 57 BC for 32 Roymans 2004, chapter 6, with further references. the coming battle against the Romans (BG 2.16). 33 Caesar, BG 5.27 and 6.5. See the discussion in Roymans/ 30 See Roymans/Dijkman, this volume, fig. 9. Dijkman, this volume, Section 4. 31 For a comprehensive discussion, see Roymans 2004, 15 Fig. 11. Distribution o f silver triquetrum coins o f the Lith group (variants b, c, h, i, t, u) and the presumed shift o f production from the oppidum at the Diinsberg to the Eburonean region. Triangle: hoard find. 9 TH E V O L U M E OF E M I S S I O N S OF S C H E E R S 29 A N D 3 I C O I N A G E S The number of coin dies identified for the various classes of Scheers 29 and 31 staters has risen sharply with respect to the numbers given in Scheers’ 1977 synthesis. For the Scheers 29 and 31 coinages, we have been able to arrive at a precise quantification and seriation of the number of dies used (see Appendices 1 and 2). Table 3 gives an overview of the number of obverse dies identified for both coin series. This count is based on coins from the hoards discussed here and those found elsewhere. For the Scheers 29 class I— III staters, we now have a total of 147 dies, while 75 dies have been documented for the Scheers 31 staters. Given that there are still many discontinuities in the seriation tables shown in Appendices 1 and 2, the actual number of dies will certainly have been greater. Based on this data we can obtain an overall picture of the volume of the emissions, assuming a production rate of 1,000 coins per die, a figure that is among the lowest estimates usually given for the production capacity of Late Iron Age coin dies.34This 34 Haselgrove 1984, 90; cf, also the results o f recent experi- makes a much higher production rate (5,000 - 10,000 ments with coin striking in De Jersey 2009, 257, and coins per die) not unrealistic. Cf. for example the coins Faucher et at. 2009. The hoards o f Heers andThuin pro- o f the Eburonean dies O 18-R 30 from the Heers hoard, vide evidence for the use o f extremely w orn dies, which l6 means that at least 147,000 Scheers 29 staters will have been produced, while the number of Scheers 31 staters certainly exceeded the 75,000 items. The emission of Eburonean staters attained only a half the size of that for Nervian staters.35 Nevertheless, both coin series, certainly those of the Eburones, can be considered among the smaller gold emissions of Belgic Gaul. no. of dies est no. of coins kilograms of fine gold Scheers 29, class l-lll 147 147,000 455 (52% gold) Scheers 31, class l-ll 75 75,000 128 (30% gold) Table 3. N um ber o f identified (obverse) dies and the estimated volume o f the Scheers 29 and 31 gold series. In the northern half of Belgic Gaul, the area between the Scheldt and Sambre rivers (roughly the ter ritory of the Nervii) has emerged as the region where gold circulated most intensively. This raises the question of where all the gold used for these emissions came from. We can only conjecture about this.36 It may have been resmelted gold that had previously circulated in the form of ornaments or coins, newly extracted gold from the adjacent Ardennes Massif, or imported gold from other regions, including Britain. IO TH E P R E S E N C E OF G O L D O R N A M E N T S IN T H E H O A R D S The gold hoard of Fraire also contained a simple but solid gold armband. A fragmented gold buffer torque Was recently found at Orp-le-Grand, with a Scheers 31 stater in the immediate vicinity. They may have come from a hoard that had been disturbed by ploughing, and which we can therefore also place in the Fraire/Amby horizon. In addition, the terminal knob of a gold torque was found as a stray find in Thuin, in close proximity to the hoards ofThuin-1 and -2. Lastly, a twisted gold armband with animal-headed terminals was recendy found in a coin hoard near Philippeville. These ornaments are discussed in greater detail in the contributions later in this volume. The presence of gold torques and arm jewellery, either intact or fragmented, in coin hoards is a well-known phenomenon in the central northern zone of Belgic Gaul (fig. 12).37 Examples are the hoard find of Frasnes-lez-Buissenal, unearthed in the 19th century, and the recently excavated hoards of Niederzier and Beringen. These finds suggest that the owners of large numbers of gold coins also wore gold ornaments. The torques and armbands were symbols of power and authority; they underlined the fact that the owners belonged to an elite milieu. In addition, gold torques had divine associations.38 O f interest is the presence of parts of ornaments in the hoards. There are different possible interpreta tions.The first is that these objects were ploughed up in recent or sub-recent times and suffered damage as a result. The highly fragmented torque of Orp-le-Grand, with almost all the components recovered, appears to be an example o f this. However, fragmented torques and armrings are also known from hoards found in situ, such as those of Niederzier and Beringen. This may have been hackgold, intended for res melting and for which only the weight mattered. An alternative third option is that fragments of orna ments were not anonymous scrap gold but, like intact jewellery, were ‘animate’ objects with a specific 35 These calculations tally quite well with Haselgrove’s ear lier estimates. H e (1984, Table 2) estimated the number o f dies for Scheers 29 (class I —III) and Scheers 31 (class I —II) at 166 and 40 items respectively. 37 For a recent overview o f the evidence, see Fitzpatrick 2005. 38 For a broader discussion, see Van Impe et a/. 1997/1998, 75-92; Fitzpatrick 2005, esp. 172-173. 36 Cf. Haselgrove 1984,91-92; Roymans 1990,139-140. 17 Fig. 12. Distribution o f Late Iron Age gold ornaments (torques, armrings) in Belgic Gaul, a hoard find associated with gold coins; b cult place; c river find ‘biography’ referring to earlier owners. For example, the old jewellery o f deceased leaders may have been deliberately divided up and distributed amongst followers or descendants.39 I I TH E A R C H A E O L O G I C A L C O N T E X T S OF T H E H O A R D S A key question in the study of coin hoards concerns the archaeological contexts in which these are found. We are currendy poorly informed about the find contexts of the hoards under discussion. We will briefly set out the available information so that we can then arrive at some conclusions. Three hoards were found in modern arable land (Amby, Heers, Orp-le-Grand), where they had been pardy dispersed by (sub-)modern ploughing. The other hoards were discovered by metal detectorists at other locations, but these, too, seem to have been scattered across an area of several tens of square meters. This brings us to the fundamental question whether the coin ensembles should be interpreted as single hoards, or are we perhaps dealing with several hoards?40A practice o f multiple hoarding of Late Iron Age coins and other objects has been documented for cult-sites (see below). O ur impression is that most cases discussed in this volume represent single hoards which had been more or less spread by later cultivation O n the cultural biographical approach o f objects, see Kopytoff 1986. O n ceremonial gift exchange and the constitution o f person and society in early medieval Anglo-Saxon society, see Bazelmans 1999,168 ff. Cf. also John Chapman’s work (2000) on the fragmentation o f objects. 40 Cf. the discussion in Haselgrove 2012. or other disturbances of the top soil. A control excavation carried out at Amby demonstrated that a hoard can be spread over an area of more than 800 square meters as a consequence of (sub-)modern ploughing, and at Orp-le-Grand fragments of the same gold torque were found scattered over a distance of c. 80 m. Another argument pleading for single hoards is the lack of time-depth in the coin ensembles of the Fraire-Amby horizon. Only in the case of the gold hoards from the Bois de Luiseul atThuin do we have good arguments for assuming a multiple-deposition site (see below). We know the exact locations of the Heers and Amby hoards; the latter was in fact salvaged thanks to a control excavation. In both cases the hoards were buried at off-site locations. The Orp-le-Grand hoard is probably the remains of a ploughed-up gold hoard, buried in highly eroded ground where later a GalloRoman villa developed. Only the approximate find spot is known for the Fraire hoard. It was allegedly found on a wooded slope beside a stream, a spot that has yielded no other reported finds.The gold hoards ofThuin come from two different locations, the peripheral zone of a Late Iron Age fortification and a site immediately outside it. The latter site is an area measuring 20 by 40 m in the Bois de Luiseul. In addition to the two gold hoards, a hoard of potin coins, a few isolated Late Iron Age coins and a gold torque frag ment were deposited there. Other find categories are not represented, nor is material from Roman times. This was most likely an open air cult place, but a controled excavation will have to establish this. Possible parallels are the cult-sites of Snettisham (UK), where at least 12 Late Iron Age hoards comprising mainly torques, coins and ingots were excavated within a large enclosure, and Hallaton (UK), where a cluster of 16 Late Iron Age coin hoards has recently been found inside a ditched enclosure.41Although discrete deposits, the Hallaton hoards are similar in composition and may have been buried at the same time. We may conclude that three of the hoards under discussion (Heers, Amby, Fraire) appear to involve off-site depositions at places that were only used once as deposition sites; there is no evidence here for the presence of a setdement or cult place.42 One instance (Thuin-1) involved a deposition in a Late Iron Age fortification and the two other hoards ofThuin involved a multiple-deposition site, probably a cult place. 12 P O S S I B L E R E A S O N S FOR T H E D E P O S I T I O N OF T HE H O A R D S A major point of discussion in the study of coin hoards is the reasons for their burial. Before we examine this question more closely, we would like to make three observations. The first is that the hoards of the Fraire/Amby horizon consist almost entirely of local coinages normally in circulation in these regions. This suggests that the people who buried the hoards were representatives of the local population.43The second observation is that all the hoards involved shallow burials and were therefore in principle retriev able. Unlike depositions in wet contexts like rivers and marshes, they could be dug up again and put back into circulation. The third observation concerns the lack of chronological depth in the composition of the coin hoards of the Fraire/Amby horizon. Representatives of older coinages do not feature at all; Snettisham: Stead 1991; Fitzpatrick 2005, 175-176. Hal while the hoard is placed at the end o f the 2nd century laton: Score 2006; idem 2012; Haselgrove 2012. BC.This discontinuity may be deceptive, however, given This is a marked difference from two older gold hoards that only a small part o f the settlement has been investi from the same region, that o f Niederzier (Gobi et al. gated. 1991) and Beringen (Van Impe et al. 1997/1998, 13 ff), 43 This contrasts with the older gold hoards o f Beringen where excavations have suggested that they were buried and Niederzier in the same region, which consist of in settlement contexts and may have been ritual deposi coinages that were imported from regions further away. tions. The problem here is that the most recent settle See Gobi et al. 1991; Van Impe et al. 1997/1998,13 ff. m ent traces in Beringen are dated to the 3rd century BC, 19 these appear to have disappeared entirely from circulation and been converted into new coinages with a lower weight and gold content. The relative contemporaneity of the hoard finds, their dating in the period of the Caesarian conquest and their composition of local coinages argues for a deposition by local groups or individuals in a period of social stress. Assuming a connection with the Roman conquests and the risks associated with warfare, we can arrive at various possible interpretations concerning the reasons behind this hoarding practice. A well-known interpretation is that of Versteckdepots - hoards buried in times o f turmoil and intended to be dug up again later. There are question marks about this, however. Profane hoards buried for safety are more likely to be found on house compounds at settlement sites than at off-site locations or a cult place. A second interpretation is that of votive offerings to a deity by communities or individuals to express their gratitude for success in battle. This votive practice, documented in the writings of classical authors on Celtic-Germanic peoples,44 often involved all or part of the spoils of war and the depositions were in principle intended as permanent offerings to the deity. It is in this sense that we can interpret coin hoards found at cult places. The difficulty here though is that the hoards of the Fraire/Amby horizon would suggest the military success of native groups, whereas Caesar s reports show that the opposite was the case. The assumption that profane Versteckdepots and sacred votive deposits were mutually exclusive catego ries is overly simple, however, and it brings us litde closer to understanding hoarding practices in times of war.45 In pre-Roman Gaul the motives behind these depositional practices were undoubtedly more complex than the scarce classical sources suggest. For example, ritual motives may also have played a role in the temporary burial of hoards in times o f crisis. Hoards might have been placed under the temporary protection of a deity for safety reasons. The place of burial may also have had supernatural associations or even have been a regular cult place. A further possibility is that the owners vowed a portion of the buried wealth to a deity in exchange for military success and the protection of the entrusted valuables. We can break through this problematical dichotomy between profane Versteckdepots and ritual hoards by making a distinction between the reasons for burying hoards and the form that depositions took. The immediate reason could be a crisis, but the form may have been ritually determined. This viewpoint allows us to understand significant patterns in the distribution in time and space of the hoard finds in the Fraire/Amby horizon. It seems to be no coincidence that the southern group (the hoards of T h u in Fraire and PhilippeviUe) belongs to the older phase of this horizon, and the northern group (the hoards of Heers, Orp-le-Grand and Amby) to the younger phase (see above, section 4).This pattern (fig. 1) ties in nicely with the picture outlined by Caesar of large-scale warfare, in the territory of the Aduatuci in 57 BC and in that of the Eburones in 54-53 BC. 13 TH E W I T H H O A R D S THE F R O M T H U I N C A E S A R I A N A N D T H E I R C A M P A I G N OF P O S S I B L E 57 L IN K BC A special situation presents itself in Thuin, where three gold hoards, several stray gold coins and the terminal of a gold torque were found in or immediately outside the fortification in the Bois du Grand Bon Dieu (fig. 13). This éperon barrée-type fortification occupies a plateau of more than 13 ha. and can be reached on the eastern side via a narrow, 60 m wide finger of land. Apart from a coupe through the rampart and the ditch, archaeological investigations have yet to take place there. An important clue is pro vided by a radiocarbon date for charcoal from the rampart, which gives a date between 90 BC and 60 AD (68.2% probability).46A dating in the last centuries BC is confirmed by the find of Late Iron Age metal 44 For an overview o f the discussion and relevant historical 45 sources, see for example Roymans 1990,77,83,130; Nick 46 2005,141 ff. 20 See also Nick 2005,143-144. Scheers/Creemers/Roymans/Van Impe, this volume. Fig. 13. Topography o f the Late Iron Age fortification (brown) at Thuin and the location o f gold finds and R om an lead sling bullets, a gold hoard; b isolated gold coin(s); c concentration o f sling bullets; d iron tools; e bronze ornaments and appliques objects, including iron tbols, bronze ornaments and appliques, and a gold coin of the Eburones. There are no finds at all from the first two centuries AD, from which we may conclude that the settlement definitely did not survive into Roman times.We do not know whether the fortification was permanently inhabited in the Late La-Tene period, or whether it was only used on an incidental basis in times of crisis. The seriation table for coin dies of the Nervian staters from the three hoards (Appendix 2) shows that these must have been buried at almost the same time, namely the early 50s BC.The two stray Eburonean staters must also have already been in circulation at that time since they were struck using the earliest dies of 21 class la.47This raises the question of whether the gold depositions atThuin may be associated with the same historical event. It is our hypothesis that we are dealing here with the oppidum of the Aduatuci, described by Caesar, which was seized by the Romans in 57 BC and whose entire population of 53,000 was sold into slavery.48*We presume that the area between the Sambre and Meuse rivers in present-day Belgium belonged at that time to the Aduatuci. On the basis of Caesars description, their territory can roughly be defined as lying between that of the Nervii and the Eburones, and flanking both sides of the Sambre. An important argument in support of our hypothesis is that the topography o f the fortification at Thuin is a remarkable match for Caesar’s description of the oppidum of the Aduatuci: '[The Aduatuci] col lected all their possessions into one oppidum, which had remarkable natural defences. It was ringed around by very high, sheer rocks except at one point, where there was a gently sloping approach no more than 200feet wide. They hadfortified this place with a very high double wall on which they were now setting heavy stones and great wooden beams sharpened to a point.’ 49 It was therefore an éperon barrée with a narrow access way that was 200 feet wide (about 60 m).The only respect in which Caesar’s description differs from the situation in Thuin is the presence of a double wall. It must be said, however, that the remains of a second rampart still to be investigated more closely can be seen on the southwestern side of the principal r a m p a rt atThuin.80 The recent discovery by a metal detectorist o f a large number of Roman lead sling bullets lends addi-1 tional support for our proposed identification of the fortification of Thuin as the oppidum of the Aduatuci laid siege to by Caesar. The sling bullets appeared in two separate concentrations: on the wall near the main entrance of the fortification and on the other side of the Biesmelle river in the Bois de Luiseul (fig. 1 3 ). The concentrations contained examples of biconical, pointed oval and rounded oval types (fig. 14). Based on their weight and shape, it is difficult to date them more precisely th a n the late republic or earliest imperial era; they were rarely used by the Roman army after the Augustan period.51 In any event, the concentrations of lead sling bullets suggest that the fortification was once besieged by the Roman army in the Late Iron Age or earliest Roman times. The distribution of these projectiles suggests that the Roman attacks focused on the principal rampart and on another target across the Biesmelle river in the Bois de Luiseul which cannot be identified more closely.52 Are there other possible historical explanations for the Roman military action atThuin? One alterna tive is to identify the fortification as the winter camp of Caesar’s general Cicero in 54 BC.This Roman camp lay in Nervian territory and was attacked by a mixed force of Eburones, Nervii and Aduatuci dur ing the revolt of 54 BC.53The problem here, however, is that the sling bullets would have been deployed by the Roman defenders of the camp, whereas their concentration at the m ain ra m p a rt suggests that they were in fact used by the attacking side. Moreover, this interpretation also means that the link with the gold depositions would fall by the wayside.This is because the gold depositions atThuin belong to the older phase o f the Fraire/ Amby horizon, which certainly pre-dates the Eburonean revolt. O f course, we cannot rule out that we are dealing with a Roman army action in the decades after Caesar’s Gallic 47 Respectively dies 0 2 and R 2 (Bois du Grand B on Dieu), and 0 4 and R 5 (Bois de Luiseul). Similar early dies 50 Bonenfant/Huysecom 1982, 105 and fig. 1. Caesar (BG occur in the Fraire hoard. See Appendix 1. 2.30) also mentions the R om an army’s circumvallation Caesar, BG 2.33. Brulet (2008, 31-34, and esp. fig. 34) o f the oppidum over a distance o f five miles. It is not suspects that the location o f the battle described by Caesar as taking place along the river Sabis in 57 BC between the Romans and the Nervii together with their known whether traces are still present in the landscape. 51 Volling 1990; Schlott 1999,47-48; Schiilze-Forster 2002, 86-88; Poux 2008, 365-371. allies is to be found in the vicinity ofT huin, but there is 52 Caesar’s description (BG 2.30-33) o f the hostilities at no archaeological evidence at all to support this. O ther the oppidum o f the Aduatuci tells us that fighting also scholars locate this batde on the Sabis further to the west along the river Selle. Cf.Turquin 1955; Deru 2009,181. 22 49 Caesar, BG 2.29. occurred direcdy outside the fortification. 53 Caesar, B G 5.38-53. Fig. 14. R om an lead sling bullets from the Late Iron Age fortification atT huin. Wars that is not historically documented. Arguing against this, however, is the complete absence of early Roman find material inside the fortification. In our view it is a plausible hypothesis to interpret the dispersed gold depositions at Thuin as the reflection of a single historical event and to associate the large numbers of lead sling bullets with this same event, namely the conquest by the Romans of the oppidum of the Aduatuci in 57 BC.54The coin deposi- 54 There are at present no serious alternative candidates for the oppidum o f the Aduatuci m entioned by Caesar. Cf. the discussion in Scheers/Creemers/Roymans/Van Impe, this volume, 83 ff. 23 tions in the Bois de Luiseul can best be interpreted as portable wealth hidden at a cult place at a time of crisis. In his biography of Caesar, Suetonius writes that Caesar was guilty of the large-scale plunder of oppida and sanctuaries in Gaul and of enriching himself enormously with the wealth stored there, most notably in the form of gold: ‘In Gaul he pillaged shrines and temples of the gods filled with offerings, and oftener sacked oppida for the sake of plunder than for any fault. In consequence he had more gold than he knew what to do with, and offered it for sale throughout Italy and the provinces at the rate of 3000 sesterces a pound.’ Since the usual price of gold was 4000 sestertii, Caesar gready inflated the Italian gold market.55The oppidum of the Aduatuci which he conquered in 57 BC will have been an apposite example. Following the capture, not only will the entire population have been sold as slaves, but the Roman army will have systematically plundered the fortification. Caesar will have been chiefly interested in the portable wealth hidden there in the form of coins and jewellery. Only the hoards under discussion remained out of Roman hands, but these will have represented only a fraction of the gold that was seized there. Our hypothesis is that the mass deportation of the Aduatuci after the fall of their oppidum meant that part of the portable wealth buried beforehand in the soil was never recovered, giving rise to an archaeological hoard horizon. Clearly, the above attempt at historical interpretation of the Thuin hoards raises a host of unanswered questions; it is the hope that future archaeological research will supply the answers. We would therefore argue for a relaunching of investigations at Thuin in order to test our hypothesis. 14 P O T E N T I A L THE S E T T L E M E N T S T E R R I T O R Y OF T H E W I T H C E N T R E F U N C T I O N S IN E B U R O N E S How should we picture the settlement pattern among the Eburones? Although the evidence is still rather thin and finds little support as yet from excavations, we are clearly not dealing here with a world made up entirely of small hamlets and isolated farmsteads. The settlement pattern from the Late Iron Age has proven to be more complex and hierarchical than we imagined until recently. Settlements with centre functions existed at the regional or subregional level. Here we will discuss four locations which are likely to have occupied a special place in about the middle of the 1st century BC (fig. 15). The first is the cult place of Empel at ’s-Hertogenbosch in the Dutch river region (fig. 16). A monu mental temple complex where the deity Hercules Magusanus was venerated stood here in Rom an times near the confluence of the Dieze and the Meuse. Starting as an open-air sanctuary, this cult place had its origins in the LaTene D1 period. The earliest objects (LaTene D l) include fibulae of the Middle La Tene and Nauheim types. The number of finds rose sharply for LaTene D2 and comprised for the most part hundreds of fairly late Celtic coins. They included 24 gold staters of the Eburones (Scheers 31), four of the Nervii (Scheers 29) and 12 silver rainbow cups of the triquetrum type, most of them of the Lith group with additional marks. The sanctuary must already have possessed a supralocal significance during this phase.56 Secondly, in the Dutch river area there is the major settlement complex at Kessel/Lith on the south ern bank of the Meuse at the ancient confluence of the Rhine/Waal and Meuse (fig. 17). This complex has been almost completely destroyed in past decades as a result of sand and gravel extraction so that the information available to us consists almost entirely o f large numbers of stray dredge finds.These show that the site did not acquire a supralocal significance until LaTene D l, the period of the Nauheim fibulae. The Late Iron Age material can be divided roughly into two groups: material from an extensive settle ment situated on the southern levee of the Meuse and a ritual find complex originating from an antique bed of the Meuse. The latter comprises large quantities of human and animal bone material, weaponry, 55 Suetonius, Dm M . 54.2. Cf. the discussion in Roymans 1990,143. 24 56 For Empel, see Roymans/Derks 1994; Roymans/Aarts 2005. Fig. 15. Location o f potential settlements with centre functions in the territory o f the Eburones. Fig. 16. General plan o f the sanctuary o f Empel, with traces o f a Gallo-Roman temple and a pre-Rom an open air cult place. After Roymans/Aarts 2005, fig. 3. 25 Fig. 17. Topographical reconstruction o f the Meuse/Waal confluence at Kessel/Lith in the Late Iron Age. After Roymans 2004, fig. 7.13. a. late-medieval river dikes; b. (sub-)modern river forelands; c. probable river course; d. zone w ith ritual depositions in ancient river bed, e. setdement complex from Late Iron Age and Early R om an period; f. native-Roman rural setdement; g. meander belt from Early/Middle Iron Age; h. pleistocene sand deposits belt hooks, etc. This cult place evolved in Roman times into a monumental temple complex. The coins of Kessel/Lith include 23 silver rainbow cups, the majority of which are Lith-group coins bearing additional marks, while gold staters are conspicuous by their absence. It is conjectured that Kessel/Lith functioned in La Tene D as a religious, political and artisanal centre of a subtribe of the Eburones and later of the developing Batavian polity.57 A third find site is located in the village of Berg, several kilometres east of the Roman town ofTongres. Almost no information is known about this site, but past decades have seen regular discoveries of Celtic coins scattered across a large area and belonging for the most part to the same horizon as the hoards of Amby and Heers.58The finds include eight gold staters of the Eburones (Scheers 31), one uni face stater (Scheers 24-1), two epsilon staters of the Nervii (Scheers 29) and several silver rainbow cups, some of them of the Lith group with an additional mark. There are also two unstruck coin blanks for gold staters with weights of 5.39 and 5.54 gr. It is conjectured that a settlement and/or cult place with a supralocal significance was situated in Berg in La Tene D2,59 but excavations will have to confirm this. 57 Roymans 2004,14 ff. and chapter 7. 58 Scheers/Creemers, this volume, 136, fig. 7. 26 59 Van Heesch 2005,257. The fourth location is the Late Iron Age fortification at Caestert, south of Maastricht. Known for some time, this fortification is wedged in between the Jeker and Meuse valleys. It has an Ehrang-type rampart construction, consisting of a murus gallicus with vertical beams at the front. Recent research has demonstrated that the fortification is considerably larger than originally assumed, covering c. 35 ha.60 We understand little as yet of the nature and function of the fortification. Even its precise dating has not been established. The main difficulty here is the lack of dateable find material, which means we have had to rely entirely on a dendrochronological dating of several pieces of carbonised wood from the rampart construction and the ditches. This has produced a dating of 31 BC, which implies that it could also be a Roman military camp from the post-Caesarian period. A recent examination of several new wood sam ples, however, was unable to either confirm or refute the dendrochronological dating. New 14C-datings of charcoal samples from the rampart and other features point to a date between c. 250 and 20 BC.61This leaves open the possibility that we are nevertheless dealing with an Eburonean fortification. Whatever the case, the extreme dearth of find material demonstrates that this fortification was only used for a brief period and was never subject to intensive habitation. It probably functioned as a place of refuge for a tribal or subtribal community in times of crisis. Future research will show whether the above settlements with centre functions were production sites for the gold staters and the Lith-group rainbow cups present in the hoard o f Amby. After all, we know from other regions in Western and Central Europe that fortified centre settlements (oppida) and impor tant cult places are potential production sites for Celtic coins. For the time being, however, there is no evidence o f this in the Eburonean area.62 Nevertheless, judging by the distribution patterns now at our disposal, we can expect that the gold staters were produced in the Tongres/Maastricht region, and the rainbow cups of the Lith group in the Kessel/Lith region. We should bear in mind, however, that both coin groups were struck at a time of crisis and that under such circumstances mints would have had a degree of mobility. 15 P O W E R R E L A T I O N S A N D C O I N USE Another view has been proposed in recent years regarding the social functioning of Late Iron Age socie ties in the Lower Rhine region. In contrast to the traditional picture of simple, highly egalitarian and relatively isolated societies with a virtual absence of social dynamic, we now have a model in which societies were fully subject to social changes, at the level of both ordinary local communities and of tribes and subtribes.63We should not view the Lower Rhine tribes mentioned by Caesar as centralised political systems, but as loosely structured, fluid confederations of subtribes, each with their own leaders and held together by alliances and clientship relations. There was, however, an increasing degree of social hierarchy and complexity, and it is within this constellation that we should understand the explosive rise in the circulation of gold coins and jewellery in the Late Iron Age.64 Relying on historical information, it is generally accepted that high-value gold and silver coins circu lated mainly in the political military sphere in Late Iron Age societies.65 Creighton sees the rise of armed groups of horsemen - whom Tacitus called comitatus - as one of the most important social innovations of the Late Iron Age in Western Europe.66These warrior bands lay at the heart of the power and authority of individual leaders and they were the force behind the process of social hiérarchisation. Warriors had a 60 Verhoeven 2011, 5-8. sen/Roymans 2006. 61 Verhoeven 2011, 59-62. 64 Roymans 1990, 27; Creighton 2000,13. 62 Only the presence o f several unstruck gold flans at 65 Haselgrove 1984; Roymans 1990, chapter 6. Tongres-Berg points in this direction. 66 Creighton 2000,14 ff. 63 See the discussion in Roymans 2004, chapter 2; Gerrit- 27 personal, semi-sacred bond of loyalty to their leader, who in turn offered protection and regular rewards to his followers. According to historical sources, the comitatus had become a key social institution. Especially among the Lower Rhine Germani, these comitatus were a highly developed phenomenon in the Late Iron Age, as is apparent from the prominent role played by horsemen in warfare and raids. Caesar reports that the Eburonean leader Ambiorix was constandy accompanied by a retinue of equites when he fled from the pursuing Roman troops.67 He also mentions a raid into Eburonean territory by 2000 Sugambrian horsemen, who gave Caesar a nasty surprise when they attacked a Roman army camp;68 they were probably the comitatus of a Sugambrian war leader, who is not mentioned by name. Elsewhere, Caesar refers to a raid carried out by a large group of Tencteri and Usipetes horsemen into the territory of the Ambivariti (probably a subtribe of the Eburones) along the Meuse; a smaller group of 800 horse men is said to have stayed behind in their homeland.69 Caesar himself was personally acquainted with the qualities of the Germanic comitatus. His retinue included a band of 400 Germanic horsemen who functioned as his bodyguard. Somewhere at the start of his Gallic campaigns, Caesar must have received this cavalry escort from an unspecified war leader probably of Ubian or Chatto-Batavian origin.70The historical information points to a marked rise in the significance of horsemanship among the Lower Rhine Germani in the Late Iron Age. It is not easy to find direct archaeological evidence for the rise of the comitates in the Late Iron Age. Nevertheless, there are several clues for the Lower Rhine region. Firstly, there is archaeozoological evidence from the Rhine/Meuse delta, which shows that horses were kept in almost all Late Iron Age settlements.71 If we assume that horsemen operated in groups rather than individually, this increase in the use of horses may reflect the growing significance of the comitatus system. Secondly, we can point to the exceptional length of Late LaTene swords in this region (and in a wider area too), which suggests that they were used as cavalry swords. Thirdly, there is the introduction o f gold in the form of coins and torques from the mid-2nd century BC. Gold opened up new opportunities for individuals to build up positions o f power. Torques were kept and worn to demonstrate a leader’s status, while coins could be distributed as a means of shaping reciprocal relations between leaders and followers. Coins represented ‘portable and transferable symbols of authority’ and marked the involvement of individuals in supralocal power networks.72 Leaders could only maintain their comitatus by giving regular rewards to their followers.This occurred to such an extent in Central Gaul that it led to the creation of full-time bands of professional warriors in the service of noble leaders.73This was less the case among the Lower Rhine Germani, where we should think in terms of followers operating in comitatus networks on a part-time basis. Here there were also cavalry formations among tribes who did not mint coins, such as the Sugambri, Tencteri and Usipetes. These mobile groups of horsemen were deployed in small-scale raids on hostile neighbouring tribes. Leaders could use the spoils acquired in this way, especially in the form of cattle,74 to present to their followers. Particularly in the Lower Rhine region, with its strong pastoral tradition, cattle and horses will have been important gifts in the political military sphere.75 And yet there also arose in this region the practice of using high-value coins in such exchange circuits in the Late La Tene period. It seems likely that there was a direct fink between the emergence of new forms of authority, based on command of a comitatus, and the increasing use of gold in the form of coins and jewellery. Also important 67 Caesar, BG 6.30; 6.43. 72 Creighton 2000,31. 68 Caesar, B G 6.35. 73 Roymans 1990, 40. 69 Caesar, BG 4.12. 74 Cf. Caesar, B G 6.35, on a raid into Eburonean territory 70 Speidel 1994,12 ff. 71 Roymans 1996, Table 5. Horse bones are surprisingly rare, on the other hand, in settlements from the Early and Middle Iron Age. 28 by Sugambrian horsemen who were intent on accumu lating as m uch cattle as possible. 75 Roymans 1996,47. is the relationship with the appearance of regional sanctuaries in the Eburonean area from LaTène D1 onwards. The deposition of weaponry at cult places of this kind suggests a connection with the domain of warfare and raiding. Assuming a close relationship between politics and religion in the public cult, the rise of regional sanctuaries can be linked to the process of social hiérarchisation described above. Cult places perhaps offered new rituals and symbols through which social relations could be redefined and institutionalised as relations of inequality.76 I 6 C O N C L U S I O N S What then are the key findings of the studies presented in this volume? They can be summed up in the following points: 1. The hoards under discussion belong to the same circulation period, which we have labelled the Fraire/Amby horizon and which was concentrated in the 50s BC. 2. Three gold hoards appear to have been buried at off-site locations that were used only once for the deposition of valuables, while at Thuin the hoards had been buried in a Late Iron Age fortification and a nearby cult place. 3. The coin hoards represent a high intrinsic value and their owners can be regarded as members of an upper social echelon. 4. The study of the coin hoards has meant a major adjustment of the chronological framework for Late Iron Age gold emissions de type beige. We can now distinguish an early and a late group. The production and circulation of epsilon staters of the Nervii and Scheers 31 staters of the Eburones are concentrated in Haselgrove’s stage 4 and more specifically the 50s BG, while the Scheers 24 uniface staters of class I-III appear to be somewhat older. 5. The silver rainbow cups of the Lith group circulated from the early 50s BC and are therefore slighdy older than has previously been accepted.They should probably be regarded as an Eburonean emission. 6. We can identify two political centres within Eburonean territory in the m id-lst century BC - the Tongres/Maastricht region in the south and the Kessel/Lith region in the north. 7. Given their late dating, the coin hoards from the Fraire/Amby horizon can be considered within the historical context of the Caesarian conquests.They therefore represent key documents for the history and archaeology of the Low Countries, marking the phase in which the expanding Roman empire violently subjugated the tribes living there. 8 . The gold hoards of Thuin, whose burials were probably contemporaneous, were located at a Late Iron Age fortification which - as indicated by the presence of numerous lead sling bullets - was once besieged by the Roman army. This is possibly the oppidum of the Aduatuci, which was captured by Caesar in 57 BC. 9. The commonly applied simple dichotomy between profane Versteckdepots and ritual hoards is not very productive when it comes to reasons for burying coin hoards. We propose that a distinction be made between the reasons for burial and the form that this took. Both appear to be strongly governed by crisis situations, especially warfare. 9. The method used in this volume for developing a comprehensive seriation of coin dies for particular coinages has generated important new insights and deserves to be followed up. We wish to end with a general comment on historical interpretations of Late Iron Age gold hoards. There is a long tradition within archaeology, and Celtic numismatics in particular, of finking the burial of hoard finds to historically documented wars and migrations. One of the best known examples is the gold hoard of Tayac in southwest France, which is considered one of the key chronological beacons of Celtic 76 Derks 1998,183. 29 numismatics. This hoard was long regarded as the war chest of the roaming Cimbri at the end of the 2nd century BC.77 For several decades criticism has been justly levelled at such historical associations, which are often insufficiently substantiated and sometimes prove to be completely anachronistic. At the same time attention has shifted to alternative —and in particular ritual §§ interpretations of hoards. Neverthe less, it is important that we remain open to historical interpretations. For instance, it is still plausible that the violent Roman conquest of especially the central and northeastern zones of Northern Gaul had a major impact on native gold circulation. This brief phase of barely a decade corresponds to a surge in the production of gold coins, a peak in the burial of coin hoards and ultimately to the virtual disappearance of native gold circulation. This theme has been explored in more detail in publications by Haselgrove (1984) and Roymans (1990) on the late gold emissions of Northern Gaul. However, Haselgrove’s (1999) plea for an extended chronology of these Northern Gallic gold emissions put paid to the old association with the Roman conquest. With this publication, we wish to show that the earlier dating is not tenable for a number of coin series and that the model of a direct link between the gold depositions and the Roman conquest is more attractive than ever. R E F E R E N C E S Andrikopoulou-Strack, N., 2001: Eburonen —und was dann?, in G. Brands et al. (eds), Rom und die Pro vinzen. Gedenkschriftfür Hanns Gabelmann, Bonn (Beihefte Bonner Jahrbücher, 53), 163-172. Bazelmans, J., 1999: By weapons made worthy. Lords, retainers and their relationship in Beowulf, Amsterdam (Amsterdam Archaeological Studies 5). Bechert,T , 1982: Römisches Germanien zwichen Rhein und Maas, München. Bievelet, C.H., 1967: L’exploration archéologique de Bavai: four de verrier, ornements muraux et moules à flans monétaires, Revue du Nord 49, 623-635. Bonenfant, P./E. Huysecom, 1982: Retranchements préhistoriques à Thuin (Hainaut), Campagne de fouilles 1981, Annales d’Histoires de l’Art et d‘Archéologie 4 , 103-113. Boucly.J.L., 1966: Une « monnaie » à Bavai, Revue du Nord 191, 545-555. Brulet, R., 2008: Les Romains en Wallonie, Bruxelles. Chapman, J., 2000: Fragmentation in archaeology. People, places, and broken objects, London. Creemers, G./S. Scheers, 2007: Wichtige Fundstücke als Zeugnisse der Eburonen in Nordostbelgien, in Krieg und Frieden. Kelten, Römer, Germanen, Bonn, 169-174. Creighton, J., 2000: Coins and power in Late Iron Age Britain, Cambridge (New Studies in Archaeology). De Jersey, Ph., 2009: Some experiments in Iron Age coin production and some implications for the pro duction of Gallo-Belgic E, in J. Van Heesch/I. Heeren (eds), Coinage in the Iron Age. Essays in honour of Simone Scheers, London, 257-269. De Laet, S.J., 1961: Les limites des cités des Ménapiens et des Morins, Helinium 1, 20-34. Delestree, L-, 1993: Les monnaies gauloises du sanctuaire d’Estrées-Saint-Denis (Oise), Revue archéologique de Picardie 1-2, 35-40. Delmaire, R ., 1983: Corpus des trésors monétaires antiques de la France, II. Nord - Pas-de-Calais, Paris. Delmaire, R ., et al., 1994: Carte archéologique de la Gaule 62. Le Pas-de-Calais, Paris. Delmaire, R., et al., 1996: Carte archéologique de la Gaul 59. Le Nord, Paris. Delmaire, R ., 1997: Corpus des trésors monétaires antiques de la France,VIII/2. Picardie, Paris. Dengis,J.-L., 2010: Trouvailles et trésors monétaires en Belgique. VI. Les monnaies gauloises, Wetteren (Collec tion Moneta 114). 77 Kellner 1970. For other examples, see Van Impe et at. 1997/1998,105-106; Fitzpatrick 2005,161-162. 30 Derks,T., 1998: Gods, temples and rituai practices. The transformation of religions ideas and values in Roman Gaul, Amsterdam (Amsterdam Archaeological Studies 2). Deru, X., 2009: Cadres géographiques du territoire des Nerviens, Revue du Nord 91, no. 383,181-201. Dhénin, M., 1996: Les monnaies gauloises de la collection Fontaine, in R . Delmaire/L. Notte (eds), Trouvailles archéologiques dans la région de Bapaume. Prospections etfouilles d’Edmond Fontaine (1926-1987), Arras (Mémoires de la Commission départementale d’Histoire et d’Archéologie du Pas-de-Calais 32), 249-258. Dourif, (Dr), 1880: Note sur des objets antiques découverts à La Sauvetat. Lue à l’Académie de Clermont-Ferrand, le 3 juin 1880, Clermont-Ferrand, 7-8, pl. 1. Doyen,J.-M., 1982: U n statère d’or nervien trouvé aux Bons-Villers (Luttre-Liberchies, Hainaut), Bulletin du Cercle d’études numismatique, 19,1-4. Doyen, J.-M., 1984: U n statère d’or gaulois découvert à Liberchies-Bons-Villers (province de Flainaut), Bulletin du Cercle d’études numismatiques 21, 4-6. Doyen, J.-M., 1987: Les subdivisions ‘aux segments de cercles’ du type BN 8030: état de la question, in Mélanges offerts au Docteurf.-B. Colbert de Beaulieu, Paris, 315-330. Doyen, J.-M., 1990: Monnaies inédites de la collection Stroobants : Ciney, Fontaine-Valmont, Mons et Taviers, Amphora 59, 64. Doyen, J.-M., 2005: Monnaies gauloises de sites et trésors monétaires de Gaule Belgique. Trouvailles isolées et dépôts (1995-2005) (première partie), Bulletin du Cercle d’Etudes Numismatiques 42,157-187. Doyen, J.-M., 2009: Les monnaies gauloises du sanctuaire de Fontaine-Valmont (Hainaut, Belgique). Essai de synthèse, in J. Van Heesch/I. Heeren (eds), Coinage in the Iron Age, Essays in Honour of Simone Scheers, London, 85-97. DRSAC 1870 - Documents et Rapports de la Société paléontologique et archéologique de l’arrondissement judiciaire de Charleroi, 3, 51. Dufirasnes, J., 1999: Quelques objets, datant de la préhistoire à la période moderne, découverts dans les déblais du canal à Pommeroeul, Vie archéologique 52,29-53. Faucher, Th./F. Téreygeol/L. Brousseau/A. Arles, 2009: A la recherches des ateliers monétaires grecs: l’apport de l’expérimentation, Revue Numismatique 165,43-80. Fischer, B./C. Gruel, 2001: Catalogue des monnaies gauloises, in M. Reddé/S. von Schnurbein (eds), Alésia. Fouilles et recherchesfranco-allemandes sur les travaux militaires romains autour du Mont-Auxois (19911997), 2 (Mémoires de l’Académie des Inscriptions et Belles-Lettres 22), 21-39, pl. 1-28. Fitzpatrick.A., 2005: Gifts for the golden gods. Iron Age hoards of torques and coins, in C. Haselgrove/D. Wigg-Wolf (eds), Iron Age coinage and ritual practices, Mainz (Studien zu Fundmünzen der Antike, 20), 157-182. Fontaine, E ./R . Dhenin/M . Dhenin, 1973: Monnaies et pierres gravées antiques trouvées àVaulx-Vraucourt, Bulletin de la Commission départementale du Pas-de-Calais 9, 221-234. Gerritsen, F./N. Roymans, 2006: Central places and the construction of tribal identities.The case of the Late Iron Age Lower Rhine région, in C. Haselgrove (ed.), Celtes et Gaulois. L’Archéologieface à l histoire. Les mutations de lafin de l’Age du Fer, Glux-en-Glenne (Collection Bribracte 12/4), 251-266. Gôbel, J./A. Hartmann/H.E. Joachim/V. Zedelius, 1991: Der spâtkeltische Goldschatz von Niederzier, BonnerJahrbücher 191, 27-84. Graff, Y./Lenoir/J. Warnimont 1983: Les monnaies gauloises de Liberchies (Bons-Villers), V. Romana Contact, 1981-1982,23, no. 188. Gricourt, D./A.Jaques, 2002: Monnaies antiques découvertes à Saint-Blangy, in R . Delmaire et al, Chro nique numismatique (XXI), Revue du Nord 84,181. Gricourt, D./A. Jacques, 2009: Les monnayages tardifs des Atrébates, témoins de la destinée éphémère d’un royaume celto-romaine, in J. van Heesch/I. Heeren (eds), Coinage in the Iron Age. Essays in honour of Simone Scheers, London, 123-153. 31 Gricourt, D./A. Jaques/G. Prilaux, 2003: Quelques figurations zoomorphes des Dioscures celtes. À pro pos d’une série inédite de petits bronzes atrébates « aux segments de cercles (type BN 8030), Cahiers numismatiques 40, 155, 23-37. Haselgrove, C., 1984: Warfare and its aftermath as reflected in the precious metal coinage of Belgic Gaul, Oxford Journal ofArchaeology 3, 81-105. Haselgrove, C., 1999: The development of Iron Age coinage in Belgic Gaul, The Numismatic Chronicle 159,111-168. Haselgrove, C., 2005: Introduction, in C. Haselgrove/D.Wigg-Wolf (eds), Iron Age coinage and ritual prac tices, Mainz (Studien zu Fundmünzen der Antike, 20), 9-22. Haselgrove, C., 2012: Beyond Hallaton, inV. Score (ed.): Hoards, hounds and helmets. A conquest period ritual site at Hallaton, Leicestershire (Leicester Archaeological Monograph, 21). Heinrichs, J., 1999: Zur Verwicklung Ubischer Gruppen in den Ambiorix-Aufstand d.J. 54 v.Chr. Eburonische und ubische Münzen im Hortfund Fraire-2, Zeitschriftfür Papyrologie und Epigraphik 127, 275-293. Heinrichs,J., 2009: Eburonisches Erbe. Sch. 217 I (a) zwischen Tungern und Sunukern, in J.Van Heesch/I. Heeren (eds), Coinage in the Iron Age. Essays in honour of Simone Scheers, London, 187-219. Huysecom, E., 1981: Dépôt monétaire celtique àThuin (Hainaut), Archéologie 1, 24-25. Huysecom, E./J. Wargnies/H.G. Bachmann, 1987: Le dépôt de statères celtiques unifaces —type Scheers 24 —du “Mont d’O r” (comm. De Leuze-en-Hainaut), Archaeologia Belgica 3, 101-116. Joachim, H.-E., 2007: Die späte Eisenzeit am Niederrhein, in Krieg und Frieden. Kelten, Römer, Germanen, Bonn, 48-58. Kellner,, H.-J., 1970: Der Fund vonTayac, ein Zeugnis des Cimbrenzuges 1,Jahrbuch fur Numismatik und Geldgeschichte 20,13-47. Kopytoff, J 1986: The cultural biography of things: commoditization as process, in A. Appadurai (ed.), The social life of things. Commodities in cultural perspective, Cambridge, 64-91. Lagadec, J.-P./A. Lieger, 1998: La circulation monétaire celtique en Lorraine. Catalogues des communes ayant livré des monnaies gauloises dans les départements de Meurthe-et-Moselle, Meuse, Moselle et des Vosges, Archaeologia Moseïlana 3, 9-60. Laval, P., 1999: Section à ornières sur la route romaine Arlon-Tongres, Vie archéologique, 52,27-53. Loscheider, R ., 1998: Untersuchungen zum spâtlatènezeitlichen Münzwesen desTrevererlandes, Archaeo logia Moseïlana 3, 61-226. Lund, A., 1998: Die ersten Germanen. Ethnizität und Ethnogenese, Heidelberg. Magermans K./S. Saerens, 2005:Trouvailles dans le vicus romain de Kester (Brabant Flamand): la collec tion Rudy Van Cutsem, Bulletin du Cercle d’études numismatiques 42, 118-121. Mignot, Ph., 1982: Statère nervien de Romerée (Doische, prov. de Namur), Revue belge de numismatique 128,223-224. Nick, M., 2005: Am Ende des Regenbogens... Ein Interpretationsversuch von Hortfunden mit keltischen Goldmünzen, in C. Haselgrove/D.Wigg-Wolf (eds), Iron Age coinage and ritual practices, Mainz (Studien zu Fundmünzen der Antike, 20), 115-155. Nouwen, R., 1997: Tongeren en het land van de Tungri (31 v.Chr.-284 n.Cht), Leeuwarden/Mechelen (Maaslandse Monografieën, 59). Pichon, Bl., 2002: Carte archéologique de la Gaule. 2. L ’Aisne, Paris. Poux, M., 2008: L’empreinte du militaire tardo-républicain dans les faciès mobiliers de La Tène finale, in M. Poux (ed.), Sur les traces de César. Militaria tarde-républicains en contexte gaulois, Glux-en-Glenne (Collection Bibracte, 14), 299-432. Raepsaet-Charlier, M.-Th., 1994: La cité des Tongres sous le Haut-Empire: problème de géographie historique, BonnerJahrbücher 194, 43-59. Rossez, C., 2010: Occupation du sol et circulation monétaire de la fin de l’Age du Fer à la fin de l’Anti quité dans la région de Merbes-le-Château (Hainaut, Belgique), Bulletin du Cercle d’études numisma- 32 tiques, 47, 277-308. Roymans, N., 1990: Tribal societies in Northern Gaul, An anthropological perspective, Amsterdam (Cingula 12). Roymans, N., 1996: The sword or the plough. Regional dynamics in the romanisation of Belgic Gaul and the Rhineland area, in N. Roymans (ed.), From the sword to the plough. Three studies on the earliest romanisation of Northern Gaul, Amsterdam (Amsterdam Archaeological Studies 1), 9-126. Roymans, N., 2004: Ethnic identity and imperial power. The Batavians in the Early Roman Empire, Amsterdam (Amsterdam Archaeological Studies, 10). Roymans, N., 2007: O n the latènisation of Late Iron Age material culture in the Lower Rhine/Meuse area, in S. Möllers/W. Schlüter/S. Sievers (eds), Keltische Einflüsse im nördlichen Mitteleuropa während der mittleren und jüngeren vorrömischen Eisenzeit, Bonn, 311-325. Roymans, N., 2007: Understanding social change in the Late Iron Age Lower Rhine region, in C. Haselgrove/T. Moore (eds), The Later Iron Age in Britain and beyond, Oxford, 478-491. Roymans, N./T. Derks (eds), 1994: De tempel van Empel. Een Herculesheiligdom in het woongebied van de Bataven, 's-Hertogenbosch. Roymans, N./J. Aarts, 2005: Coins, soldiers and the Batavian Hercules cult. Coin deposition at the sanc tuary of Empel in the Lower Rhine region, in C. Haselgrove/D.Wigg-Wolf (eds), Iron Age coinage and ritual practices, Mainz (Studien zu Fundmünzen der Antike, 20), 337-359. Roymans, N./J.Äarts, 2009: Coin use in a dynamic frontier region. Late Iron Age coinages in the Lower Rhine area,Journal ofArchaeology in the Low Countries 1, 5-26. Roymans, N./L.Verniers, 2009: Glazen La Tène-armbanden in het gebied van de Nederrijn.Typologie, chronologie en sociale interpretatie, Archeobrief 13, no. 4, 22-31. Roymans, N./W. Dijkman, 2010: De Keltische goud- en zilverschat vanAmby (gern. Maastricht), Utrecht. Scheers, S., 1977: Traité de numismatique celtique II. La Gaule belgique, Paris. Scheers, S., 1990a: Le monnayage, in G. Leman-Delerive/ Beaussart (eds), Les Celtes en France du Nord et en Belgique, VIe-Ier siècle avant J.-C.,Valenciennes-Bruxelles, 144-152. Scheers, S., 1990b: Les monnaies du trésor de Fransnes-lez-Buissenal, in G. Leman-Delerive/ Beaussart (eds), Les Celtes en France du Nord et en Belgique, Vie-Ier siècle avant J.-C.,Valenciennes-Bruxelles, 159. Scheers, S., 1994: Keltische munten, in M. Rogge/L. Beeckmans (eds), Geld uit degrond. Tweeduizend jaar muntgeschiedenis in Zuid-Oost-Vlaanderen, Zottegem, 19-36. Scheers, S., 1995: Lés statères bifaces du type Lummen-Niederzier. U n monnaiage Eburon antérieur à la conquête Romaine, in M. Lodewijckx (ed.), Archaeological and historical aspects ofWest-European societies. Album amicorum André van Doorselaer, Leuven (Acta Archaeologica Lovaniensia, 8), 87-94. Scheers, S., 1996: Frappe et circulation monétaire sur le territoire de la future civitas Tungrorum, Revue belge de numismatique et de sigillographie 142, 5-51. Scheers, S./J. van Heesch/R. van Laere, 1991: Uitmuntend goud, Brussel. Scheers, S./J. van Heesch, 1994: Grez-Doiceau: monnaies gauloises et romaines, Chronique de l’Archéologie wallonne, 2,12-13. Scheers, S./G. Creemers 2002: U n trésor monétaire des Éburons à Heers (Limbourg, Belgique), Lunula. Archaeologia protohistorica X, Bruxelles, 80-85. Schlott, Chr., 1999: Zum Ende des spâtlatènezeitlichen Oppidum auf dem Dünsberg, Montagnac. Schulze-Förster, J., 2002 : Die latènezeitliche Funde vom Dünsberg, Marburg (Ph.D. dissertation Universität Marburg). Score, V, 2006: Rituals, hoards and helmets: a ceremonial meeting place of the Corieltavi, Trans. Leice stershire Archaeological and Historical Society 80,197-207. Score,V., (ed.), 2012: Hoards, hounds and helmets. A conquest period ritual site at Hallaton, Leicestershire (Leices ter Archaeological Monograph, 21). Severs, L., 1980: Trouvailles de monnaies gauloises, Bulletin du Cercle d’études numismatiques, 17, 3, 68-71. Severs, L., 1980: Statère d’or nervien à Basse-Wavre, occupation pré-romaine au Champ de l’Hosté ?, 33 Amphora, 21, 37-39. Reprinted in Amphora 59, 4-6. Sicherl, B., 2009: Namenlose Stämme. Nordwestdeutschland am Vorabend der römischen Okkupation, in S. Berke/M. Zelle/M. Tatter (eds), 2000Jahre Varusschlacht. Mythos, Stuttgart, 43-57. Sills, J., 2003: Gaulish and British gold coinage, London. Speidel, M., 1994: Riding for Caesar. The Roman emperor’s horse guard, London. Stead, I.M., 1991:The Snettisham treasure: excavations in 1990, Antiquity 65, 447-464. Thirion, M., 1962: Le trésor de Fraire, monnaies gauloises en potin, Revue beige de numismatique 108, 67-109. Thirion, M., 1976: Une monnaie en argent des Nerviens, Bulletin du Cercle d’études numismatiques 13, 54-59. Turquin, P., 1955: La bataille de la Seile (du Sabis) en 1’an 57 avant J.-C., Les Etudes Classiques 23, 2,113154. Vanderhoeven, A./M. Vanderhoeven 2004: Confrontation in archaeology. Aspects of Roman military presence in Tongeren, in F.Vermeulen/K. Sas/W. Dhaeze (eds), Archaeology in confrontation. Aspects of Roman military presence in the Northwest, Ghent (Archaeological Reports Ghent University 2), 143-154. Van Es,W.A., 1981: De Romeinen in Nederland, Haarlem. Van Heesch,J., 1998: De muntcirculatie tijdens de Romeinse tijd in het noordwesten van Gallia Belgica. De civitates van de Nerviërs en de Menapiërs (ca. 50 V.C.-450 n.C.), Brussel (Koninklijke Musea voor Kunst en Geschiedenis. Monografie van Nationale Archeologie). Van Heesch, J., 2005: Celtic coins and religious deposits in Belgium, in C. Haselgrove/D. Wigg-Wolf (eds), Iron Age coinage and ritual practices, Mainz (Studiën zu Fundmünzen der Antike, 20), 247-263. Van Impe, L., et al. 1997/1998: De Keltische goudschat van Beringen (prov. Limburg), Archeologie in Vlaanderen 6,1-124. Verhoeven, M.F. 2011: Een aanvullende archeologische evaluatie en waardering van het plateau van Caestert (Riemst, provincie Limburg),Weesp (RAAP-rapport 2162). Vermeulen, E, 1992: Tussen Leie en Schelde. Archeologische inventarisatie en studie van de Romeinse bewoning in het zuiden van de Vlaamse Zandstreek, Gent (Archeologische Inventaris Vlaanderen, Buitengewone reeks 1). Volling.Th., 1990: Funditores im römischen Heer, Saalburg-Jahrbuch 45, 24-58. Weiller, R . 1983: Fundmünzen der Römischen Zeit in Luxemburg, III, Luxemburg. Woimant, G.-P, 1995: Carte archéologique de la Gaule. 60. L’Oise, Paris. Zieghaus, B., 1995: Der Münzfund von Grossbissendotf. Eine numismatisch-historische Untersuchung zu den spätkeltische Goldprägungen in Südbayern, Munich (Ausstellungskatalog Prähistorische Staatssammlung München 27). 34 A P P E N D I X G O L D I. S E R I AT I O N STATERS H O A R D S OF T H E OF T H E TABLE FOR E B U R O N E S , F R A I R E / a C O I N CLASS DIES I, OF F R O M M B Y H O R I Z O N . Heers Amby Orp-le-Grand - - - - - - THE T HE O = obverse die; R. = reverse die. Fraire Philippeville Class la P2 0 1- R 1 0 2—R1 F 1,2 0 2 -R 2 - P3 - H1 A 20 0 3 -R 4 - - - A5 - 0 4 -R 5 - - - A 10 - 0 5 - R6 F3 - - - 0 5 -R 7 - P1 - A 117 - H2 A2 H 3,4,5,6 A 18,9 0 3 -R 3 0 6 -R 8 - 0 6 -R 9 - 0 7 - R9 - H7 - - 0 8 -R 1 0 F4 P4 H8 - - 0 8 - R 11 - - - A4 - 0 9 - R 12 - A3 - 0 10 - R 13 H9 - - H 10 A 6,7 - H 11,12 - - 0 10 — R 14 - - 0 11 - R 15 - 0 12 — R 16 - - H 13 - 0 1 2 - R 17 - - H 14 - 0 12 - R 18 - - H 15 - - 0 13 - R 18 - - - A 15 - - H 16 0 13 — R 19 - - 0 13-R 20 - - A8 - 0 1 3 - R 21 - H 17 - - 0 1 4 - R 18 - H 18 - - 0 1 4 - R 22 - - H 19 - - 0 1 4 - R 23 - - H 20 - - 0 1 5 - R 24 - - H 21 A 17 - 0 1 5 - R 25 - - H 22 0 1 6 - R 26 - - H 23 A 19 - 0 1 7 - R 27 - - 0 18 - R 28 - - H 24 A 11 01 0 1 8 - R 29 - - - A 16 - - H 25-116 A 12-14 - A 21-30,119 _ 0 18 - R 30 A1 - Class lb 0 19 - R 31 0 20 - R 32 - - - A 31-35,115 - 0 21 - R 33 - - - A 36-38 - 35 A P P E N D I X 2. G O L D E P S I L O N FR O M THE S E R I A T I O N STATERS H O A R D S OF TABLE OF THE THE FOR C O I N N E R V I I FRA IRE / DIES OF (S C H E E R S A M B Y THE TYPE 29) H O R I Z O N . O = o b v e rs e d ie ; R = re ve rse d ie . Fraire Thuin-1 Thuin-2 Thuin-3 Phi I ippeville Heers C lass la 0 1- R 1 F9 0 2 -R 1 - T 1-1 0 4 -R 2 - TI-2 0 4 -R 3 - 0 5 -R 4 - 0 6 -R 5 - 0 3 —R1 O6- R 6 H 117 - - - TIII-1 - - - TII-1 - - - TI-3 - - - - - p5 - 0 6 -R 7 - TI-4 - TIII-2 0 7 -R 8 - TI-5 - - - - 0 7 -R 9 - TI-6 - - - - 0 8 -R 1 0 F10 - - - - - 0 8 -R 1 1 F 11,12 - - - - - 0 9 —R 12 F13 - - - - - 0 9 —R 13 - - T 11-2 - - - 0 9 -R 1 4 F 14,15 0 10 —R15 - - - - - - - T 1-14,15 - - - - 0 10 —R 16 - T 1-16 - 0 10 —R 17 - - - - - 0 11 - R 18 F 16 - - - 011 - R 19 F 17-19 - - 0 11 - R 20 - - T 11-3 - T 11-4 0 12- R 21 H 118 - - - - - - 013-R 22 F 20 - - - - - 0 13-R23 - TI-22 T 11-5 - - - 0 14- R 23 F 21 - - - - - 0 15- R 24 - - T 11-6 - - TI-18 - - - - 0 15-R25 0 15-R26 - T 1-19 - TIII-3 - - 015-R 27 - - T 11-7 - - - 0 15-R28 TI-20 - - - 0 15-R 29 F 22 - - - - - 0 15-R 30 - T 1-21 - - - - 0 1 6 - R 31 - TI-9 - 0 16-R32 - T 1-10 - - 0 16-R33 F 23 - - - - 0 17 - R 34 - TI-17 - - 018-R 35 F 24 T 1-11,12 - - 0 18-R36 - 36 T 11-8 - - - _ . Fraire 0 19- R 36 Thuin-1 Thuin-2 Thuin-3 _ F 25 - 0 20 - R 37 0 21 - R 37 F 26 0 21 - R 38 - 0 2 2 - R 39 - TII-9 - T 11-10 - TI-23 - - TI-24 - - - - TI-25 - - 0 23 - R 42 - - T 11-11,12 - 0 2 4 - R 43 - T 1-13 T 11-13 - T 11-14 T 11-15,16 0 25 - R 44 Heers _ - 0 22 - R 41 0 22 - R 40 Philippevlile - - H 119 ip ia - 0 26 - R 45 F 27 - 0 26 - R 46 F 28 - 0 27 - R 47 F 29 TI-8 0 28 - R 48 F 30 0 29 - R 49 F 31,32 - 0 3 0 - R 50 F 33 - - - - 0 31 - R 51 F 34 - - - - - TI-7 - - - - 0 3 2 - R 52 - - - - - H 120 - - - - - 0 33 - R 53 - - T 11-17 - - - 0 34 - R 54 - - T 11-18 - - - 0 3 5 - R 55 - - - TIII-4 - - 0 3 6 - R 56 - - - TIII-5 - - 0 37 - R 57 - - T 11-19 - - 0 38 - R 58 - - T II-20 - - 0 38 - R 59 F 35 - - - 0 39 - R 60 F 36 - - - - - 0 40 - R 61 - T 1-41 - - - - 0 41 - R 62 F 37,38 - T 11-21 - - 0 4 2 - R 63 - H 121 0 43 - R 64 - TI-26 - - H 122 0 4 4 - R 65 - - TIII-6 - - _ _ H 123 - - ■ - - H 124 Class Ib 0 45 - R 66 F 39 0 45 - R 67 F 40 0 45 - R 68 0 46 - R 68 - T II-22 - . m - - - - TI-27 ■ F 41 TI-28 - - - 0 46 - R 69 - - TII-23 - - - 0 47 - R 70 F 42 - - - - - F43 - - - 0 4 8 - R 71 F 44 TI-42 0 49 - R 72 - - TII-24 - - 0 50 - R 73 - TI-36 - - 0 50 - R 74 0 48- R 70 - - TII-25 - - 0 50 - R 75 - T I-37-39 TII-26 - - 0 50 - R 76 F 45 TI-40 TII-27 0 5 0 - R 77 F 46 - TII-28 - - 37 Fraire 0 50 - R 78 F 47 Thuin-1 Thuin-2 Thuin-3 Philippeville - - - - - Heers - 0 50 - R 79 F 48 - 0 51 - R 80 F 49 T 1-30-32 T II-29 - 0 51- R 81 F 50 - - - - - 0 51 - R 82 - T 1-33 - - - 0 52 - R 83 F 51,52 T 1-34,35 T II-30 - - 0 5 2 - R 84 F 53 - - 0 5 2 - R 85 F 54 - 0 53 - R 86 - T 1-29 - 0 5 4 - R 87 F 55 - - - - 0 55 - R 88 - - - - - - - TIII-7 " - Class lia 0 56 - R 89 F 56-60 T 1-47 T II 31,32 - - 0 56 - R 90 F 61-65 T 1-48-51 T II-33,34 TIII-8 - - 0 5 6 - R 91 - - - - P-9 - 0 5 6 - R 92 - - TII-35 - - - 0 57 - R 93 F 66,67 TII-36 - - 0 57 - R 94 F 68 T II-37 - 0 5 7 - R 95 - - - 0 57 - R 96 F 69 - 0 58 - R 97 F 70 - - 0 58 - R 98 - T 1-43 TII-38 0 58 - R 99 F 71,72 T 1-44 0 5 9 - R 100 F 73,74 0 59-R101 F 75 T 1-54 0 5 9 - R 102 F 76 T 1-55 0 60 — R 103 F 77 T 1-45,46 - TIII-9 H 125 - - - - P-10 - - - - - - - T 111-10 - - - - - - T 1-52 - - - - - 0 6 0 - R 104 F 78 T 1-53 - 0 61- R 105 F 79 - TII-39 - 0 61- R 106 - - TII-40 - - - T 11-41,42 - - - - - - T 111-11 - - - T 111-12 - - - 0 62 — R 107 0 63-R108 - T 1-56 0 6 3 - R 109? - - 0 6 4 - R 110? - - - T 1-57 - - P-6 T 1-58 T II-43 - - T 1-59-61 Tl 1-44,45 TIII-13 - - Class Mb 0 65 - R 111 0 65 — R 112 0 66 — R 112 F 80-82 0 66- R 113 T I-62-64 T 11-46 T 111-14 - 0 67-R112 F 83 - - - - 0 68 - R 114 F 84,85 T 1-67-71 T II- 47,48 T 111-15 0 6 8 - R 115 - TI-73 - 0 68 - R 116 - - T 11-49 - 0 6 8 - R 117 F 86 TI-72 - - - 0 6 9 - R 118 F 87 - T 11-50 0 6 9 - R 119 F 88-90 TI-65,66 T 11-51,52 T 111-16-20 P-7,8 38 - Fraire 0 6 9 - R 120 F 91 0 69- R 121 F 92,93 0 70- R 122 F 94 0 71- R 123 - 0 7 2 - R 124 Thuin-1 Thuin-2 Thuin-3 - - - Philippeville Heers _ - - - - - - H 126 - - - - H 127 - 0 72 - R 125 - - - 0 7 3 - R 126 - - - - H 129,130 0 7 3 - R 127 - - - - H 131-133 0 7 3 - R 128 - - - - 0 7 4 - R 129 - - - 0 7 5 - R 130 - - - - H 138 0 7 5 - R 131 - - - H 139-147 0 7 5 - R 132 - - - H 148-150 H 128 H 134-136 H 137 0 7 5 - R 133 - - - - - H 151,152 0 7 5 - R 134 - - - - - H 153,154 39 APPENDIX 3 . LIST OF ASCRIBED TO EBURONES. 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. 29. 30. 31. 32. 33. 34. 35. 36. 37. 40 THE SITES OF SCHEERS 3 I STATERS, Empel, De Werf. Cult place. 23 examples; 10 class la, 2? class lb and 11 class II. Roymans 2004, Appendix 4.2. Rossum, St.Andries. Context unknown. Class la. Roymans 2004, Appendix 4.2. Meteren, Zes Morgen (= Geldermalsen, Honsgemet). Rural settlement. Two pieces, both class la. Roymans 2004, Appendix 4.2. Deil, Hooiveld? Cult place? Two pieces, both class la. Roymans 2004, Appendix 4.2. Est,Tieflaar. Rural settlement. Class la. Roymans 2004, Appendix 4.2. Randwijk, Hokkerden. Rural settlement.Two pieces, class la (plated bronze) and lb. Roymans 2004, Appendix 4.2. Nijmegen, Hunerberg. Roman cam Class lb (bronze). Roymans 2004, Appendix 4.2. Houten, Schalkwijkseweg. Rural setdement. Class la. Plated bronze.Vos 2009,191, fig. 5.5. ‘Betuwe’. Exact location unknown. Class la. See www.muntenbodemvondsten.nl no. 29260. Oirschot, Kasteren. Rural setdement. Class la. Roymans 2004, Appendix 4.2. Hapert, Hoogpoort. Hoard/cult place? Class II. Roymans 2004, Appendix 4.2. Bergeijk, Enderakkers. Rural setdement. Class la; 5.68 g. Roymans 2004, Appendix 4.2. Weert, Veldbeemd. Rural setdement. Class la. Roymans 2004, Appendix 4.2. Thorn, R o o f tile factory. Rural setdement. Class II. Roymans 2004, Appendix 4.2. Antwerp, Class I. Scheers 1977, 441. Asse, vicinity. Class la. Scheers 1977, 442. Asse, Roman settlement. 2 class II. Doyen, 2005 177; Dengis 2010,19-20. Brussels, vicinity. Scheers 1977, 442; Dengis 2010, 33. Rotselaar. Rural setdement. Class la. Scheers 1996, fig. 2; Dengis 2010, 114. Leuven. Class II. Scheers 1977, 442. Molembeek-Wersbeek. Class la. Scheers 1996, fig. 2; Jean Elsen,Vente 58, 1999, no.734; Dengis 2010,96 (Molenbeek-Wersbeek) and 128 (Tienen,Testelt). Jesseren. Class la. Unpublished. Detector find D. Hermans 2009. Inf. S. Scheers. Rijkhoven, Kleine Spouwen. Class lb; 5.57. Unpublished. Detector find Luc Meyers 1998. Inf. S. Scheers. Bolder, Riemst. Class la. Unpublished. Coll. B. Emons. Inf. Tim Vanderbeken, ZOLAD, Riemst municipality. Inden, Geuenich.Two pieces, both class la, from Merovingian burial. Roymans 2004, Appendix 4.2. Amby, Heihof. Class II, 4.96 gr. Unpublished. Coll. Derks. Maastricht, Randwyck. Rural settlement. Class II. Roymans 2004, Appendix 4.2. Maastricht, Amby, Hagenstraat. Hoard find. 40 pieces; 20 class la and 19 class Ib.This publication. Tongres, Berg. 10 pieces: 6 class la; 4 class II. Settlement and/or cult place? Scheers/Creemers, this volume, Appendix 2. Tongres? Class I. Bronze. Scheers et al. 1991, 35; Roymans 2004, Appendix 4.2. Tongres, vicinity. Class I. Scheers et al. 1991, 36; Roymans 2004, Appendix 4.2. Heers. Hoard find. 82 pieces, all class la. Scheers/Creemers, this volume. Wange. Class la. Unpublished. Inf. M. Billiau. Unpublished. Waremme.Voie de Nivelle. Class lb; 5.40 gr. Unpublished. Detector find D. Hermans 2006. Orp-le-Grand,Jauche. Class la. Hoard find? Dengis 2010, 72.Van Impe/Scheers, this volume. Petit-Hallet. Cult place. Class II. Scheers 1996, fig. 2; Dengis 2010,108. Braives. Settlement? 11 pieces: 2 class la; 1 class lb; 7 class II. Scheers 1977, 442; Scheers 1996, fig. 2; Roymans 2004, Appendix 4.2. Some unpublished. 38. Liberchies. Nucleated settlement. 3 pieces: 1 class la (bronze), 2 class II (1 gold, 1 bronze). Scheers 1977, 442; Scheers 1996, fig. 2. Roymans 2004, Appendix 4.2. 39. Namur. Bed of river Meuse. Class la. Scheers 1997, 443; Dengis 2010,100. 40. Vervoz. Nucleated settlement. Class II. Scheers et al. 1991, 37; Dengis 2010, 37. 41. Haulchin. Merovingian cemetery. Class la. Roymans 2004, Appendix 4.2. Dengis 2010, 66. 42. Fontaine-Valmont. Cult place. Class la. Scheers 1996, fig. 2. Roymans 2004, Appendix 4.2. Dengis 2010,51. 43. Fraire. Hoard find. 4 pieces, all class la. Scheers 1984. Scheers/Creemers, this volume. 44. Marche-en-Famenne. Class II. Roymans 2004, Appendix 4.2. 45. Boviolles (F). Oppidum. Class la. Roymans 2004, Appendix 4.2. 46. ‘Limburg’ (N). Class II. See www.muntenbodemvondsten.nl no. 27085. 47. ‘Departement des Ardennes’ (F). Class lb. Unpublished. Inf. S. Scheers. 48. Best, Hagelaarweg. Class la. Isolated detector find 2010. No other finds have been reported from this site. Unpublished. Inf. N. Arts, Eindhoven. 49. Beaumont, Thirimont. 2 pieces; 1 class lb and 1 class II. Dengis 2010, 52 (mentions erroneously Fontaine-Valmont as findplace). 50. Riemst, Maastrichter Steenweg. Unpublished Class la. Inf. Tim Vanderbeken, ZOLAD, Riemst municipality. 51. Heers. Direction of Oreye. Class la; 5.61 g. Doyen 2005,183. Coll. Delvaux, Brussels. Dengis 2010, 67. 52. Horpmaal, Keiberg. Class II; 5.28 g. Unpublished. Detector find Luc Meyers 2003. Inf. G. Creemers, Tongeren. 53. Neerharen. Class II; 5.3 g. Unpublished. Detector find Luc Meyers 2005. Inf. G. Creemers, Ton geren. 54. Vechmaal (Heers). Class II. Roman villa. Unpublished detector find Bart Smets. Coll. Bart Smets. 55. Thuin, Bois de Luiseul. Class la. Isolated detector find. Coll. S. Scheers. Scheers/Creemers, this volume. 56. Thuin, Bois du Grand Bon Dieu. Isolated find from Late Iron Age fortification by R . Fauville. Class la. Scheers/Creemers/Roymans/Van Impe, this volume. 57. Borgloon, Mellenstraat. Class II. Unpublished. Detector find Jan Helsen (Borgloon). Coll.J. Helsen. 58. Vechmaal (Heers), Rigoosberg. Class II. Roman rural settlement. Scheers/Creemers, this volume, Appendix 2, no. 26. 59. Roclenge-sur-Geer. Class II. Roman villa. Scheers/Creemers, this volume, Appendix 2, no. 27. 60. Vicinity of Philippeville. Hoard find. 4 pieces, all class la. Scheers, this volume. 61. Kester, Gooik, Rom an vicus. 2 class II. Magerman, Saerens 2005,118; Dengis 2010, 75. 62. Tielt-Winge, Gempemolen. Class II. Coll. S. Scheers. Unpublished. 63. Kemexhe. Roman settlement. Class II (bronze). Dengis 2010, 74. 64. Aiseau-Presle. Class II. Doyen 2005,158; Dengis 2010,14. J. Eisen, 97,13-9-2008, no. 7 65. Nimy. Class II. Inf. J.-M. Doyen. Unpublished. 66. Vermand (F). 1 class la (bronze) and 1 class II (bronze). Münz Zentrum Köln, 68,25. 67. Hapert, Steenakkers. Rural settlement. Class la. Unpublished. Detector findWil Biezen 2011. Coll. Biezen. 41 APPENDIX I - 111 , 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 4 . LIST ASCRIBED TO OF SITES THE SCHEERS 29 STATERS, CLASS Castricum, Oosterbuurt. Found during excavation in a posthole of a native-Roman house. Roymans 2004, 51. Houten, Dwarsdijk. Class I. Unpublished detector find by Kees Leenheer. Aalst (N), Eendenkade. Rural settlement. Found on field by H. de Graaflf, Heusden. Roymans 2004, 51. Ressen, Kerkenhof. Class I. Roymans 2004, 51. Found on field by Henk Houterman in 1977. Gent (bronze). Class I. Unpublished detector find. Information prof. Jos Benders, Nijmegen. Doesburg. Roymans 2004, 51. Empel (4 coins, 2 bronze; 3 class I, 1 class unknown). Roymans/Aarts 2005. Acht. Class I. Isolated detector find near Late Iron Age/native-Roman cemetery. Inf. Nico Arts, Eindhoven. Roymans 2004, 51. Deurne. Class III. Scheers 1977, 403; Dengis 2010, 79, no. 79. Strijtem, 1906, in a hoard o f 12 staters, 6 staters class III. Scheers 1977,403, 899; idem 1994,30-31; Van Heesch 1998, 292; Dengis 2010,123, no. 354. Ternat, class I.Van Heesch 1998, 292; Dengis 2010,124, no. 357 (wrongly class III). Grez-Doiceau, class III (bronze). Inf. F.Wouters; Scheers 1996,27, note 58; Dengis 2010,60, no. 129. Unpublished. Jodoigne, class I. Inf. F. Wouters. Scheers 1996, 27, no. 58 and 60; Dengis 2010, 73, no. 177. Waterloo, vicinity (Liberchies?), class II. Oxford coin catalogue CCI 00.1088. Wavre, class I. Severs 1980a, 70-71; Severs 1980b, 37-39; Severs 1990, 4-6; Dengis 2010,139, no. 401. “land van Aalst”, class I. Scheers 1977, 403; Scheers 1994, 30;Van Heesch 1998, 225; Dengis 2010, 13, no. 2. Dendermonde, class II.Van Heesch 1998, 241; Dengis 2010, 41, no. 78 (wrongly class III). Kwaremont, 1816, hoard, anepigraphic staters and staters withVIROS: Scheers 1977, 404, 887; Scheers 1994, 31;Van Heesch 1998, 263; Dengis 2010, 77, no. 193. Lokeren, 1936, class I. Scheers 1977, 404;Van Heesch 1998, 266; Dengis 2010, 88, no. 216. Moerbeke-Waas, class II. Scheers 1977, 404; Van Heesch 1998, 271; Dengis 2010, 95, no. 248. 78 The following Belgian fmdspots have not been retained: Another identical stater, weighing only 5,41g, was pre - Baarle-Hertog, cited by Dengis 2010, 23, no. 29, refers sented at J. Elsen, list 88, april 1986, no. 9. B oth coins are to Bar-le-Duc, France, and concerns an eye-stater o f the forgeries: they have the same oval outline, show a break Remi. at exactly the same place while the types are identical. - Braives (Dengis J.-L. 2010, 32, no. 50 after Scheers - Herderen, 1989 (Dengis 2010, 69, no. 159) the find- 1996, 27, note 58 ).There has been a regrettable confu place is given in the J. Elsen, Auction catalogue 17, sion with the Eburones staters with triskeles. 17-11-2990, no. 8. The coin is identical to the stater of - Frasnes-les-Buissenal (Dengis 2010, 55, no. 133) con the hoard o f Fraire no. 86 and has the same weight. The tained only uniface staters; cf. Scheers 1990b. findspot must be wrong and has not been retained. We - Kermt, cf. Scheers 1996, 27, note 58; Roymans 2004, also have not retained some French findspots: Aizy-Jouy 155; Dengis 2010,74, no. 181.The coin is identical with (Aisne), cf. Delmaire 1997, 57; Pichon 2002, 90, no. 8. the stater no. 23 o f the hoard o f Fraire. It apparently is a The short description “horse, crescent and wheel” also counterfeit with a low weight (5,76 g), although it reput fits a Suessiones stater. edly was found about 1975 on an old school ground. 42 OF N E R V 11 . 78 21. Meilegem, class III. Scheers 1994, 30; Van Heesch 1998, 269; Dengis 2010, 93, no. 237. 22. Oudenaarde, vicinity, class I, class II in the Scheldt; 3 anepigraphic staters in the neighbourhood . Scheers 1977, 403; Scheers 1994, 30;Van Heesch 1998,277; Dengis 2010, 106, nos 287-288. 23. Velzeke, Ruddershove, class I (bronze). MuseumVelzeke. Unpublished. 24. Amougies, hoard (?), one known ex. class III: Scheers, 1977, 404, 873; Van Heesch 1998, 225; Dengis 2010,15, no. 09. 25. Blicquy,Ville d’Anderlecht, 2 class II (bronze) one from the sanctuary, the other from the Roman setdement, 1 class III (bronze). Scheers 1994, 32; Van Heesch 1998, 234; Dengis 2010, 27, no. 42. Another class III bronze stater probably from the sanctuary. Unpublished. 26. Elouges, several staters (hoard?) in the cemetery of Monceau.Van Heesch 1998, 245. 27. Fontaine-Valmont, Les Sarrasins. 3 class I, 1 class II, 1 class III. Some unpublished. Doyen 2009, 90 (3 class I, II and III); Dengis 2010, 50, no. 105 (1 class I). 28. La Buissiere (hoard, doubtful), 1 class III.Van Heesch 1998, 263-264; Dengis 2010, 78, no. 196. 29. Liberchies, 1 class I, 1 class II, 1 class III. Scheers 1981, 2-3; Doyen 1982, 1-4; Doyen 1984, 4-6; Dengis 2010, 81, 83, no. 208. 30. Monceau-sur-Sambre, prob. class I. DRSAC 3, 1870, 51, note; Rossez 2010, 283 (mentions mis takenly 2 staters). 31. Nimy, 1 class III. Inf. J.-M. Doyen; Dengis 2010,102, no. 273. 32. Nouvelles, Roman villa, 1 stater class I or II (plated).Van Heesch 1998,275; Dengis 2010,103, no. 276; Rossez 2010, 283. 33. Peissant, 1867 hoard, 6 anepigraphic staters: Scheers 1977,404, 892;Van Heesch 1998, 279; Dengis 2010,107, no. 290. 34. Pommeroeul, vicus, class Ill.Van Heesch 1998,282; Dufrasnes 1999,51; Dengis 2010,109, no. 295 (mentions mistakenly 2 staters). 35. Thuin, hoard Thuin-1, Bois du Grand Bon Dieu; 73 staters, 42 class I and 31 class II. Huysecom 1981, 24-25;Van Heesch 2005, 263; Dengis 2010, 125, no. 359. See this book. 36. Thuin, mixed hoard Thuin-2, Bois de Luiseul. 52 staters, 30 class I and 22 class II, 1 quarter stater class l.Van Heesch 2005, 263. See this book. 37. Thuin, hoard Thuin-3, Bois de Luiseul, 20 staters, 7 class I and 13 class II. Van Heesch 2005, 263; Dengis 2010,125, no. 361 (mentions only 5 staters); Rossez 2010, 283 (mentions only 5 staters). See this book. 38. Thuin, Bois de Luiseul: class II isolated stater. Coll. Scheers. See this book. 39. Thuin, class III (bronze). Doyen 2005,185; Dengis 2010,126, no. 359; Rossez 2010, 283. 40. Thuin, unknown location, class I. Inf.T. Decraeker. Unpublished. 41. Tournai, class I. Unpublished. 42. Prov. Hainaut, class I, class III.Thirion 1962,107; Dengis 2010, 61, no. 132 (mentions only 1 stater). 43. Oleye, vicinity, class I. Scheers 1977, 404; Dengis 2010,104, no. 279. 44. Berg, 2 class I. Dengis 2010,24, no. 34 = 130, no. 369 (Tongeren) (1 stater).The second one unpub lished. See this book, appendix 2, nos. 7 and 8. See Scheers/Creemers, this volume, 147, appendix 2. 45. Heers, mixed hoard, 38 staters, 8 class I and 30 class II. Dengis 2010, 68, no. 155 (composition of nos. 7 and 8 to be revised). See this book. 46. Sint-Truiden, vicinity, 1 class I. Scheers 1977,404; Scheers 1996,27, note 58; Dengis 2010,120, no. 338. 47. La Roche-en-Ardenne, class III. Laval 1999, 51; Dengis 2010, 78, no. 195. 48. Ciney, vicinity, class I. Doyen 1990, 64; Dengis 2010, 36, no. 63 (wrongly class III). 49. Fraire, mixed hoard, 87 staters, 47 class I and 39 class II, I unknown. Scheers 1984; Scheers 1990a, 147, 264, 265; Dengis 2010, 54, no. 111/1-2. See this book. 43 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. B0. 81. 82. 83. 84. 85. 86. 44 Romerée, vicinity, class I. Mignot 1982, 223-224; Doyen 1990, 64; Dengis 2010,113, no. 309. Yves-Gomezée, class II. Scheers 1977, 405; Dengis 2010,141, no. 406. Philipville, vicinity, mixed hoard;, 6 staters, 1 class I, 5 class II. See this volume. Holnon, class II. Scheers 1977, 400; Pichon 2002, 256. Pommiers, class I (bronze). Unpublished. Vermand, vicinity, hoard, 18 staters class I. Inf. H. Pion andTh. Decraeker. Unpublished. Rugles, class I. Scheers 1977, 400. Montboissier (Eure-et-Loire). Scheers 1977, 401. Boviolles (Meuse), class I. Musée Epinal; Lagadec/Liéger 1998, 39. Avesnes-sur-Helpe, vicinity, 1 anepigraphic stater. Scheers 1977, 402 ; Delmaire 1996, 107; Van Heesch 1998, 309. Bavai, in the sand-pits, at least 3 anepigraphic staters. Scheers 1977,402;Van Heesch 1998, 313. Cagnoncles (or Carnières or Rieux); class III. Delmaire 1996,155;Van Heesch 1998, 318. Cambrai, vicinity, class II. Delmaire 1996,167. Condé-sur-l’Escaut, Bois de Bonsecours, ca. 1975; class I. Delmaire 1996, 153; Van Heesch 1998, 320. Crochte, hoard (doubtfiil), before 1870. Some anepigraphic staters. Delmaire 1983, 28; Delmaire 1996,162;Van Heesch 1998, 320. Douai (?); class I (doubtfiil). Unpublished. Douai, vicinity (?), class II. In f T. Decraeker. Unpublished. Escaudin, 1843, some anepigraphic staters. Delmaire 1996,199. Flaumont-Waudrechies (?), Catelet or Camp d’Avesnelles, hoard, class I.Van Heesch 1998,324. The hoard reputedly consisted of three uniface staters, but the Museum of Avesnes displays two uniface staters together with the cast of a Nervii stater. Flines-les Mortagne, hoard, 14 staters (1 class III). Scheers 1977,402 (wrongly situated at Rouillon); Delmaire 1996, 249;Van Heesch 1998, 324. Flines-lez-Râches, Mer de Flines, before 1872, at least 1 anepigraphic stater. Delmaire 1996, 250. Ledringhem, mixed hoard, 1852, large number of class III and IV staters: Scheers 1977, 402, 888; Delmaire 1996, 338;Van Heesch 1998, 327. Mostly unpublished. Maubeuge, class I. Scheers 1977, 402; Delmaire 1996, 338; Van Heesch 1998, 329; Rossez 2010, 282. Phalempin, 1981, class I. Delmaire 1996, 353;Van Heesch 1998, 33. Rieux-en-Cambresis, class I. Unpublished (see also Cagnoncles). Seclin , class I, class III. Scheers 1977, 402; Delmaire 1996, 406;Van Heesch 1998, 334. Valenciennes, faubourg de la Porte de Famars, 1835,1 anepigraphic stater: Delmaire 1996, 419. Viesly, class I: Scheers 1977, 402; Delmaire 1996, 425 (mentions mistakenly a uniface stater);Van Heesch 1998, 337. Dept. Nord, class III. Unpublished. Beauvais, vicinity, class II. Scheers 1977, 402; Woimant 1995,158-159. Estrées-Saint-Denis, Le Moulin des Hayes, sanctuary, class I (bronze). Delestrée 1993, 38-39; Woimant 1995, 454. Senlis, vicinity, class II. Scheers 1977, 402. Aubigny-en-Artois, hoard, only class II: Scheers 1977,402,874; Delmaire 1983, 62; Delmaire 1994, 161 (mentions more types, but the source is doubtful). Hesdin, class III (bronze). Inf.Th. Decraeker. Unpublished. Saint-Omer, vicinity, class I. Scheers 1977, 402; Delmaire 1994, 566. Saint-Venant, 2 anepigraphic staters. Delmaire 1983, 80; Delmaire 1994, 402. Thérouanne, vicinity, 1 class 1,1 bronze stater. Scheers 1977, 403; Delmaire 1994, 86. 87. 88. 89. 90. 91. 92. 91. 94. 95. 96. 97. 98. Dept. Pas-de-Calais, class III. Unpublished. La Sauvetat (Puy-de Dome), class I. Dourif 1880, 7-8. Dieppe, arr. (Seine-Maritime), 2 anepigraphic staters. Scheers 1977, 403. Hiermont, 1866 class II. Scheers 1977, 403. Toutencourt, class I, class II. Scheers 1977, 403. dept. Somme, 1 anepigraphic stater. Scheers 1977,403. between Cologne and Bonn, class I. Scheers 1977, 405. between Emmerich and Kleef. Unpublished. Biel (CH), vicinity, class I. Scheers 1977, 405. Nyon (CH), vicinity. Scheers 1977,405. Silchester (UK), class I (bronze). Reading Museum 00200; Scheers 1977, 405. county Sussex (UK), class I. Glendining 20-11-1961, no. 25 (Doubleday Sale); Oxford Celtic Coin Index 61 0377. 45 APPENDIX STATERS, 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. 46 5 . LIST ASCRIBED OF TO SITES THE OF SCHEERS I 5 2 -V QUARTER NERVII. Blicquy, Roman vicus and sanctuary; 2 class I.Van Heesch 1968, 223. Fontaine-Valmont, site des Castellains, Roman sanctuary; 33+, class I. Doyen 2005,157-159, nos 2-10 (9 coins, one with wheel variant). Unpublished (24 coins). Ghlin, Long Coron, 1 class I.Van Heesch 1996, 250; Dengis 2010, 58, no. 120. Liberchies, BonsVillers, Roman vicus, 1 class I. Graff/Lenoir/Warnimont 1983,23, no. 188 ; Doyen 1987, 322, class XVI, no. 6. Mons, vicinity, Mont Eribus, 1 class I. Doyen 1987, 325. Thuin, Bois de Luiseul, sanctuary ?, hoard 2,1 class I. Doyen 2005,185; Dengis 2010,215, no. 360; Rossez 2010, 283. See this volume. Thuin, Bois de Luiseul, sanctuary ?, 1 class 1,1 class II. Coll. Scheers. See this volume. Kruishoutem, Kapellekouter, Roman vicus and sanctuary, 4 class 1 ,1 class II, 1 class III. Vermeulen 1992, 73-74 (4 of 6 coins); Van Heesch 1998, 262 ( 4 of 6 coins). Velzeke, Ruddershove, Roman vicus, 1 class I.Van Heesch 1998, 301 (only 1 of this type); Dengis 2010,134, no. 378 (only 1 of this type). Fraire, hoard, 9 class 1,1 class II. Doyen 1987, 322, class XIV, 1, class XVI, 1-4; Dengis 2010, 54, no. 111. See this volume. Namur, river Sambre, 1 class I. Mahieu 1905,215-216; Scheers 1977, 647, no. 13; Doyen 1987,322, class XVI , no. 7; Dengis 2010,100, no. 263. Philippeville, vicinity, hoard, 3 class 1,1 class III. See this volume. Berg (?), 1 class I. Scheers/Van Heesch/Van Laere 1991, nos 16-17; Dengis 2010, 130, no. 369. Scheers/Creemers, this volume, 149, Appendix 2, no. 67. Titelberg, oppidum, 3 class I.Weiller 1983, 319, no. 228/345-347, pi. XI, 345. Saint-Thomas, oppidum Vieux-Laon, 2 class I. Unpublished. Vermand, Le Calvaire, Roman vicus, 4 class 1,1 class III. Unpublished. Villeneuve-Saint-Germain, oppidum, 2 class II. Unpublished. Vichy (?); 1 class I. Doyen 1987, 382, class XVI, 5. Alésia, Roman camp C, 1 class I. Fischer/Gruel 2001, 36, no. 655. Bavay, 1 class I. Biévelet 1967, 623-624; Boucly 1966, 545-555;Thirion 1976, 54-59. Bouvines, 1 class I. Paris, BnF, Cabinet des Médailles. Documentation A. Barthélemy. Lettre de Rigaux fils, 22 avril 1872 (drawing). Chemy, 1 class I. Unpublished. Viesly, 1 class I. Unpublished. Saint-Laurent-Blangy, on the site of a Roman fort, 2 class I. Gricourt/Jacques 2002,179-182; Gricourt/Jacques/Prilaux 2003, 23-37; Gricourt/Jacques 2009,136-138,148. Vaulx-Vraucourt,Voie Jacqueline ; 1 class I. Fontaine/Dhénin/Dhénin 1973, 222, no. 2; Delmaire et al. 1994, 330, no. 303/3; Dhénin 1996, 251, no. 13 ; Gricourt/Jacques 2009,136-138,148. T he gold hoard o f Fraire Simone Scheers / Guido Creemers 1 2 2.1 2.2 2.3 2.4 2.5 3 Discovery of the hoard and context of the site Composition of the hoard Gold staters of the Eburones Silver rainbow staters Gold staters of the Nervii Quarter staters of the Nervii The gold bracelet Dating of the hoard Summary References Appendix 1. Descriptive list of the coins from the Fraire hoard Plates I D I S C O V E R Y OF T H E H O A R D A N D C O N T E X T OF THE SITE In early September 1984 the first author (Simone Scheers) received a telephone call from Jean Elsen, director of a Brussels auction house, saying that he had purchased a Celtic gold hoard. He asked me to identify it and to provide a brief description for the auction catalogue; in return, I would be given an opportunity to examine the find in the auction house. According to Elsen, the hoard had been found in the Belgian municipality of Fraire, scattered across just a few square metres in a wood on the slopes o f a stream valley. No traces o f a container were found. Initially, 35 Nervian staters and a bracelet were discovered, probably with the aid o f a metal detector. Later, a second and then a third group of coins were found, 69 coins in total, spread across a somewhat larger area. For the rest, no details have been provided about the precise find site, nor the name of the finder and seller.1The find was sold at auction in Brussels and is nowadays dispersed over a small number of private and public collections in Belgium and abroad. Achille Trotin (Valenciennes), who later followed up several oral accounts in Fraire, concluded that the gold hoard must have been found east o f La Maroquette and west of or close to the western foothills of the Bois des Minières (fig. I).12 It seems that the hoard had been found by an employee of the ‘ancienne scierie’ (sawmill) during an afternoon walk between the mill and the ‘ancienne verrerie’ (glassworks). The coins then allegedly ended up in the hands of the mill owner, who sold them in Brussels. This account is fairly consistent with the information given at the time by Jean Elsen. It would mean that the gold hoard of Fraire had been found on the sloping banks of the Jaune stream. The paucity of context data points to a one-off deposition at an off-site location. 1 Scheers 1984. 2 Personal letter addressed to Simone Scheers, 11 May 2010 . 47 ~j j I La navette 1 km| Bois des Minières 3o Fig. 1.Topography o f Fraire, with approximate indication o f the find spot o f the gold hoard (1) and o f the hoard o f potin coins (2). 3 isolated finds o f Celtic coins; 4 trajectory o f R om an road; 5 Gallo-Roman/M erovingian cemetery at La M aroquette’; 6 ‘ancienne scierie’; 7 *ancienne verrerie’ The municipality of Fraire (formerly Fraire La Grande) is situated in the Belgian province of Namur, at the centre of the area between the Sambre and Meuse rivers. Two significant areas of high ground - the plateaus of Philippeville (approx. 300 m) and Fraire (approx. 270 m) occupy the central part of this area. The latter is intersected by the valley of the Fairoul river.3 Fraire is situated in an area o f fairly rich agricultural land, in a transition zone between fertile loess soils to the west of the village and stony, loamy, forested soils atop limestone and schist to the north and south of the village. The district is rich in traces of historical iron mines and quarries. Still today, most topographical maps feature places like Bois des Minières and Carrière. In the 1970s most quarries were abandoned and filled in. We also find various crayats des sarrasins - iron slag heaps o f indeterminate age - which were exploited once again at the end of the 19th century. As a result they have now largely disappeared and are preserved only in the local toponymy. It is not clear when iron mining began, but it probably goes back to the LaTène period. The Roman road north of Fraire is pardy constructed of iron slag. The etymology of Fraire - Ferrarias (AD 868), Ferrières (AD 1092, 1147) - also points to iron extraction.4The region around Fraire was situated in the territory of the Aduatuci, or possibly the Nervii, during the Late Iron Age.5 In Roman times it belonged to the civitas Tungrorum.6 Evrard 1997,5. ■ chapter on the coin hoards ofT huin and the synthesis by http ;//users.swing.be/ tor-4601/ curhisto.htm 5 De Laet 1961; Faider-Feytmans 1952; Mariën 1980, 34, and 472. See also the discussion in this volume in the 48 Roymans/Scheers. 6 Raepsaet-Charlier 1994; Scheers 1996,41. In about the mid-19th century there was intense archaeological activity in the municipalities of Fraire and Fairoul, which unearthed a Rom an and a Merovingian cemetery, as well as traces of iron extraction (including at La Maroquette) 7 In addition, Roman coins were found in the centre of Fraire in 1883 and in 1976.8The presence of remains from the Roman and Merovingian periods is without doubt linked to the trajectory of a major Roman road, that of Bavay-Trier, just north of Fraire.9 Yet another Late Iron Age coin hoard is known from the Fraire area. Discovered around 1872, it contained 25 potin coins of the au rameau type (Scheers 190, class 4).10The chronology of these coins, traditionally ascribed to the Nervii, is still uncertain, but we do know that they were already in circula tion at the time of Caesar’s conquests.11There is some confusion about the exact location of the find site: the hoard is said to have been discovered close to the Merovingian cemetery (La Maroquette) east of the village.12 Going by the original labels for the coin hoard in the museum of Namur, Thirion concluded that the hoard came from Fairoul.13 However, this is inconsistent with the findings made in 1872-1877 (cf. earlier footnotes). In our view, there is sufficient cause to regard the environs of the cemetery of La Maroquette as the find site. For the rest, one or two similar potin coins are said to have been found at Les Minières (southeast of the village), and another at Fairoul, about 1.5 km southwest of Fraire.14 2 C O M P O S I T I O N OF T HE H O A R D The gold hoard under discussion comprises 105 coins and a gold bracelet. The inventory comprises 87 Nervian staters (class I and II), four Eburonean staters (class la), four silver rainbow staters, ten Nervian(?) quarter staters with circle segments, as well as the gold bracelet (fig. 2). 2.1 GOLD,STATERS OF THE EBURONES The four gold staters of the Scheers 31 type with triskeles (nos. 1-4) belong to class la.15The small num ber of Eburonean staters can be explained by the hoard’s location in the area between the Sambre and Meuse rivers, which is outside the normal circulation area for Eburonean coins. The four staters belong to the earliest emissions of class la. O n the reverse, they feature a circle with a dot below the horse. For later emissions, this was changed to a circle with a ring in the centre. Two of the coins are examples of the very earliest emissions, conspicuous in their fine, detailed engraving of a horse surrounded by a range of sm all ornaments. These ornaments usually disappeared in later emissions, a fact which, despite the small number o f coins, is significant for the hoard’s chronology. A bibliography in Knaepen-Leserenier 1970, 97-99. See 12 .Evrard 1997,10 and 15; http://users.swing.be/tor-4601/ also Del M armol 1884,429; De Loë s.d., 187, for a report curhisto.htm. The Merovingian cemetery La Maroquette o f the R om an cemetery; Del M armol 1876, 304 for the is situated just east o f the present-day centre near house Merovingian cemetery. no. 46, Rue La Maroquette. This was confirmed for us by Annales de la Société Archéologique de Namur 16,1883, 490; one o f the inhabitants during on-the-spot inquiries. Doyen 2007. 13 Thirion 1962. Brulet 1970; Brulet 2008, 64-66,74,76-77; Del Marmol 14 Evrard 1997,10; Oger 1900, 95; Scheers 1977,746, 882; 1875, specifically 8, 12-13; Evrard 1997, 13-14; Marien Annales de la Société archéologique de Namur 13,1875, 522. 1980, map 476-477. Cajot 1877, 208 also shows clearly that the hoard was Scheers 1977, 746; Thirion 1962;Van Heesch 2005, 259. found in the vicinity o f the *cimetière franc* during exca Cf. Scheers/Creem ers/Roymans/Van Impe, this volume, vations in 1872. The find site lay ‘très loin*from the stray 79, note 21, on a similar hoard o f potin coins fromThuin. coins o f Les Minières and Fairoul. 49 Fig. 2. Coin types represented in the Fraire hoard. Scale 2:1. 1 Scheers 29 stater (Nervii); 2 Scheers 31 stater (Eburones); 3 silver rainbow stater; 4 Scheers 152-V quarter stater (Nervii?) The weights are variable.Two staters weighing 5.63 and 5.58 g fall within the average weight range for the series (5.40 to 5.80 g).At 4.77 and 4.65 g, the other two coins are much lighter.This large varia tion could point to a careless adjustment of individual weights, but some examples may already have been plated at this earliest stage. Despite the weight differences, the die combinations —with several identical dies —prove that they were all struck in the same mint (see p. 35, Appendix 1). 2.2 SILVER RAINBOW STATERS Once again, the hoard contains only four examples of this coin type (nos. 5-8). They feature on the obverse a triskeles within a laurel wreath and on the reverse a configuration of circles framed by a zigzag line.1516Three coins, which bear no additional marks (Roymans, variant a), range in weight from 5.88 to 6.25 g. The fourth coin, which is somewhat less well preserved, weighs 5.92 g and has the additional mark of a small circle at the bottom right (Roymans, variant b).The reverse is a die match with four silver rainbow cups from the hoard of Maastricht-Amby.17 Here too, the small number of rainbow cups in the hoard can be explained by the location of the find site far outside the normal circulation area for these coins. 15 La Tour 1892, pi. XXXV 8859; Scheers 1977, pi. IX, 254. 16 Roymans 2004,72-73. 50 17 Roymans/Dijkman, this volume, 210, nos. 82-85. 2-3 GOLD STATERS OP THE NEEVII Given the find site’s proximity to Nervian territory, it is not surprising that the 87 Nervian staters of the Scheers 29 type make up about 83% of the hoard. One coin could not be documented. All 86 remaining staters belong to classes I and II (nos. 80-94).18We note that the class I staters (55%) are slighdy more numerous than those of the later class II (45%). These same relative proportions of classes I and II are found in the hoards ofThuin. A closer inspection of class I reveals that class la coins predominate (35%). This would suggest that class la, the earliest emission, was larger than the later one. Another salient fact is that although the hoard contains 15 class lib staters (nos. 80-94), the most recent group of staters of this class, characterised by a fine, compact style, is prominent in the hoard of Heers but absent from that of Fraire. The staters of Fraire are characterised by a good state of preservation. Worn coins, a feature of the Heers hoard, are absent. Although many die links have been established with coins from the other hoards, the large variation shows that these were circulation coins. Another possibility is that this part of the hoard constitutes a gift or payment from a central coin reserve. The weights exhibit individual fluctuations but for the most part fall within the average weight range for Nervian staters (5.80 to 6.10 g).This applies to both class I and class II. 2.4 QUARTER STATERS OF THE NERVII? These small coins of the Scheers 152 type, class V, form three subclasses.19The class I coins feature three parallel circle segments on the obverse, surmounted by a circle, terminating in a dot below, and with a C-shaped arc before them. The entire image is surrounded by zigzag fines. The reverse shows a horse to the left with a wide open mouth. There is a four-spoked wheel above the horse’s back and a horizontal C-shaped arc with a dot below.20The coins in class II feature an equivalent obverse. On the reverse, the horse is surrounded at the top, in front and below by circles with a dot; as in class I, the one at the bottom is a horizontal C with a dot.21The Fraire hoard contained nine coins from class I and one from class II. Although these small coins are called quarter staters, we cannot be certain of this. That is because they usually contain little gold, and are gold-plated or made entirely of bronze, resulting in variable weights. Only in exceptional instances do they weigh 1.30 g or slightly more. Four coins therefore make up a stater of 5.20 g, which is far below the average weight for staters (5.80 to 6.10 g).That they indeed served as subsidiary coins for the anepigraphic Nervian staters is confirmed by quarter staters bearing the inscription VIROS,22 which tie in with the staters with the same inscription.23 These small coins were barely known before the hoard of Fraire was discovered. They are now found regularly in the territories of the Nervii and their neighbours, including at Blicquy, Kruishoutem, Fontaine-Valmont, Liberchies and the hoard of Philippeville. La Tour 1892, pi. XXXV, 8760; Scheers 1977, pi. VIII, 20 Scheers 1977, 644, type 152, class V, pi. XIX, 537; 216. Doyen 1987, 315 and 328, class XVI. 21 Doyen 1987, 317 and 328, class XIV. See Roymans/Scheers, this volume, fig. 5. The subdivi sion into three subclasses is not yet found in Scheers’ 22 Doyen 1987, 315 and 328, class I. Traité o f 1977; coins o f subclass II and III were then as 23 Scheers 1977, type 29, class IV, pi.VIII, 219. yet unknown. 51 Fig. 3. Gold bracelet from the Fraire hoard (scale 1:1) and detail o f the decoration (scale 2:1). 2.5 THE GOLD BRACELET The open bracelet (no. 105) is formed by a 3-mmthin curved bar with a sbght thickening at the ends, which are 4 mm in diameter.24 It is o f solid gold and weighs 37.50 g. The internal diameter is 52 x 63 mm. The decoration is confined to two incised parallel lines on the thickened ends (fig. 3). This piece of jewellery remains a unique item with no known counterparts.25 Its presence in the coin hoard of Fraire means that the hoard can be categorised as a coin hoard with ornaments. Five such hoards have been found in Belgium — at Frasnes-lez-Buissenal,26 Beringen,27 Philippeville, Orp-le-Grand and Thuin (the buffer terminal of a gold torque).28 All instances involved torques or bracelets. 3 DATING OF THE HOARD As with the Heers hoard, the importance of this coin hoard lies in its mixed character. It was the first Belgian hoard containing a combination of Eburonean staters, silver rainbow cups and gold coins of the Nervii. The four Eburonean staters belong to the earliest emissions of class la. This early group is also present in the hoards of Heers and Maastricht-Amby. The strong association between the three hoards is evident in the die combinations, with many of the coins coming from the same dies.29The lack o f later Eburo nean staters of classes la and lb render an early dating for the hoard likely. In the light of the hoard of Maastricht-Amby, it is now possible to date the four rainbow cups with greater accuracy. Their presence in the hoard of Fraire allows us to place the beginning of minting quite early, in our view c. 60 BC.30 Their presence in the hoard of Maastricht-Amby and their distribution pattern give reason to believe that they were minted by a northern subgroup of the Eburones.31 The Scheers 29 staters of the Nervii belong to classes I and II, but the recent group of class lib staters, which makes up the bulk of the Nervian staters in the hoard of Heers, are absent altogether. The situa tion is therefore analogous to that of the Eburonean triskeles staters, of which only the earliest emissions O u r thanks to Luc Van Impe for this description. the gold finds from Thuin, see elsewhere in this volume. 25 Hautenauve 2005,273-279. 29 Cf. Roym ans/Scheers, this volume, 35, Appendix 1. 26 Hautenauve 2005,194-195; Cahen-Delhaye 1990, 153- 30 See the study on the hoard o f Maastricht-Amby and the 155; Scheers 1990,159. 27 Hautenauve 2005, 192-194; Van Impe et al. 1997/98, 9-132. 28 For the hoards o f Philippeville and Orp-le-Grand, and 52 general conclusions in Roymans/Scheers, this volume. 31 Cf. the contributions o f Roym ans/Dijkm an and R oy mans/Scheers in this volume. are represented. Only early Nervian staters occur, which here too suggests an older dating than for the hoard of Heers. Further, there is a striking resemblance to the Nervian staters from the three hoards from Thuin, which also contain only early staters and are therefore almost contemporaneous with the hoard of Fraire.32The exceptional state of preservation of the Nervian coins of Fraire proves that they were not long in circulation before being buried. We note with interest the presence of ten Scheers 152-V quarter staters in the hoard of Fraire. The gold hoards ofThuin-2 and Philippeville also contain these coins. Their presence in these hoards proves that they were minted at the same time as the staters. A class I coin was found in camp C at Alesia,33 thus yielding a terminus ante quern of 52 BC for the start of minting. Although the hoard of Fraire belongs to the same horizon, on the basis of the information presented here we suggest a somewhat earlier dating than the hoards of Heers and Maastricht-Amby. SUMMARY The significance o f the coin hoard of Fraire, with a total of 105 coins and a gold bracelet, lies in its mixed composition. The vast majority of the coins are Nervian gold staters, in association with Eburonean sta ters, local quarter staters and silver rainbow cups. This association of different coin types makes the hoard of Fraire an important reference point for research into the chronology of late gold emissions in North ern Gaul. An analysis of the coin dies of the Nervian staters from Fraire reveals a striking resemblance between this hoard and the three hoards from nearby Thuin; they must have been buried at virtually the same time.The synthesis elsewhere in this volume argues for a dating in the early 50s BC and a link with the initial phase of Caesar’s conquest of Northern Gaul. For the rest, we have scant information about the context of the coin hoard, but it was probably an off-site location on the slopes of a stream valley. REFERENCES Brulet, R ., 1970: Chaussée de Bavay à La Meuse, Archéologie 2, 68. Brulet, R ., 2008: Les Romains en Wallonie, Bruxelles. Cajot, F., 1877: Les Gorduni. Étude de géographie et de numismatique gallo-belge, Annales de la Société archéologique de Namur 14,1877,195-212. De Laet, S.J., 1961: Les limites des cités des Ménapiens et des Morins, Helinium 1, 20-34. Deles trée, L.-P./M.Tache, 2002: Nouvel Atlas des monnaies gauloises. I. De la Seine au Rhin, Saint-Germainen-Laye. Del.Marmol, E., 1873: Rapport sur la situation de la Société archéologique de Namur pendant l’année 1871-1872, Société' archéologique de Namur (Reprint), 267-276. Del Marmol, E., 1875: Route romaine de Bavay à La Meuse, Annales de la Société archéologique de Namur 13,1-21. Del Marmol, E., 1876: Rapport sur la situation de la Société archéologique de Namur pendant l’année 1875, Société archéologique de Namur (Reprint), 299-310. Del Marmol, E., 1884: Rapport sur la situation de la Société archéologique de Namur pendant l’année 1883, Société archéologique de Namur (Reprint), 429-449. De Loe A., s.d.: Forma Orbis Romani. Carte archéologique de la Belgique romaine, répertoire statistique et bibli ographique avec carte, s.l. 32 See Roymans/Scheers, this volume, 36 ff., Appendix 2. 33 Fischer/Gruel 2001, 36, no. 655, pl. 24. 53 Doyen, J.-M., 1987: Les subdivisions « aux segments de cercles » du type BN 8030 : état de la question, Mélanges offerts au DocteurJ.-B. Colbert de Beaulieu, Paris, 315-330. Doyen, J.M., 2007: Trouvailles, Bulletin du Cercle d’Etudes numismatiques, 44. Evrard, J., 1997: Fraire-la-Grande et Fairoul, s.l. (revised édition from Courrier de Philippeville, 1958). Faider-Feytmans, G., 1952: Les limites de la cité des Nerviens, L’Antiquité Classique 21, 338-352. Fischer, B./K. Gruel, 2001: Catalogue des monnaies gauloises, in M. Reddé/S. von Schnurbein (eds), Alésia, Fouilles et recherchesfranco-allemandes sur les travaux militaires romains autour du Mont-Auxois (19911997) 2. Le matériel, Paris (Mémoires de l’Académie des Inscriptions et Belles-Lettres 22). Hautenauve, H., 2005: Les torques d’or du second Age du Fer en Europe. Techniques, typologies et symbolique, Rennes (Travaux du Laboratoire d’Anthropologie Rennes 44). Knaepen-Lescrenier, A.M., 1970: Répertoire bibliographique des trouvailles archéologiques de la province de Namur, Bruxelles (Répertoires archéologiques Série A: Répertoires bibliographiques 9). La Tour, PL de, 1892: Atlas de monnaies gauloises, Paris. Leman-Delerive, G./E. Warmenbol 2007: Dépôts et sites cultuels en contexte ‘humide’ dans les civitates des Nerviens et des Atrebates, in Ph.Barral/A. Daubigney/C. Dunning/G. Kaenel/M.-J. RoulièreLambert (eds), L’âge du Fer dans l’arc jurassien et ses marges. Dépôts, lieux sacrés et territorialité à l’âge du Fer, Actes du 29e colloques international de l’AFEA Bienne, 5 —8 mai 2005, vol. 2, Besançon, 426-427. Mariën, M.E., 1980: Belgica Antiqua. De stempel van Rome, Antwerpen. Oger, A., 1900: Nos fouilles de 1897 à 1899, Annales de la Société archéologique de Namur 24, 75-98. Raepsaet-Charlier, M .Th., 1994: La cité des Tongres sous le Haut-Empire, BonnerJahrbücher 194, 43-59. Roymans, N., 2004: Ethnie Identity and Impérial Power. The Batavians in the Roman Empire, Amsterdam (Amsterdam Archaeological Studies 10). Scheers, S., 1977: Traité de numismatique Celtique II. La Gaule Belgique, Paris (Centre de Recherches d’Histoire Ancienne 24, Série Numismatique) . Scheers, S., 1984: Le trésor de Fraire-2 (1981 —1984),Jean Eisen —liste, 6-8. Scheers, S., 1996: Frappe et circulation monétaire sur le territoire de la future Civitas Tungrorum, Revue belge de Numismatique et de Sigillographie 142, 5-51. Thirion, M., 1962: Le trésor de Fraire. Monnaies gauloises en potin, Revue belge de Numismatique 108, 67-112. Van Heesch, J., 1998: De muntcirculatie tijdens de Romeinse tijd in het noordwesten van Gallia Belgica. De civitates van de Nerviërs en de Menapiërs (ca. 50 V.C.-450 n.C.), Brussel (Monografieën van Nationale Archeologie 11). Van Heesch, J., 2005: Celtic Coins and Religious Deposits in Belgium, in C. Haselgrove/D.Wigg (eds), Ritual and IronAge Coinage in North-West Europe, Berlin-Durham (Studiën zu Fundmünzen der Antike 20), 247-263. Van Impe, L./G. Creemers/R. Van Laere/S. Scheers/H. Wouters/B. Ziegaus, 1997-1998: De Keltische goudschat van Beringen (prov. Limbourg), Archeologie in Vlaanderen 6, 9-132. 54 APPENDIX i: DESCRIPTIVE FRAIRE HOARD. LIST OF THE COINS FROM THE The numbering corresponds to the numbering of the coins in Plates 1-10. O = obverse die; R. = reverse die. The hoard was sold by Jean Elsen auctioneers and is nowadays dispersed over a small number of public and private collections in Belgium and abroad.34 No. weight (g) die no. analysis % references Au Ag Cu g o ld staters type Scheers 31, class la, Eburones 1 4.77 0 2 -R 1 2 5.63 0 2- R1 3 4.66 0 5 -R 6 4 5.58 0 8 -R 1 0 Scheers 1984, no. 31 Scheers 1984, no. 32 44 33 23 Tongeren, Gallo-Rom. Museum PMP-R 6219 Brussels, National Bank s ilv e r ra in b o w staters. triquetrum type 5 6.25 variant a 6 6.09 variant a 7 5.88 variant a 8 5.92 variant b Scheers 1984, no. 33 10 85 5 Scheers 1984, no. 35; coll. Ferdy Willems Scheers 1984, no. 34 g o ld staters type Scheers 29, class la, N e rvii 6.04 0 1- R 1 Scheers 1984, no. 12 10 5.95 0 8 -R 1 0 Cabinet Albuquerque, Rouen, 23-1-1993, no. 141 11 5.96 08 12 5.97 0 8 -R 1 1 Scheers 1984, no. 2. Monnaies et Medailles, Bâle, 507, 1988, no. 51; Mon- 13 5.83 0 9 -R 1 2 Scheers 1984, no. 6. Elsen, 62, Dec. 1983, no.7; Elsen, 14,7-10-1989, 14 5.91 0 9 — R 14 15 5.85 0 9 - R14 16 5.92 0 11 - R 18 Scheers 1984, no. 15 17 5.96 0 11 - R 19 Scheers 1984, no. 4. Monnaies et Médailles, Bâle, 498, avril 1987, no. 33 18 5.90 0 11 - R 19 19 5.84 0 11t R 19 9 R 11 naies et Médailles, Bâle, 516,1988, no. 119 no. 14; Elsen, 66,16/17-6-2001, no. 112 Scheers 1984, no. 18. MünzZentrum, 13-11-1984, no. 1421 ; Lanz, 29-4-1985, no. 21; Crédit Suisse, 8,27/28-10-1987, no. 627 49 41 10 coll. F. Willems. Elsen, 41, février 1982, no. 58 Monnaies et Médailles Bâle, 498, avril 1987, no. 34; Monnaies et Médailles Bâle, 516, novembre/décembre 1988, no. 120; Auctiones AG, 1-10-1986, no. 316; Auctiones AG, 8-11-1990, no. 1208 20 5.89 0 1 3 -R 2 2 Cabinet Albuquerque, 27,18-4-1991, no. 214; 5,25-11-1987, no. 15; 3,25-5- 21 5.94 0 14 - R 23 Scheers 1984, no. 13; Argenor, 22-4-2004, no. 219 22 5.97 0 1 5 - R 29 Poinsignon, Strasbourg, 42, juin 1997, no. 9 23 5.94 0 1 6 - R 33 Elsen, 55, Mai 1983, no.5; Elsen, 88, avril 1986, no. 9 24 5.79 0 1 8 -R 35 Crédit de la Bourse, Paris, octobre 1992, no.61; Cabinet Albuquerque, 1987, no. 248. 34 W here applicable, the auction catalogues or auction cited for each coin, house lists that mention isolated items from the hoard are 55 Rouen, 11-12-1994, no. 129. Crédit de la Bourse, Paris, 24-11-1995, no. 117 25 6.16 0 19 - R 36 26 5.94 0 21 - R 37 27 6.02 0 26 - R 45 28 5.96 0 26 - R 46 29 5.97 0 27 - R 47 30 6.00 0 28 - R 48 Scheers 1984, no. 16. Cabinet Albuquerque, Rouen, 3, 25-5-1987, no. 247 Scheers 1984, no. 1 CGF Paris, 4,26-6-1998, no. 66. Burgan, 30-9-1994, no. 245; Vedrines/Poindessault, 14-12-1995, no. 297; Vedrines/Poindessault, 9-9-1998, no. 207 0 29 - R 49 Brewi, 7, augustus 1987, no. K 22 31 5.85 32 5.88 0 29 - R 49 Scheers 1984, no. 17. Weil, 15-1-2006, no. 200 33 5.90 0 30 - R 50 Vedrines/Poindessault, 5-6-1985, no. 311 Banque populaire du Nord, Paris, liste 35, mai 1985, no. 57 34 5.91 0 31 - R 51 35 5.91 0 38 - R 59 36 5.66 0 39 - R 60 37 5.83 0 41 - R 62 38 5.93 0 41 - R 62 Scheers 1984, no. 8 g o ld staters type Scheers 29, class lb, N e rvii 39 5.93 0 45 - R 66 Scheers 1984, no. 11 (0)-10 (R) 40 5.94 0 45 - R 67 coll. S. Scheers 41 5.85 0 46 - R 68 42 5.85 0 47 - R 70 43 6.00 0 48 - R 70 44 5.79 0 48 - R 71 0 50 - R 76 Elsen, 46,21-9-1996, no. 70 45 5.88 46 5.92 0 50 - R 77 Scheers 1984, no. 7 47 5.95 0 50 - R 78 Scheers 1984, no. 14 48 6.05 0 50 - R 79 Burgan, 22-12-1984, no. 10 49 5.89 0 51 - R 80 Banque populaire du Nord, 38, avril 1986, no. 46; Banque populaire du No 50 5.83 0 51 - R 81 Giessner Miinzhandlung, 3-3-1997, no. 19; Rauch, 19-4-1993, no. 24; Lanz, 51 5.99 0 52 - R 83 Scheers 1984, no. 10 (0)-11 (R) 52 5.92 0 52 - R 83 53 5.94 0 52 - R 84 54 5.98 0 52 - R 85 55 5.97 0 54 - R 87 Crédit de la Bourse, octobre 1986, no. 267 16-11-1984, no. 18 Elsen, 2,14-11-1984, no. 6 g o ld staters type Scheers 29, class lia, N e rvii 56 5.90 0 56 - R 89 57 5.87 0 56 - R 89 58 5.87 0 56 - R 89 59 5.81 0 56 - R 89 60 5.84 0 56 - R 89 61 5.93 0 56 - R 90 Nomisma/Banque populaire du Nord, mars 1985, no. 228 62 5.78 0 56 - R 90 Banque populaire du Nord/Crédit de la Bourse, Avril 1987, no. 867; Banque 63 5.99 0 56 - R 90 Banque populaire du Nord, 37, décembre 1985, no. 18 Cabinet Albuquerque, 11,29-9-1988, no. 212 populaire du Nord, 37, décembre 1985, no. 17 56 Scheers 1984, no. 24. Cabinet Albuquerque, Rouen, 25-5-1989, no. 294 64 5.92 0 56 - R 90 CGF Paris, 5,12-12-1998, no. 640 65 5.88 0 56 - R 90 Scheers 1984, no, 30 66 5.75 0 57 - R 93 Argenor, on line 2006, no. AM 743 67 5.84 0 57 - R 93 Scheers 1984, no. 28 68 5.87 0 57 - R 94 Scheers 1984, no. 20. Monnaies et Médailles Bâle, 483, novembre/ 69 5.84 0 57 - R 96 70 5.84 0 58 - R 97 décembre 1985, no. 57 Blancon, 8,1990, no. 30; Blancon, 7,1989, no. 43 71 5.86 0 58 - R 99 Scheers 1984, no. 23; Poinsignon, 30-6-1987, no. 12 72 5.96 0 58 - R 99 Scheers 1984, no. 22 73 5.99 0 5 9 - R 100 Cabinet Albuquerque, Rouen, 14,26-1-1989, no. 1; Elsen, 66,16/17-6-2001, 74 5.89 0 59 - R 100 Vedrines/Poindessault, Paris, 29-12-1984, no. 244 75 5.95 0 5 9 - R 101 Scheers 1984, no. 3. Monnaies et Médailles, Bâle, 474, février 1985, no. 3 76 5.84 0 5 9 - R 102 77 5.86 0 6 0 - R 103 0GN Paris, liste 64, hiver 2009, no. 318 78 5.93 0 60 - R 104 Scheers 1984, ho. 5. Elsen, 113, janvier 1989, no. 8; Elsen, 14,7-10-1989, no. 79 5.89 0 61 - R 105 no.113 13 Cabinet Albuquerque, 11-7-1987, no. 5; Cabinet Albuquerque, 25-5-1987, no. 246 g o ld staters type Scheers 29, class lib , N e rvii 80 5.85 O 6 6 - R 112 81 5.91 0 66 - R 112 82 5.92 O 66 - R 112 83 5.93 0 6 7 - R 112 84 5.86 0 68 - R 114 Scheers 1984, no. 26. Burgan, 22-12-1984, no. 9 Banque populaire du Nord, avril 1986, no. 47; Banque populaire du Nord/ Crédit de la Bourse, Octobre 1986, no. 268 Thirion, automne 1992, no. 8; Thirion, Automne 1990, no. 3; Banque populaire du Nord/Crédit de la Bourse, avril 1987, no. 866 ; Banque populaire du Nord, 37, décembre 1985, no. 19 85 5.87 O 6 8 - R 114 Scheers 1984, no. 27 86 5.88 O 68 - R 117 Elsen, 17, 17-11-1990, no. 8 reputedlyfound at Herderen (Limburg, B.) 87 5.86 0 6 9 - R 118 Scheers 1984, no. 21. Delestrée-Tache 178 88 5.84 O 69 - R 119 Scheers 1984, no. 19 89 5.96 0 69- R 119 Elsen, liste 58, août 1983, no. 7 90 5.89 0 69- R 119 Scheers 1984, no. 25 91 5.87 0 69 - R 120 Scheers 1984, no. 9. Thirion, automne 1989, no. 4 92 5.89 0 69^ R 121 93 5.83 0 6 9 - R 121 Scheers 1984, no. 29; Peus, 4-11-1998, no. 6 94 5.98 0 7 0 - R 122 Baldwin, 67/68,28/29-9-2010, no. 3263 quarter staters type Scheers 152-V, class I, N ervii? 95 1.13 Scheers 1984, no. 38. Coll. S. Scheers 96 0.61 coll. S. Scheers (broken) 97 1.07 Scheers 1984, no. 36 98 0.80 Scheers 1984, no. 37. Coll. S. Scheers 99 1.29 100 1.09 101 1.21 Scheers 1984, no. 39 57 102 1.05 103 1.10 quarter staters type Scheers 152-V, class II, N ervii? 104 1.03 Monnaies et Médailles Bâle, liste 471, octobre 1984, no. 19 gold bracelet 105 Solid gold bracelet with slightly thickened ends; 37.50 g. 58 Scheers 1984, no. 40. PLATES All coins are depicted on the scale 2:1. The numbering of the coins corresponds to the numbering in Appendix 1. 59 FRAIRE, 6o PLATE I FRAIRE, PLATE 2 6l FRAIRE, 62 PLATE 3 FRAIRE, PLATE \ 63 FRAIRE, 64 PLATE 5 PRAIRE, PLATE 6 65 PP/ FRAIRE, 68 PLATE 9 FRAIRE, PLATE I O 69 Three gold hoards from T huin Simone Scheers / Guido Creemers / Nico Roymans / Luc Van Impe 1 2 2.1 2.2 2.3 2.4 3 4 Site and find contexts Composition of the hoards Thuin-1 Thuin-2 Thuin-3 Terminal knob of a gold torque Dating o f the hoards Discussion References Appendix 1. Descriptive fist of coins from theThuin-1 hoard Appendix 2. Descriptive list of coins from the Thuin-2 hoard Appendix 3. Descriptive list of coins from the Thuin-3 hoard Plates I SITE AND FIND CO N TEXTS1 Thuin is situated in the Belgian province of Hainaut near the confluence of the river Sambre and its tribu tary the Biesmelle (fig. l).The area on the southern bank of the Sambre is characterised by hills with a stony loam topsoil that covers sandstone and schist. Immediately to the south ofThuin are two forested hills, the Bois du Grand Bon Dieu and the Bois de Luiseul, separated by the Biesmelle valley. The former contains a pardy ramparted and ditched fortification of the éperon barré type, which is enclosed on the southern and western sides by the Biesmelle valley and on the northern side by the R y valley.12The site lies at an altitude of 170 m and about 20 m above the adjacent valley. An excavation by the Université Libre de Bruxelles, headed by Pierre Bonenfant and Eric 1luyseconi, has been conducted on the eastern side, at the entrance to the éperon barré. This work uncovered the remains of a fortified site from the Neolithic and Iron Age, with the Iron Age site covering an area of 13 ha. Radiocarbon dating of charcoal from the principal rampart of the fortification has given a calibrated date of between 90 BC and 60 AD (with 68.2 % probability), thus making a date of the most recent fortification in the Late La Tène period very plausible.3 1 ville and Walter Lederçq/Eugène Warmenbol (Univer We would like to thank all persons for having provided sité Libre de Bruxelles). information used in this study: Anne de Breuck (Koning Boudewijnsdchting/Fondation R o i Baudouin), Anne Cahen, Nicolas Cauwe and Claire Massart (Musées 2 3 Bonenfant/Huysecom 1981; Conreur 2000. Bonenfant/Huysecom 1981, 109. Radiocarbon date of Royaux d’Art & d’Histoire de Bruxelles), Jacqueline 2020 + / - 50 BP (Lv-1287). The calibration according Cession-Louppe (Musée royal de Mariemont), Xavier to Oxcal 3.10 - Oxford Radiocarbon Acceleration U nit D eru (Université Lille), Jean-Marc Doyen, Roland Fau- differs from older calibrations. Cf.Van Heesch 1998, 81. 71 Fig. 1. Topography o f the Late Iron Age fortification at Thuin, the location o f the three gold hoards (T-l,T-2, T-3) and other finds, a gold hoard; b isolated gold coin(s); c concentration of R om an lead sling bullets; d metal tools; e bronze orna ments. An earth and stone rampart, still measuring 3 m high and 9 m wide, controlled the approx. 60 m-wide entrance to the plateau (fig. 2). Beyond the rampart was a ditch 1.60 m deep and 3.80 m wide.4 The remains of a second rampart are visible immediately to the southwest of the rampart; they have yet to be examined more closely.The fortification is shaped like an irregular pentagon (fig. I).5 In contrast to the nearby sites of Rouveroy, Lompret and Olloy, which are of the mums gallicus type, Thuin is an éperon barré of the Fecamp type, with solid ramparts without internal beams or stone facing, and with a wide flat-bottomed ditch in front. Fortifications of this type occur predominantly in the western part of Belgic Gaul, mainly during LT D2 (70-20/15 BC).6 No archaeological finds from this fortification are known in the literature. However, we were able to study material collected there over the years by a local amateur archaeologist. He found scattered near the slope of the R y stream (fig. 1) three bronze La Tène necklaces or bracelets, two bronze appliqués in the shape of a bull’s head, and a terminal knob from a sword hilt or knife handle (?) in the shape of a double-faced human head (fig. 3).7 Various iron tools of Late Iron Age type were found in the centre 4 Bonenfant 1998; Bonenfant/Huysecom 1981; Frébutte et For similar armrings, neck rings and bull’s head appliqués, al. 1992,18. see Leman-Delerive/Beaussart 1990, 79, 168-169, 245 5 Brulet 2008, 28, fig. 28; Cahen-Delhaye 1984,155. fig. 68, 246 fig. 69, 249 fig. 77, 250 fig. 79, 254 fig. 89; 6 Fichtl 1994, 41-42,176; Leman-Delerive 2006,184. Swinkels 1994, 86, fig. 3-d, 90; Roymans/Dijkman, this 72 Fig. 2. Section o f the main wall near the entrance to the fortification atT h u in during the 1980 excavation. Photo Université Libre de Bruxelles. of the fortification, including three socketed axes, an anvil and the sheath of an ard share (fig. 3).8A sur prising discovery was that of two concentrations of Roman lead sling bullets, one of which was found on the rampart at the main entrance to the éperon barré and the other outside the fortification in the Biesmelle valley (fig. 1). Although difficult to date precisely, these bullets (fig. 4) point to a siege by the Rom an army at some time during the Late Republican or Augustan-Tiberian period.9Finally, an isolated Eburonean stater (type Scheers 31, class la) was found inside the fortifcation.10 In Rom an times there will hâve been a setdement on the plateau outside the fortification, as attested to by the presence of a cemetery on the ‘Le Petit Paradis’ site directly east of the entrance to the forti fication.11 The graves date from the late 2nd and the 3rd century AD. Significantly, however, there is a Perrin 1990,78-79; Lambot 2006; Schonfelder 2006. publication, fig. 19; A. Duval, in Barbieux 1992, 77-81. A dating from the end o f the 3rd and especially the 2nd 9 century BC appears likely, although they can occur up until the 1st century AD. An identical example to the 8 See the discussion in Roymans/Scheers, this volume, 20 ff. 10 Information M. Doyen. The Eburonean stater derives hum an head o f cast bronze (?) is known from Deil in the from die 0 2 (obverse) and R 2 (reverse). These dies Dutch river area. However, this is an unpublished stray are also represented in the coin hoard o f Philippeville. find whose dating remains uncertain. Nevertheless, this See Roymans/Scheers, this volume, 35, Appendix 1. site has yielded several gold Scheers 31 staters, as well as The coin is therefore contemporaneous with the ‘early’ silver rainbow staters: Roymans 2004, 52, 97. hoards o f the Fraire/Amby horizon. Such axes feature among the typical find material from Late Iron Age sites. Cf. for example Collis 1984, 87-93; 11 Brulet 2008, 28,185,186, fig. 27; Faider-Feytmans 1965; M arien 1980, 59,474. 73 74 Fig. 3. Late Iron Age ornaments and tools from the fortification atT huin. Scale 1:2, except f scale 1:1. a-c bronze LaTène armrings and neck rings; d-e bronze appliqués with representations o f a bull's head; f bronze terminal knob in the shape o f a double-faced human head; g iron sheath o f an ard share; h iron socketed axe; i iron anvil; j bronze ring complete absence of finds from the first two centuries AD within the fortification, which suggests that the Late Iron Age habitation did not continue into the Roman period. Three Late Iron Age gold hoards were discovered in the 1980s at two different locations - the Bois du Grand Bon Dieu and the Bois de Luiseul. Using a metal detector, R obert Coenen unearthed a coin hoard in May 1980 on the northwestern boundary of the fortification on the plateau of the Bois du Grand Bon Dieu (fig. 1). The hoard, labelled Thuin-1, contained 73 staters of the Scheers 29 type (fig. 6).12The coins lay dispersed across an area of three by one metre, at a depth of 30 to 40 cm. The hoard had probably originally been stored in a container of organic material (leather or textile), of which no traces remain. The complete coin hoard, with the exception of six items, was purchased by the Koning Boudewijn Stichting for the Koninklijke Musea voor Kunst en Geschiedenis in Brussels.13Then in 1981 the Thuin-2 hoard was discovered, a little further south on the edge of the Bois de Luiseul plateau (fig. I),14 a site separated from the éperon barré by the Biesmellë valley.The coins were dispersed across several metres at a depth of 15 to 20 cm. There were no traces of a container. In total, 52 epsilon staters of the Scheers 29 type of the Nervii and one quarter stater with ‘circle segments’ (Scheers 152-V, class 1) were collected, three of which were purchased by the Musée Royal de Mariemont;15 the remainders were sold. The Thuin-3 hoard was discovered by the same individual in 1982, roughly 20 m away from the Thuin-2 find site. The coins were scattered across an area of several square metres at a depth of 15 to 20 cm. Once again, there was no trace of a container. Thuin-3 comprised 20 staters of the Scheers 29 type of the Nervii. The hoard has now vanished without trace, but we do have photos (albeit of inferior quality) of the coins. In addition to the two gold hoards, the finder collected several more coins (fig. 5) within a zone measuring approx. 40 by 20 m: an Eburonean stater Scheers 31, class la,16two quarter staters Scheers 152V, classes I and III,17 and a silver quinar bearing the legend SEQVANOIOTVOS.18There were also two Nervian staters Scheers 29.19 It is quite possible that some of these coins were originally part of one of the hoards. In addition, 10 potin coins of the au rameau type (Scheers 190 type, class IV) were collected 12 Brulet 2008,33; Huysecom 1981; Leman-Delerive/War- 18 Type La Tour 5351, weighing 1.76 g. Seventeen quinarii menbol 2007; Scheers 1996,41-48;Van Heesch 1991;Van o f this type were found at Alesia, including one example Heesch 1998,35. from Camp C (Fischer/Gruel 2001, nos. 590-603, who only cite 14 coins). These quinarii circulated during the 13 We wish to thank Anne de Breuck (Koning Boudewijn Stichting), Claire Massart and Anne Cahen (Koninklijke Gallic Wars, hut continued in circulation for a long time. Musea Kunst en Geschiedenis) for their help with this M inting may even have begun before the Gallic Wars, study and for granting permission to publish. although there is no hard evidence for this. The coin from T huin belongs to the early emissions o f finely fin 14 In 2010 we visited the find spot together with the finder, ished types, but its poorly defined relief and low weight who pointed out the exact location. (normally 1.90 - 1.94 g) lead us to suspect that the coin 15 Brulet 2008, 33; Cession-Louppe 1993, 46-47; Scheers had been in circulation for quite some time. It is also the 1996, 41-48;Van Heesch 1998, 35. northern-m ost find o f this type. 16 This was an early plated example (3.75 g) o f class la, struck w ith die 0 4 (obverse) and R 5 (reverse), as known 19 Class I (5,96 g) and class II (5.91 g) respectively. from the hoard o f Fraire and Philippeville (see R oymans/Scheers, this volume, 35, Appendix 1). 17 0.51 and 0.81 g respectively. Simone Scheers collection. 75 Fig. 4. Rom an lead sling bullets found in two different concentrations at the main entrance o f the fortification and in the Biesmelle valley. Scale 1:1. 76 Lg. 5. Late Iron Age coin finds from the cult place (?) in the Bois de Luiseul atT huin. Scale 2:1 (nos. 1-5) and 1:1 (nos. 6-15). stater Scheers 31-1 (Eburones); 2 stater Scheers 29-11 (Nervii); 3 stater Scheers 29-1 (Nervii); 4 quarter stater Scheers 152-V, ass I (Nervii); 5 quarter stater Scheers 152-V, class III (Nervii); 6-15 hoard ofp o tin coins Scheers 190-IV (Nervii) 77 Fig. 6. The hoard o f T huin-1. Overview o f the coins. Photo Koning Boudewijnstichting/Fondation R o i Baudouin. across an area of about 10 x 20 m (but outside the concentrations of the Thuin-1 andThuin-2 coins); these probably originally constituted a coin hoard (fig. 5). Lastly, a buffer terminal from a gold torque was found (fig. 7).20There is a conspicuous absence of fibulae and other metal finds from the Late Iron Age; finds from Roman times are also entirely absent. The site in question appears to have been used for a short time in LTD2 for special depositions. We can but speculate about the reasons for burying the three gold hoards of Thuin, especially given the absence of good context data gleaned from controled excavations. In the case of Thuin-I, this was a one-off deposit in the peripheral zone of an oppidum, quite possibly a Versteckdepot. The situation was different in the Bois de Luiseul, where multiple depositions of principally gold coins and a torque frag20 The Eburonean stater and the potin coins are now in the Simone Scheers collection. The other items have been retained by the finder. 78 Fig. 7. Buffer terminal o f a Late Iron Age gold torque found in the Bois de Luiseul atT huin. Scale 2:1. ment were made within a small area in the space of perhaps just a few years.21 These may have been ritual depositions at a cult place.22 The presence of the terminal from a gold tor que close to the Thuin-2 and Thuin-3 hoards calls to mind the well-known gold hoards of Niederzier, Beringen and Frasnes-les-Buissenal (see below), where coins were associated with jewellery and which are beheved to be ritual hoards. Sig nificantly, it has been established (see below) that the three gold hoards ofThuin must have been buried at about the same time, which suggests the possibility of a link with a specific historic event.23 We do not know for certain which tribal polity the area around Thuin belonged to in the Late Iron Age. In Rom an times the southern bank of the Sambre atThuin is said to have lain just outside the civitas Nerviorum in the territory of the Tungri. As Faider-Feytmans and others have stated, the boundaries of the Nervian territory are based on those of the medieval diocese of Cambrai. We do not know to what extent this reflects the situation in Roman times.24 Based on the medieval diocese boundary, the eastern boundary of the Roman civitas Nerviorum is generally said to have been marked here by the Hantes; 21 T he Scheers 190 potin coins already circulated at the 25 potin coins. They are a bit less heavy, w ith weights time o f the Caesarian conquest. Typological variation ranging from 4.10 to 1.89 g, although only 6 coins have together with differences in weight and diameter allow a lower weight than the lightest T huin coin and some o f for a rough classification, but the chronology o f these these are incomplete or clipped. In the hoard o f Peissant groups remains uncertain. There is no doubt that the (B) the potin coins were associated with Scheers 29 gold .earliest group contains the heavy coins, weighing over staters. Scheers 1977, 892. For the dating o f the silver 6.00 to 5.00 g,but most coins are lighter and the weight . quinar with the legend SEQVANOIOTVOS, see note 18 above. fluctuates between 4.99 and 2.00 g.This is also the case for the 10 coins from Thuin. Their weights run from 22 4.76 to 2.25 g, although some o f the weights under 3.00 focus. g may be due to the brittle or w orn condition o f these coins. The presence o f a light-weight rameau-potin in camp B at Alesia, confirms that these coins circulated O f possible significance here is the presence on the site o f a rocky outcrop, which may have functioned as a ritual 23 24 See the discussion in Roymans/Scheers, this volume, 20 ff. Boucly 1978; Brulet 2008, 53; D e Laet 1961; Faider- during the Gallic War (Fischer/Gruel 2001, no. 676). Feytmans 1952,342 fig .l/ 345 fig. 2,350; Frebutte 1992, Weight (3.44 g), diameter (20 mm) as well as the type of 17 fig. 3; Marien 1980, 34, and map 472; Raepsaet- the Alesia coin are closely related to the T huin coins. The Charlier 1994, 45, 52; Scheers 1996, 41-48; Van Heesch burial o f this small hoard during the Gallic War is a dis 1998,12/27 note 8;Wightman 1985,53. tinct possibility. A similar hoard found at Fraire contained 79 it then followed the Sambre for some distance before moving further northwards along the PietonLasne-Dijle route. This would place Thuin outside Nervian territory. The situation is of course much murkier for the Late La Tene period. Most scholars suppose that here too the situation in the Roman or medieval period can be taken as the norm. Others (BouclyWightman andVan Heesch) have adopted a more nuanced position. They believe that the area between the Sambre and Meuse may have been incorporated in Nervian territory at that time. Deru has recently made a further evaluation of all known hypotheses. Basing himself primarily on numismatic data (the distribution of Nervian gold staters), he states that the zone to the south of the Sambre at Thuin belonged to Nervian territory.25 However, we cannot establish which tribal group the area around Thuin belonged to at the time of Caesar’s conquests without looking at the territory of the Aduatuci. Judging by Caesar’s comments, this group must have been located between the Eburonean and Nervian territories. This would make the region on either side of the Belgian part of the Sambre, including the area between the Sambre and Meuse, the most likely candidate. If we accept this position, the environs o f Thuin formed part of Aduatucian territory. 2 COMPOSITION 2.1 T H U IN - I OF THE HOARDS The hoard (fig. 6) is made up exclusively of staters of the Scheers 29 type, ascribed to the Nervii. The 73 coins belong to classes I and II. The coins had clearly been in circulation before being removed and concealed, but their good state of preservation shows that this must have been for only a short time. Numerous die associations link this hoard to those of Fraire, Philippeville, Thuin-2 and Thuin-3, and Heers,26 which shows that they belong to the same horizon. The hoard contained 17 staters o f class lib (nos. 57-73), but as with the hoard of Fraire, Thuin-1 lacks the most recent class lib emissions that are so strongly represented in the hoard of Heers.27 2.2 T H U IN -2 This hoard consists solely of Nervian coins: 53 staters of Scheers 29 type and one quarter stater of Scheers 152 type, class V Gold staters of the Nervii The 53 staters belong to Scheers 29 type, classes I and II. They are in an excellent state of preservation, which shows that they could only have circulated very briefly. The Thuin-2 hoard has many dies in com mon with the hoards of Fraire andThuin-1, suggesting that the coins must have been in circulation at 25 D eru 2009. However, there is a danger o f circular reason Iron Age ‘coin territories’ and tribal groups m entioned tion o f Late Iron Age coinages to specific tribes is based by Caesar, see Haselgrove 1999, 119. For a recent over on correspondences with the tribal map o f N orthern view o f the distribution o f Late Iron Age coins and coin Gaul drawn by Caesar. Conversely, it is therefore risky to hoards in the region o f Thuin/Fontaine-Valmont, see reconstruct territories o f the tribes named by Caesar on Rossez 2010. the basis o f the distribution of certain coin types. Moreo ver, it is always possible that different groups (in this case the Nervii and Aduatuci) used the same coin type. For 8o the general problem o f correspondences between Late ing in his argument. We must not forget that the attribu 26 27 See Roymans/Scheers, this volume, 36 ff., Appendix 2. Heers nos.126-154. See Roymans/Scheers this volume, 39, Appendix 2. the same time. A further argument is that despite the presence of ten class lib staters, this hoard also lacks the most recent emissions of this class that are so characteristic of the hoard of Heers. The quarter stater of the Nervii The quarter stater of the Scheers 152-V type, featuring a wheel above the horse, belongs to class I.28The Fraire hoard also contained nine quarter staters of this type. Their presence in the hoards together with the staters shows that they were contemporaneous. 2.3 THUIN-3 This hoard consists of 20 staters of the Scheers 29 type. A study of the hoard is hampered by the fact that it has vanished; poor-quality photos of the coins are all that remain. For this reason, the identification of the coin dies is uncertain in some instances. We know the weight of only a single stater. The 20 staters belong to classes I and II. Here too, class lib is represented by eight staters (nos. 13-20). There are no examples of more recent emissions, however; these are present only in the hoard of Heers (Heers nos. 126-154). 2-4 TERMINAL KNOB OF A GOLD TORQUE The buffer terminal of a gold torque (fig. 7) was discovered on the same site where the Thuin-2 and Thuin-3 hoards were found (see above).The finder identified it as a stray find, but we cannot altogether rule out that it was originally part of one of the gold hoards. The buffer terminal consists of a concave, spherical knob showing severe secondary damage; the knob has been flattened along both sides and a piece of beading is bent. The knob is made of 0.9 mm thick gold plate and the intact spherical edges are 19 mm in diameter. It weighs 5.41 g and is 20.5 mm in length. The upper side, originally open, is closed with a biconical ‘spool’, the attachment of which is concealed beneath some beading. The closed upper edge of the spool is also finished with beading. The endplate on the spool features a rectangular 4 mm high hook, consisting of two blocks soldered together.The upper block measures 5 x 3.5 mm and is 3.7 mm thick. There are few parallels for buffer terminals of this type. The closest parallel is the terminal knob of a twisted torque from Soucy-Montgobert (dép. Aisne, F) :29 here the three twisted gold threads terminate in concave, spherical knobs, one of which has a rectangular keyhole and the other the hook. In contrast to the hook from Thuin, that of Soucy was cast in one piece. Hook- and T-shaped fasteners were a fairly common feature of continental golden torques from the 3rd century BC onwards, on both richly decorated and plainer examples. Equally important, however, is the similarity between the biconical spool with beaded edges that supports the hook, and the almost identical components on the buffer terminals of the torques from Beringen and Niederzier (fig. 8).30 The similarities between Thuin, Beringen and Niederzier therefore suggest a dating for the Thuin knob from the mid-2nd to the m id-lst century BC. This dating means that the torque fragment and the gold hoards may have been deposited at the same time. Whether the torque knob was scrap metal intended to be melted down again is a question we are unable to answer. See Roymans/Scheers, this volume, fig. 5. 30 Joachim/Zedelius 1980; Joachim 1991; ibid. 1999; ibid. Éluère 1986. Hautenauve, in Ugaglia 2001, 123-124; 2007; Göbel et al. 1991;Van Impe et al. 1997/1998; Hau- Hautenauve 2005, 229-230 (cat. no. 124). tenauve 2005,76-77. 8l o f the gold torques. Photo VIOE Brussels. 3 DATING OF THE HOARDS The composition ofThuin-1 is strongly reminiscent of that of Fraire. Here too class I (42 coins) is better represented than class II (31 coins), almost 58% and 42% respectively. Moreover, the oldest staters of class la are the most numerous, with 27 examples (37%).This reinforces our suspicion that class la staters were minted in far greater numbers than the more recent classes. Given their strong similarities, the hoards of Thuin-1 and Fraire are contemporaneous and a little older that those of Heers and Maastricht-Amby. The composition ofThuin-2 also closely resembles that of Fraire and Thuin-1. Here too, class I (30 coins) is more strongly represented than class II (22 coins), 58% and 42% respectively.This proportion is identical to what we observed for the hoard of Thuin-1. The oldest staters of class la are the most numer ous, and in the same proportion, with 22 examples (42%).This once again confirms that class la staters were minted in particularly large numbers. Because of their identical composition, the hoards ofThuin-2, Thuin-1 and Fraire are without doubt contemporaneous. Chronologically speaking, theThuin-3 hoard also belongs to the same horizon as the hoards ofThuinl,T huin-2 and Fraire, although the composition is slightly different. Class I, with seven coins (35%), is less well represented than class II, with 13 coins (65 % of the hoard). This difference may perhaps be accounted for solely by the small number of coins. These four hoards —Fraire, Thuin-1, Thuin-2 and Thuin-3 - clearly form a compact group with numerous die identities.31 Because the most recent class lib staters (Heers nos. 126-154) are systematically absent, the hoards antedate those of Heers and MaastrichtAmby. In an absolute sense, we can assume a dating in the early 50s BC.32 4 DISCUSSION The above data shows that the three coin hoards are strictly contemporaneous in a numismatic sense, and we should therefore consider the possibility that they were buried at the same time in response to a specific historical event. For this reason, we view Thuin as a serious contender for the oppidum of the Aduatuci, which was besieged and conquered by Caesar in 57 BC and which has not yet been convinc ingly located. The significance of the fortification of Thuin has so far been seriously underestimated. Its considerable size (13 ha) and the find material from the Late Iron Age point to a certain centre function for a tribal community Caesar reports that the Aduatuci withdrew as a group in 57 BC to a single forti fication.33 He describes a place with steep slopes in a rocky environment, with on one side an approach that was no more than about 60 m wide. The Aduatuci had fortified that spot with a double wall. Thuin is a good match for this description. Even though Bonenfant has only observed a single rampart, the resemblance to the approx. 60 m wide fortified approach zone is significant.We also consider it a possibil ity that the area between the Sambre and Meuse, where Thuin is located, belonged to the Aduatuci at the time of the Roman conquest. The distribution pattern of coinages ascribed to the Nervii is not of itself an argument for attributing the area to the Nervii. These coins could just as well have circulated among the Aduatuci, who were allies of the Nervii in the anti-Roman coalition of 57 BC. Also significant is the discovery of the two concentrations of Roman lead sling bullets, a unique find for Belgium. The bullets are possibly linked to one of Caesar’s military campaigns, and his siege of the See Roymans/Scheers, this volume, 35 £F., Appendix 1 and 2. See the argumentation in Roymans/Scheers, this vol ume, 7-8. Caesar, BG 2.29. has yet to be explored. Current research suggests that rampart structures o f the Fecamp type were often erected on top o f older structures. Their simple method o f con struction was the quickest way to build a fortification. Cf. Fichd 2010,328-329. The remains o f a second rampart can be seen but this 83 oppidum of the Aduatuci in 57 BC naturally springs to mind first. One of the concentrations was located on the remains of the principal rampart at the main entrance of the Late La Tène fortification, and must be associated with an outside attack on the fort. The size and weight of these lead projectiles - a little less than the average known for Republican slingshot - and the lack of stamps, seems to suggest a date in the Augustan-Tiberian period.35 An argument against this, however, is the complete absence of early Roman find material in the hill fort. It is surprising that Thuin has only sporadically been cited in the past as a potential location for the oppidum of the Aduatuci. Most scholars probably felt that the site was too close to Nervian territory. However, if we look at the different hillforts in Wallonia, there are few that withstand a comparison with Caesar’s description. In the past the Hastedon at St.-Servais in Namur has regularly been linked to the oppidum of the Aduatuci. There the approach zone is 80 m wide and the habitable area more or less corresponds to that of Thuin (13 ha). However, the site was only in use during the Late Hallstatt/Early La Tène period, which would make any association with the Aduatucian oppidum an anachronism.36 In terms of dating, the fortification of Olloy-sur-Viroin (Namur province) is also a potential candidate as the refuge of the Aduatuci, but the area inside the two ramparts - less than 3 ha —is much too small.37 The Late La Tène fortification of the ‘Camp Romain’ at Lompret (Chimay, Hainaut) tallies better with Caesar’s description. The site has two ramparts of the mums gallicus type which seal off an éperon of 16 ha that features steep slopes. However, the access way to the plateau (about 200 m) is much wider than Caesar’s description of the oppidum of the Aduatuci.38 It goes without saying that only continuing research into the various fortified setdements in the west ern part of Wallonia will be able to demonstrate the historical importance o f these sites. For the time being, the fortification of Thuin —with three gold hoards in its immediate vicinity —remains the most likely place of refuge from which the Aduatuci staged a futile defence against the advancing Roman legions in 57 BC. REFERENCES Barbieux, J. (ed.), 1992: Le site d’Hornaing (Nord). Protohistoire, La Tène I, La Tène III, Gallo-Romain, Méro vingien, Douai (Archaeologia Duacensis 5). Bonenfant, P., 1990: Hastedon (Namur). Grande fortification du début de La Tène (475-450 av.J.C.), in G. Leman-Delerive/Ph. Beaussart (eds), Les Celtes en France du Nord et en Belgique, Valanciennes, 81-85. Bonenfant, P., 1998: Les fortifications de l’Age du Fer en Hainaut, in G. Leman-Delerive (ed.), Les Celtes. Rites funéraires en Gaule du Nord entre le Vie et le Ier siècle avantJésus-Christ. Recherches récentes en Wallonie, Namur (Etudes et Documents, série Fouilles 4), 143-148. Bonenfant, P./E. Huysecom 1981: Retranchements préhistoriques à Thuin (Hainaut). Campagne de fouilles 1981, Annales d’histoire de l’Art et d’archéologie 4 , 103-113. Boucly,J.-L. 1978: A propos de la frontière orientale de la Nervie avant la conquête romaine, Les Etudes Classiques 46, 3, 237-249. Brulet, R ., 2008: Les Romains en Wallonie, Bruxelles. 35 See Poux 2008,365-371. 38 Bonenfant 1998, 146-147; Cahen-Delhaye/Jadin 1990. 36 De Laet 1979, 526-529; Bonenfant 1990. For a recent discussion about the historical interpretation 37 Cahen-Delhaye 1990; Doyen/Warmenbol 1981; Grisart o f Iron Age fortifications in Wallonia, see Warmenbol 1972;Warmenbol/Pleuger/Van Strydonck 2010 with an extensive bibliography. «4 2010 . Cahen-Delhaye, A., 1984: Fouilles récentes dans les fortifications de l’Age du Fer en Belgique, in A. CahenDelhaye/A. Duval/G. Leman-Delerive/P. Léman, Les Celtes en Belgique et dans le Nord de la France, Actes du sixième colloque tenu à Bavay et Mons, Bruxelles (Revue du Nord, numéro spécial hors série). Cahen-Delhaye, A., 1990: La fortification du plateau des Cinkes à Olloy-sur-Viroin (Province de Namur), in G.Leman-Delerive/Ph. Beaussart, Les Celtes en France du Nord et en Belgique, 57. Cahen-Delhaye, A ./ A. Duval/G. Leman-Delerive/P. Léman, 1984: Les Celtes en Belgique et dans le Nord de la France, Actes du sixième colloque tenu à Bavay et Mons,Villeneuve d’Asq (Revue du Nord, numéro spécial hors série). Cahen-Delhaye, A./Y.Jadin, 1990: La place forte de Lompret (Sud du Hainaut Belge), in G. Leman-Delerive/Ph. Beaussart (eds), Les Celtes en France du Nord et en Belgique, Valenciennes, 51-55. Cession-Louppe, J., 1993: Archéologie du Hainaut, in Donnay, G./C. Derriks/C. Noppe et al., 30 ans d’acquisitions, tentoonstelhngscatalogus, Mariemont, 40-47. Collis, J., 1984: Oppida. Earliest towns North of theAlps, Sheffield. Conreur, M., 2000: La bataille du Sabis, Thuin. De Laet, S.J., 1961: Les limites des cités des Ménapiens et des Morins, Helinium 1, 20-34. De Laet, S.j., 1979: Prehistorische culturen in het zuiden der Lage Landen, Wetteren. De Loe, A., s.d.: Forma Orbis Romani. Carte archéologique de la Belgique romaine. Répertoire statistique et biblio graphique avec carte. Deru, X., 2009: Cadres géographiques du territoire des Nerviens, Archéologie de la Picardie et du nord de la France. Revue du Nord 91, 383,181-203. Doyen,J.M., 2009: Les monnaies gauloises du sanctuaire de Fontaine-Valmont (Hainaut, Belgique). Essai de synthèse, in J. van Heesch/I. Heeren (eds), Coinage in the IronAge. Essays in honour of Simone Scheers, London, 85-97. D o y e n , J.-M./E.Warmenbol, 1981: La fortification protohistorique d’Olloy surViroin, Bruxelles (Publications du club archéologique Amphora 11)« % Éluère, Chr., 1986: L’orfèvrerie pré-romaine au musée de Cluny, La revue du Louvre et des Musées de France 36, 97-106. Faider-Feytmans, G., 1952: Les limites de la cité des Nerviens, L’Antiquité Classique 21, 338-352. Faider-Feytmans, G., 1965: La Nécropole Gallo-Romaine de Thuin, Morlanwelz (Fouilles du Musée de Mariemont (Belgique) 2). Fichtl, S., 1994: Les Gaulois du nord de la Gaule (150 —20 av.f.-C.), Paris. Fichd, S., 2010: Reflexions sur les remparts de type Fécamp, in S. Fichtl (ed.), Murus celticus. Architecture et fonctions des remparts de l’Âge du Fer, Table ronde internationale 11, 12 octobre 2006, Glux-en-Glenne, Bibracte (Collection Bibracte 19), 315-334. Fischer, B./K. Gruel, 2001: Catalogue des monnaies gauloises, in M. Reddé/S. von Schnurbein (eds), Fouilles et recherches Franco-Allemandes sur les traveaux militairs Romains autour du Mont-Auxois (19911997), 2, Paris, 21-68. Frebutte, C./M.Vandeur/J.-M. Horemans 1992: Les Celtes et la Thudinie. Autour du trésor monétaire nervien de Thuin, Thuin (Publications du Centre d’Histoire de la Thudinie 14). Göbel, J./Hartmann A./H.-E. Joachim/V. Zedelius, 1991: Der spätkeltische Goldschatz von Niederzier, BonnerJahrbuch 191, 27-84. Grisart, E, 1972:Trois localisations nouvelles, Romana Contact 12, 3-4, 4-67. Hautenauve, H., 2005: Les torques d’or du second Âge du Fer en Europe. Techniques, typologies et symbolique, Rennes (Travaux du Laboratoire d’Anthropologie 44). Hautenauve, H ./C . Besson, 1998: Persistance de techniques d’orfèvrerie. Le torque celtique de Soucy (Aisne) et les bracelets gallo-romains de Poiré-sur-Vie (Vendée), Revue archéologique de l Ouest 15, 141-150. 85 Huysecom, E., 1981: Dépôt monétaire celtique àThuin, Archéologie 1, 24-25. Joachim, H.-E., 1991:The votive deposit at Niederzier, in S. Moscati/O.-H. Frey/V Kruta/B. Raftery/M. Szabô (eds), The Celts (exhibition Palazzo Grossi, Venezia), Milano, 532. Joachim, H.-E., 1999: Die Eburonen - Historisches und Archäologisches zu einem ausgerotteten Volks stamm caesarischer Zeit, in G. von Büren/E. Fuchs (eds),Jülich. Stadt - Territorium - Geschichte. Festschrift zum 75jährigen Jubiläum des Jülicher Geschichtsvereins 1923 e.v., 157-170. Joachim, H.-E., 2007: Die späte Eisenzeit am Niederrhein, in M. Hegewisch (ed.), Krieg und Frieden. Kelten-Römer-Germanen, Bonn/Darmstadt, 48-58. Joachim, H .-E./V Zedelius 1980: Ein bedeutender keltischer Verwahrfund aus Niederzier, Kr. Düren, Gymnasium 87, 205-210. Knaepen-Lescrenier, A.M., 1970: Répertoire bibliographique des trouvailles archéologiques de la province de Namur, Bruxelles (Répertoires archéologiques Série A: Répertoires bibliographiques 9). Kurz, G., 1995: Keltische Hort- und Gewässerfunde in Mitteleuropa. Deponierungen der Latènezeit, Stuttgart (Materialhefte zur Archäologie in Baden-Württemberg 33). Lambot, B., 2006: Assemblages métalliques dans les structures du village d‘Acy-Romance (Ardennes). In: Bataille, G./Guillaumet,J.-P. (éds), Les dépôts métalliques au second Age du Fer en Europe tempérée (Actes de la table ronde de Bibracte, Glux-en-Glenne 13-14 octobre 2004), Glux-en-Glenne, 193-209. Leman-Delerive, G., 2006: La Belgique et la Nord de la France du Ille au 1er siècle avant J.-C., inV Kruta et al. (eds), Celtes, Belges, Boïiens, Rèmes, Volques, Exhibition Catalogue, Mariemont, 179-204. Leman-Delerive, G./Ph. Beaussart (eds), 1990: Les Celtes en France du Nord et en Belgique VF- Ier siècle avant J.-C., Bruxelles. Leman-Delerive, G./E. Warmenbol 2007: Dépôts et sites cultuels en contexte ‘humide’ dans les civitates des Nerviens et des Atrebates, in Ph. Barral/A. Daubigney/C. Dunning/G. Kaenel/M.-J. RoulièreLambert (eds), L’Age du Fer dans l’arcjurassien et ses marges. Dépôts, lieux sacrés et territorialité à l’Age du Fer, Actes du 29e colloques international de l’AFEA Bienne, 5-8 mai 2005, vol. 2, Besançon, 426-427. Marien, M.E., 1980: Belgica Antiqua. De Stempel van Rome, Antwerpen. Perrin, Fr., 1990: Un dépôt d’objets gaulois à Larina (Hières-sur-Amby, hère), DARA-Lyon (Documents d’Archéologie en Rhône-Alpes 4). Poux, M., 2008: L’empreinte du militaire césarien dans les faciès mobiliers de LaTène finale, in M. Poux (ed.), Sur les traces de César. Militaria tardo-républicains en contexte gaulois. Actes de le table ronde de Bibracte, 17 octobre 2002, Bibracte (Collection Bibracte 14), 299-432. Raepsaet-Charlier, M .Th., 1994: La cité desTongres sous le Haut-Empire, BonnerJahrbücher 194, 43-59. Rossez C., 2010: Occupation du sol et circulation monétaire de la fin de l’Age du fer à la fin de l’Antiquité dans la région de Merbes-leChâteau (Hainaut, Belgique), Bulletin du Cercle d’Etudes Numismatiques 47, 3,277-308. Scheers, S., 1977: Traité de numismatique celtique IL La Gaule belgique, Paris. Scheers, S., 1996: Frappe et circulation monétaire sur le territoire de la future civitaS tungrorum, Revue belge de numismatique et de sigillographie 142, 5-51. Schönfelder, M., 2006: Ein spâtlatènezeithcher Werkzeug- und Gerätehort aus dem ostkeltischen Gebiet, in G. Bataille/J.-P. Guillaumet (eds), Les dépôts métalliques au second Age du Fer en Europe tempérée (Actes de la table ronde de Bibracte, Glux-en-Glenne 13-14 octobre 2004), 2006,109-127. Swinkels, L., 1994: Een vergoddelijkte Hercules en enkele andere bronsfiguren, in N. Roymans/T. Derks (eds), De tempel van Empel. Een Herculesheiligdom in het woongebied van de Bataven, ’s-Hertogenbosch, 82-91. Thirion, M., 1967: Les trésors monétaires gaulois et romains trouvés en Belgique, Bruxelles (Cercle d’études numismatiques. Travaux 3). Turquin, P, 1955: La bataille de la Selle (du Sabis) en l’an 57 avant J.-C., Les Etudes Classiques, 23,2,113-1 154. 86 Ugaglia, É. (ed.), 2001: L’Or de Tolosa,Toulouse. Van Heesch,J., 1991 : Le trésor gaulois de Thuin, Bruxelles (Fondation R oi Baudouin). Van Heesch, J., 2005: Celtic coins and religious deposits in Belgium, in C. Haselgrove/D. Wigg-Wolf (eds), Ritual and Iron Age coinage in North-West Europe (Studien zu Fundmiinzen der Antike, 20), Berlin/ Durham, 247-263. Van Impe, L./G. Creemers/R. van Laere/S. Scheers/H. Wouters/B. Ziegaus, 1997/1098: De Keltische goudschat van Beringen (prov. Limburg), Archéologie in Vlaanderen 6 (2002), 9-132. Warmenbol E., 2010: La Belgique gauloise. Mythes & archéologies, Brussels. Warmenbol, E./J.-L. Pleuger/M. van Strydonck, 2010: La fortification protohistorique d’Olloy-surViroin: campagne de fouilles 2009 (province de Namur, Belgique), Lunula. Archaeologia protohistorica, 1 8 ,1 3 9 ll4 1 . Wightman, E.M., 1985: Gallia Belgica, London. 87 A P P E N D I X T H U I N - I I. D E S C R I P T I V E LIST OF C O I N S F R O M T HE H O A R D . The numbering corresponds to the numbering of the coins in Plates 1-8. O = obverse die; R = reverse die. The coin composition was identified by means of surface analysis (X-ray fluorescence analysis). No. weight (g) die nos. analysis % references Au Ag Cu 52 33 15 private collection 23 private collection g o ld staters type Scheers 29, class la, N e rvii 6.00 1 0 2 —R 1 6.08 0 4 -R 2 49 28 3 5.99 0 6 -R 5 57 30 13 4 6.00 0 6 -R 7 52 26 22 5 5.98 0 7 -R 8 55 29 16 5.99 0 7 -R 9 54 30 16 29 14 2 6 7 6.02 0 32 - R 52 57 8 5.96 0 27 - R 47 51 26 23 9 5.83 0 16-R31 55 31 14 10 5.96 0 16-R 32 54 32 14 5.85 0 18-R35 55 30 15 30 14 11 12 5.98 0 1 8 - R 35 56 13 6.03 0 24 - R 43 55 27 18 14 6.01 010 — R 15 54 32 14 15 5.88 0 10 — R 15 52 30 18 16 5.94 0 10 — R 16 58 29 13 17 5.94 0 17 - R 34 53 24 23 18 5.97 0 1 5 -R 2 5 55 35 10 19 5.95 0 1 5 -R 2 6 54 32 14 20 5.97 0 15-R28 60 29 11 6.04 0 15-R30 54 31 15 22 5.89 0 1 3 - R 23 50 30 20 23 6.01 0 22 -R 39 59 28 13 24 5.96 0 22 - R 40 55 26 19 25 5.89 0 22 - R 41 55 29 16 6.01 0 43 - R 64 58 31 11 21 26 g o ld staters type Scheers 29, class lb, N e rvii 27 5.94 0 45 - R 68 51 28 20 28 5.89 0 46 - R 68 52 34 14 29 5.90 0 53 - R 86 58 32 10 9 30 5.91 0 51 - R 80 59 32 31 5.90 0 51 - R 80 56 30 13 32 5.89 0 51 - R 80 58 31 11 33 5.87 0 51 - R 82 34 private collection 5.95 0 52 - R 83 35 5.90 0 52 - R 83 50 21 29 36 5.97 0 50 - R 73 52 35 13 37 5.89 0 50 - R 75 56 32 12 88 38 5.92 0 50 - R 75 60 26 14 39 5.84 0 50 - R 75 55 31 14 40 5.93 0 50 - R 76 55 36 9 41 5.87 0 40 - R 61 49 24 27 42 5.94 0 48-R71 49 27 24 21 g o ld staters type Scheers 29, class lia , N e rvii 43 5.88 0 58 - R 98 51 28 44 5.93 0 58 - R 99 58 28 14 45 5.85 0 57 - R 95 50 20 30 46 5.82 0 57 - R 95 50 23 27 5.93 0 56 - R 89 51 18 31 15 47 48 5.84 0 56 - R 90 55 30 49 5.98 0 56 - R 90 58 31 11 50 5.89 0 56 - R 90 52 28 19 51 5.88 0 56 - R 90 59 24 17 26 24 29 52 5.92 0 60-R103 50 53 5.95 0 60-R104 51 20 54 5.94 0 59-R101 55 28 17 55 5.81 0 59-R102 53 25 22 56 5.98 0 6 3-R 1 08 52 25 23 g o ld staters type Scheers 29, class lib , N e rvii 57 5.86 0 65 - R 111 55 25 20 58 5.90 0 6 5 - R 112 57 27 16 59 5.95 0 66 - R 112 60 6.05 0 6 6 - R 112 57 32 11 11 61 5.81 0 6 6 - R 112 57 32 62 5.99 0 6 6 - R 113 52 28 20 63 5.85 0 6 6 - R 113 55 34 11 64 5.92 0 6 6 - R 113 57 31 12 65 5.91 0 6 9 - R 119 54 32 14 66 5.88 0 69 — R 119 59 31 10 67 5.79 0 6 8 - R 114 56 30 14 68 5.91 0 6 8 - R 114 56 34 10 69 5.90 0 6 8 - R 114 52 24 24 30 16 70 5.90 0 68-R114 54 71 5.92 0 6 8 - R 114 60 27 13 72 5.90 0 6 8 - R 117 54 24 22 73 5.92 0 6 8 - R 115 54 24 22 private collection private collection private collection 89 A P P E N D I X T H U I N - 2 2. D E S C R I P T I V E LIST OF C O I N S F R O M THE H O A R D . The numbering corresponds to the numbering of the coins in Plates 9-14. O = obverse die; R = reverse die. No. weight (g) die nos. references gold staters type Scheers 29, class la, Nervii 1 6.09 0 5- R4 2 5.98 0 9 - R 13 3 5.90 0 11 - R 20 4 5.96 0 12 - R 21 5 5.82 0 13 - R 23 Elsen, 35,2-7-1994, no. 13 6 5.92 0 15 - R 24 coll. S. Scheers 7 5.95 0 1 5 - R 27 8 6.00 0 1 8 - R 36 9 5.79 0 20 - R 37 10 5.86 0 21 - R 38 11 - 0 23 - R 42 12 5.96 0 23 - R 42 13 5.95 0 24 - R 43 14 5.92 0 25 - R 44 15 5.91 0 26 - R 45 16 5.93 0 26 - R 45 Elsen, 37,17-12-1994, no. 2 17 5.83 0 33 - R 53 Jacquier, liste 16, automne 1994, no. 4 18 5.92 0 34 - R 54 19 5.86 0 37 - R 57 20 5.86 0 38 - R 58 21 5.95 0 41 - R 62 22 5.87 0 42 - R 63 CGF Paris, on line, no. 215217 Musée Mariemont, photo 93-9 Elsen, 36,17-9-1994, no. 11; Elsen, 35,2-7-1994, no. 12 g o ld staters type Scheers 29, class lb, N e rvii 23 5.81 0 46 - R 69 24 5.96 0 49 - R 72 25 5.85 0 50 - R 74 26 5.94 0 50 - R 75 27 5.99 0 50 - R 76 28 5.96 0 50 - R 77 29 5.96 0 51§f R 80 30 5.89 0 52 - R 83 Elsen, 225, mai-août 2003, no. 9; Elsen, 78, 20-3-2004, no. 12; Elsen 69,16-3-2002, no. 37; Elsen, 63,16-9-2000, no. 41 Musée Mariemont, photo 93-8 g o ld staters type Scheers 29, class lia , N e rvii 31 5.95 0 56 - R 89 32 5.91 0 56 - R 89 33 5.87 0 56 - R 90 34 5.85 0 56 - R 90 35 5.85 0 56 - R 92 36 5.83 0 57 - R 93 Elsen, 103,12-12-2009, no. 13 37 5.87 0 57 - R 94 Jacquier, liste 18, automne 1996, no. 45; Jacquier, liste 16, automne 1994, ni. 5 90 Elsen, 110,10-9-2011, no. 79 38 5.87 0 58 - R 98 39 5.85 0 61 - R 105 40 5.85 0 61 - R 106 41 5.97 0 62 - R 107 42 5.99 0 6 2 - R 107 Jacquier, liste 19, été 1997, no. 22; Jacquier, liste 18, automne 1996, no. 44 g o ld staters type Scheers 29, class lib , N e rvii 43 5.82 0 65 - R 112 44 5.86 0 66- R 112 45 5.87 0 6 6 - R 112 46 5.88 0 66- R 113 47 5.88 0 6 8 -R 1 1 4 48 5.88 0 68 - R 114 49 5.86 0 6 8 - R 116 50 5.84 0 6 9 - R 118 51 5.96 0 6 9 - R 119 52 5.92 0 69 - R 119 Elsen, 35,2-7-1994, no. 11 Musée Mariemont 93-10 UBS,77,9/12-9-2008, no. 138; Jacquier, liste 25,2000, no. 34; Jacquier, liste 24, printemps 2000, no. 15 quarter stater type Scheers 152-V, class 1 53 0.72 9i A P P E N D I X T H U I N - 3 3 . D E S C R I P T I V E LIST OF C O I N S F R O M THE H O A R D . The numbering corresponds to the numbering of the coins in Plate 1 5 .0 = obverse die; R = reverse die. The weights of the coins could not be documented. The identification of the dies is not always certain, because of the poor quality of the photos. No. weight (g) die nos. g o ld staters type Scheers 29, class la, N e rvii 1 - 0 4 -R 3 2 - 0 6 -R 7 3 - 0 15 - R 26 4 - 0 35 - R 55? 5 - 0 36 - R 56? 6 - 0 44 - R 65? g o ld staters type Scheers 29, class lb, N e rvii 7 - 0 55? - R 88? g o ld staters type Scheers 29, class lla , N e rvii 8 9 0 56 - R 90 5.80 0 58 - R 98 10 0 59-R101 11 0 63-R109? 12 0 64 - R 110? g o ld staters type Scheers 29, class lib , N e rvii 13 - 0 6 6 - R 112 14 - 0 6 6 - R 113 15 - 0 68 - R 114 16 - 0 69 - R 119 17 - 0 6 9 - R 119 18 - 0 69 — R 119 19 - 0 69 — R 119 20 - 0 6 9 - R 119 92 PLATES All coins are depicted on the scale 2:1, except for those ffomThuin-3 which are on scale 1:1.The num bering of the coins corresponds to the numbering in Appendix 1-3. For the coins ofThuin-3 only scans of low quality photos were available.The black-and-white photos of six coins from Thuin-1 were received at a later stage and placed at the end of the list (Plates 7-8). 93 T H U IN -I , PLATE 3 I T H U I N - I , PLATE 2 T H U IN -I , 96 PLATE 3 T H U IN -I , PLATE 4 97 T H U I N -I , 98 PLATE 5 T H U I N - I , PLATE 6 99 T H U I N - I , IOO PLATE 7 T H U I N - I , PLATE IO I T H U I N - 2 , 102 PLATE I T H U I N - 2 , PLATE 2 103 T H U I N - 2 , 10 4 PLATE 3 T H U I N - 2 , PLATE 4 10 5 T H U I N - 2 , I 06 PLATE 5 T H U I N - 2 , PLATE 6 10 7 iVi T H U I N - 3 , I 08 PLATE I A coin hoard w ith an animal-headed gold bracelet from the vicinity of Philipp eville Nico Roymans / Simone Scheers 1 2 3 4 Discovery of the hoard The gold coins The animal-headed gold bracelet Conclusions References Appendix 1. Descriptive list of the coins from the Philippeville hoard I D I S C O V E R Y OF T H E H O A R D In December 2010 the second author received a message from M r Jean Elsen, director of an auction house in Brussels, that he had acquired a Celtic gold hoard, which —apart from some 15 gold coins —also included a decorated gold bracelet. As in the case of the hoard from Fraire,1 here too the context infor mation was minimal and the finder remained anonymous to us. According to M r Elsen the hoard was found in the vicinity of Philippeville (B), which is in the direct neighbourhood of Fraire andThuin.12The finder, however, did not want to give information about the exact findspot. The finds were reportedly collected in two concentrations within a distance of only a few meters; the first one included the bracelet and six Nervian staters, the second one four Eburonean staters and five quarter staters (see below).The fact that all the finds were found at a very close distance from each other and that all coins belonged to the Fraire/Amby horizon (see below), makes it plausible that we are dealing with the remains of a single hoard, which was slighdy scattered by later soil disturbances. This is not certain, however; we may be also dealing here with two separate hoards that had been deposited more or less at the same time. 2 T H E G O L D C O I N S In total 15 coins have been encountered in the ‘Philippeville’ hoard, belonging to four different types. Firsdy there are four staters of the Scheers 31 type, ascribed to the Eburones (fig. 1, nos. 1-4).They have all been struck with early dies of class la, which are also represented in the hoards of Fraire, Heers and Amby.3 Secondly there are six Nervian staters of the type Scheers 29, one belonging to class I and five to class II 1 Cf. Scheers/Creemers, this volume, 47 S'. 2 In fact the find place cannot be indicated more specifically than ‘in de region between Sambre and Meuse’.The region and is also the place were the finder o f the hoard lives. 3 See Roymans/Scheers, this volume, 35, Appendix 1. small town o f Philippeville is situated in the centre o f this 10 9 Fig. 1. Gold staters ascribed to the Eburones (nos. 1-4) and the Nervii (nos. 5-10) from the hoard of Philippeville. Scale 2:1. Fig. 2. Q u a r te r staters o f th e N e rv ii (nos. 1 1 -1 4 ) a n d th e R e m i (15) fro m th e P h ilip p ev ille h o a rd . Scale 2:1. (fig. 1, nos. 5-10). O f this latter group two items belong to class Ha and three to class lib. All coins have die connections with coins from the hoards of Fraire and Thuin.4 Significant, however, is that coins of the youngest ‘Heers variant’ with crosses of class lib are absent. Thirdly we have four quarter staters of the type Scheers 152, with on the reverse a wheel above a horse facing left (fig. 2, nos. 11-14).Three belong to class I and one to class III.5They probably represent a Nervian emission and have also been found in the hoards of Fraire andThuin-2. Finally, there is a fifth quarter stater (fig. 2, no. 15) of a type that was probably struck by the Remi, but which is also known from sites in southern Belgium. 3 TH E A N I M A L - H E A D E D G O L D B R A C E L E T A gold spiral bracelet with animal-head terminals represents a highly exceptional find from the Philippe ville hoard (fig. 3). It is made of a single twisted wire with two spiral-shaped windings and terminals in the form o f animal heads. The heads look ‘horsey’ (see the long snout and the nostrils of the animals) and they are bordered by a fine of pearls. The weight is 44,36 g, the outer and inner diameter respectively c. 63 and 58 mm. Although we have not been able to investigate the metal composition of the bracelet, its yellowish colour points at a high gold content, which fits in well with the evidence available for the other gold ornaments discussed in this volume.6 Gold spiral-shaped bracelets with zoomorphic terminals in varying styles are known from a wide area including the Greek-Hellenistic world, ancient Persia and the Skythian region in Eurasia. They are generally dated in the second half o f the 1st millennium BC.7Torques and armrings with zoomorphic terminals also occur in the La Tene cultural area, although they are relatively rare here. Examples are a pair of gold spiral bracelets from tomb 23 at Montefortino (I), a bronze torque supposedly fromVieille Toulouse (F) in the British Museum, and the bronze animal-headed spiral bracelet from Snailwell (GB).8 4 Roymans/Scheers, this volume, 36 ff., Appendix 2. Montefortino gold bracelets (3rd century BC): Moscati 5 Cf. Roymans/Scheers, this volume, 5, fig. 5. et al. 1991, 234. VieilleToulouse torque: Stead/Rigby 6 Cf.Wbuters, this volume, 215 ff., Section 6. 1999, ML 1708, 71 and fig. 93. Snailwell bracelet (1st 7 See for example the gold spiral bracelets with animal- century AD?): Stead 1996, 43, fig. 45. See also the gold head terminals from the so-called Oxus treasure in the hoards w ith spiral-shaped bracelets with snake-head ter collection o f the British Museum (Afghanistan, 5th/4th minals found recently in Rom ania and dated to the 1st century BC; Dalton 1964), and from the Skythian area as century BC: Constantinescu et al. 2010; Spanu/Cojocaru presented in the catalogue edited by Cahen-Delhaye 1991. 2009. We thank our colleagues Vincent Megaw, Fraser III 0 Fig. 3. Twisted gold bracelet with animal-headed terminals from the Philippeville hoard. Scale 2:1. 1 12 1 2 cm 2 cm j Fig. 4. Twisted silver bracelet with terminals in the form o f bird heads from Eijsden-Margraten (NL). Photo Municipality o f Maastricht. Scale 2:1. These latter parallels are in bronze and have a wire with a rounded or plano-convex section instead of a twisted wire. However, twisted bracelets are a well-documented phenomenon in Northwest Europe from the later Bronze Age onwards up into the Roman period.9An interesting new Late Iron Age find from the southern Netherlands is a twisted silver bracelet with terminals in the form of bird heads (fig. 4).10 We may conclude that the Philippeville armring with its combination of spiral-shaped windings, the sin gle twisted wire, and the zoomorphic terminals represents a rather unique ornament for the Northwest European Iron Age. There is, however, no reason to assume that is was imported from another cultural area. O n stylistic grounds the ornament cannot be dated more precisely than in the La Tene period. However, the association with the coins clearly suggests a date in the Late La Tene phase. H unter (National Museums o f Scotland) and Aurel 9 10 The (unpublished) bracelet was found in 2011 by Danny R ustoin for providing us with useful information and van Krevel in a field in the municipality o f Eijsden- references. Margraten. The same site also produced a fragment o f a See for example the twisted gold spiral bracelet from glass LaTène bracelet and some objects from the Rom an Toulouse (?) (F), dated to the Late Bronze Age: Ugaglia period. We thank W im Dijkman (Municipality o f Maas et al 2002,77. tricht) for providing us this information. C O N C L U S I O N S 4 Qua coin composition the hoard from the vicinity of PhilippeviUe belongs to the Fraire/Amby horizon. Coins that are typical for the younger phases of this horizon (like the Nervian class lib staters of the ‘Heers variant’ and Eburonean staters struck with younger dies of class la) are absent. We are probably dealing here with a hoard (or a combination of two hoards) that was deposited in the early phase of the Fraire/Amby horizon, or the early 50s BC." This is further confirmed by the limited traces of wear of the coins from PhilippeviUe, which is similar to the coins present in the hoards of Fraire andThuin.112 R E F E R E N C E S Cahen-Delhaye, A. (ed.), 1991: Het goud der Skythen. Schatten uit de Hermitage, Leningrad, Brussel (Koninklijke Musea voor Kunst en Geschiedenis). Constantinescu, B./E. Oberlander-Tânoveanu/R. Bugoi/V Cojocaru/M. Radtke, 2010:The Sarmizegetusa bracelets, Antiquity 84,1028-1042. Dalton, O.M., 1964: The treasure of the Oxus with other examples of early oriental metalwork, London. Moscati, S., et al. (eds), 1991: The Celts, Milano. Scheers, S., 1977: Traité de numismatique celtique II. La Gaule helgique, Paris. Spânu, D./V. Cojocaru, 2009: The Dacian hoard from Bucuresti-Herâstrân. Archaeological and archaeometallurgical approaches, Materiale §i Certâri Archeologice V, 97-116. Stead, I., 1996: Celtic art in Britain before the Roman conquest, London. Stead, I.M ./V Rigby, 1999: The Morel collection. Iron Age antiquities from Champagne in the British Museum, London. Ugagha, E., et al.: L’Or de Tolosa, Toulouse. 11 See Roymans/Scheers, this volume, 7 ff., Section 4. 12 Compare the coin plates o f the hoards from Fraire and Thuin elsewhere in this volume. ÏI4 A P P E N D I X i: D E S C R I P T I V E P H I L I P P E VILLE LIST OF T HE C O I N S F R O M THE H O A R D The numbering corresponds to the numbering of the coins in figures 1 and 2. O = obverse die; R = reverse die. The hoard was sold by Jean Elsen auctioneers and is nowadays dispersed over several private collections.12 No. weight (g) die nos. references g o ld staters type Scheers 31, class la, Eburones 1 5.36 0 5- R7 Elsen, 110,10-9-2011, no. 79 2 5.45 0 1- R 1 Elsen, 109, 18-6-2011, no. 28 3 5.51 0 2- R2 Elsen, 108, 12-3-2011, no. 19 4 5.62 0 8 - R 10 Elsen, 108, 18-6-2011, no. 29 g o ld staters type Scheers 29, class la, N e rvii 5 5.86 0 6-R6 Elsen, 108,12-3-2011, no. 22 g o ld staters type Scheers 29, class lia , N e rvii 9 5.85 0 56 - R 91 10 5.79 0 58 - R 98 g o ld staters type Scheers 29, class lib , N e rvii 6 5.92 0 65-R111 7 5.81 0 69 - R 119 8 5.84 0 69 - R 119 Elsen, 108,12-3-2011, no. 23 Elsen, 109,18-6-2011, no. 30 quarter staters type Scheers 152-V, class I, N e rvii 11 0.91 Elsen, 108,12-3-2011, no. 25 12 0.74 Elsen, 108,12-3-2011, no. 26 13 1.04 Elsen, 109,18-6-2011, no. 32 quarter staters type Scheers 152-V, class III, N e rvii 14 1.13 Elsen, 109,18-6-2011, no. 31 quarter sta te r type Scheers 152-111, Hemi 15 1.24 Elsen, 108,12-3-2011, no. 27 12 W here applicable, the auction catalogue that mentions isolated items from the hoard is cited for each coin. US Remains o f a disturbed gold hoard at O rp-le-Grand? Luc van Impe / Simone Scheers 1 2 3 4 Discovery of the gold finds and site description The gold torque The gold stater of the Eburones A ploughed hoard? References I D I S C O V E R Y OF T H E G O L D F IN D S A N D SITE D E S C R I P T I O N Over a five-year period in the 1990s, several Late Iron Age metal objects were discovered by a private individual in a field at the site o f a Gallo-Roman villa (?) south of the hamlet of Marilles (municipal ity of Orp-Jauche) (fig. I).1 The finds include a stater ascribed to the Eburones and four fragments of a gold torque, which were scattered over a distance of 70 to 80 m.The site is situated at a height of about 100 m on the Brabant plateau, which is made up of pleistocene loess soils. The site is on the edge of a gendy sloping depression, which is partially covered with colluvial deposits. Like other depressions in the vicinity, it is drained by a small stream, Le Mosembais, which eventually feeds into the Kleine Gete to the north.12 On the southern edge of this depression are several clay outcrops, which are no doubt linked to the remains o f a Roman tile kiln at several hundred meters southeast of the villa site.3 Finds from the Late Bronze Age and the subsequent Iron Age are also well represented in the vicinity.4 In particular, we can report several find sites from the Late Iron Age (fig. 1). East of Marilles at ‘Haut Tiège’ is a find site that has yielded a Late La Tène horizon, represented by ceramics, fragments of glass bracelets, iron tools and an iron sword-shaped ‘currency bar’ of the Wérimont-La Tène-Manching type.5 Further eastwards setdement finds have been found near the ‘Chapelle Sainte-Adèle’ in Orp-le-Grand and also further south at Jauche.6 Best known is the settlement at ‘Le Tierceau’, about 3.5 km northeast of Marilles in the hamlet of Maret on the western slope leading down to the Kleine Gete. The site was the subject of an archaeo- 1 The find site is located near Nodrenge, in the vicinity gives an overview o f the area east and southeast o f O rp- o f Marilles, itself part o f the present-day municipality of Jauche towards Liège. Orp-Jauches (Prov. Brabant-Wallon). The finds described 2 3 Mercenier 1963; Cahen-Delhaye 1973, 9, no. 1; Mariën 1961, 145; Mariën 1970, 229. O n ‘currency bars’, see Baeyens 1960. Mariën 1970, 125-129; Kurz 1995, 16-21, 237-239; Joachim 2007, 37-38; Marquart 2010. Rem y 1977, 18 and PI. I, 19-21. A.-M. Herinckx, in Brulet 2008, 292 and fig. 13:12. 4 5 here are in private ownership. Cahen-Dalheye, 1973, 8-12, map fig. 3. Destexhe 1987 6 Mercenier 1965, 27; Cahen-Delhaye 1973, 9, nos. 2 and 3c. 117 Fig. 1 Location o f the find site o f the gold torque at Orp-le-G rand (star) and o f some other Late Iron Age settlements in the immediate vicinity. logical investigation in 1970 and once again in 1998 of the new high-speed railway line Knking Brussels, from the Late Bronze Age to the start of the Roman striking feature is the rectangular layout of a multiple 2 TH E G O L D as part of a preventive study for the construction Liège and Aachen. The settlement was inhabited period and was conspicuously rich in ceramics. A ditch system that encloses the entire settlement.72* T O R Q U E Four torque fragments have been preserved (figs 2 and 3). The distortion of one of the fragments and the breaks suggest that the torque was torn apart when the soil was worked, probably by ploughing.8 The breaks fit together well; only one of the two terminal knobs is missing, as well as perhaps a small fragment of the tube. Mariën 1961, 145; idem 1970, 229; Mercenier 1965; 8 We would like to thank G. Schalenbourg for the photos, Preud’homme 2001 and 2003; Cahen-Delhaye 1973 and and the people at VIOE (Flemish Heritage Institute), 1974; Preud’homme et al. 1999a and 1999b; Fechner Limburg and Flemish Brabant, especially A. Detloff for 1999; Fock et al 2008, 56,89,121. the drawing o f the find. n 8 Fig. 2. Fragments o f the gold torque o f Orp-le-Grand. Scale 1:1. Photo G. Schalenbourg. The torque consists of a hollow tube with a diameter of 6 to 6.4 mm, constructed from a 0.5 mmthick gold plate. As is customary with other torques, the interior seam is open along the full length. On one of the fragments, the edge of the interior seam has a few fine indentations. The fragments are 13.5, 5.8, 9.4 and 9.6 cm long, giving a total length of 38.3 cm. Based on this figure, and the curvature of the pieces, the interior diameter o f the torque can be calculated at about 12 cm. The weights of the respec tive fragments are 21, 8,14 and 18 g, or 61 g in total. Because one of the two terminal knobs is missing, and possibly also a small piece of the tube itself, the original weight would have been several grammes more. A small pearl ring has been pushed over both slightly narrower ends of the hollow tube. This also served to conceal the join with the terminal knob. The only preserved terminal knob is about 9 mm high and is bell-shaped. Hollow gold torques were worn in different periods. Various hollow torques, made from thin gold plate, are already known from the Early Iron Age, including from graves traditionally identified as aristo cratic.9 Torques with terminal knobs are found primarily in the subsequent La Tene period. Among the group of La Tene torques we encounter few parallels for the torque from Orp with its small terminal 9 See for example D ehn 1996. Fig. 3. Drawing o f the torque fragments o f Orp-le-G rand (scale 1:1) and detail o f the terminal knob (scale 2:1). knob.There is a parallel between the fastening system of this torque and that of several other gold exam ples. The preserved terminal knob ends in a small round plate with a rectangular aperture. This is the ‘female’ knob of the fastening; its male counterpart with the clasp is missing. The shape and positioning of the aperture shows that the clasp matched those of the torques from Niederzier, Beringen, Thuin 101 and Soucy-Montgobert, and not the T-model that we find on many other gold torques. It is a type of fastening that occurred from the second half of the 3rd century BC.U Small dentate or pearl rings to support the terminal knobs or to separate different parts of the knobs were also common, although this practice spanned a longer period. We see examples of such pearl rings in the older torque from La Tene 12 and the earlier ‘baroque’ terminal knobs on the torques of Civrayde-Touraine.13 Dentate plates and bobbins are more common on other torques, such as those o f Nied erzier and Beringen .14 The small size of the terminal knob is highly unusual in the known data set of 10 See Scheers et al., this volume, figs 7 and 8. 2003, 215-231, Abb. 60,2. 11 Hautenauve 2005,76-77. 13 Hautenauve 2005,75, fig. 42. 12 Vouga 1923, 67-68, fig. 8; Muller 1991,78, fig. 19; Adler 14 Van Impe et al. 1997/1998, figs 24-25, 47-48. 120 gold torques, but it does have parallels with the solid filamentary bronze examples, such as the one from Manching.15 A litde imagination suggests a match with the stylised, blunt, oversized terminal knobs present on several sculptures or figurative images on vessels. We are thinking first and foremost of a series of scenes on the Gundestrup cauldron, although this comparison is not entirely satisfactory.16 There are definite parallels, however, with sculptures such as a figurine from Reims P e p . Marne, F.),17 or the figurines of Euffigneix (Haute-Marne, F.)18 and Paule (Cotes d’Armor, F.).19 A stylistic dating places the Euffigneix idol in the late 1st century BC or early 1st century AD, while the figure from Paule is dated earlier - to the 2nd century BC .20 The torque on the figure of Euffigneix is covered with punched dots that are similar to the dotted decoration on the terminal knob from Orp. Furthermore, the arrangement of the decoration on the knob from Orp in the form of a garland is reminiscent of that on a ceramic bowl from Bussy-le-Chateau (Marne, F.),21 which has a La Tine finale II dating. Taking into account all the above comparisons and suggestions, we believe that we can adhere to the second half of the La Tene period as the date for the torque from Orp. We can also say that the gold torque with the smaller terminal knob is no longer so very far removed from the smooth or twisted bronze armbands with small blunt terminal knobs that we know from Early Roman grave assemblages and cult places in the Low Countries .22 Here too we see a punched, although more or less linear, decora tion. Some of the terminal knobs also have an articulated construction. 3 THE GOLD STATER OF THE EBURONES An Eburonean stater of Scheers type 31, class la was found several dozen metres from one of the torque fragments. It features a triskeles on the obverse and a horse facing left on the reverse.23 The simplified reverse and the presence of a circle with a central ring below the horse show that it is a later emission. The stater is related to coins from the hoards of Amby and Heers. Interestingly, the obverse die O 8 links it to a large group of die-match coins from the Heers hoard (nos. 25-116), although this obverse die is combined with a different reverse die ‘4- R 28.24 Fig. 4. Gold stater o f the Eburones from O rp-leGrand. Scale 2:1. Weight 5.43 g. 15 Adler 2003, 240-244, Abb. 75,7. 21 Charpy/Roualet 1991, 240, no. 290. 16 Kaul 1991. See the simplified typology o f these torques: 22 Massart 1997, 65-68, with for example fig. 2, 1-5 and Adler 2003, 81-83, Abb. 15. 77 Lambot 2006,235. 18 Duval 1977,192-193, fig. 199; Duval 1989,126-127, fig. 45; Hatt 1989,122-123, fig. 95. 19 Arramond et al. 1992. M enez/Arram ond 1997,139-143. fig. 4, 1-2. Sas/Thoen 2002,170-173, nos. 80, 82, 85-87 (Wijshagen and Wijnegem); 174, nos. 90-91 (Maaseik); 174-175, no. 92 (Nijmegen). 23 Scheers 1977, pi. IX.254, Scheers type 31, class I. 24 See Roymans/Scheers, this volume, 35, Appendix 1. 20 See Van Impe et al. 1997/1998, 103-104. 121 A 4 P L O U G H E D H O A R D ? Given that the gold objects from Orp are disturbed finds, it is difficult to say anything about their original context in the soil. After all, this material came from the topsoil layer and would not have been discovered without the use of a metal detector. It is no longer possible to establish whether the Eburonean stater and the fragments of the gold torque were originally part of a gold hoard. Yet this remains a strong possibility, especially since it is difficult to imagine that the gold torque and the gold coin were objects lost by chance. In the event of a ploughed coin hoard, we might have expected more coin finds from the site. However, these sites in the Belgian loess region have been subject to considerable erosion during the past two millennia. Pedological observations of the enclosed settlement at Le Xierceau, about 3.5 km northeast of the site of MariUes, have suggested a level of erosion of between 50 and 150 cm. The lowest value was established at the top of the plateau, in the middle of the protohistoric settlement, and the highest value on the steeper zones of the slope. This means that the original surface from the Iron Age and Roman times, together with all shallow traces of habitation, have disappeared and become absorbed into the colluvial deposits located lower down .25 As we know, the association of ornaments and coins is a common phenomenon in the Late Iron Age. Examples are the hoards of Beringen, Frasneslez-Buissenal and Niederzier. Torques, on the other hand, also occur in isolation or in hoards without coins, including in setdement contexts.26 If the Orp find is indeed a ploughed hoard, the dating of the Eburonean stater shows that it belonged to the younger phase of the Fraire/Amby horizon (i.e. the later 50s BC). It is equally possible, however, that the torque and the coin were deposited independently of one another. Whatever the case, they bear new witness to the rich gold circulation in the Late Iron Age in central Belgium. REFERENCES Adler, W , 2003: Der Halsring von Männern und Göttern. Schriftquellen, bildliche Darstellungen und Halsring funde aus West-, Mittel- und Nordeuropa zwischen Hallstatt- und Völkerwanderungszeit, Bonn (Saarbrücker Beiträge zur Altertumskunde, 78). Arramond, J.-Ch./Y. Menez/Cl. Le Potier, 1992: Le camp de Saint-Symphorien à Paule dans les Côtes d’Armor, Saint-Brieuc (Archéologie et travaux routiers départementaux). Baeyens, L., 1960: Carte des sob de la Belgique-Texte explicatif de la planchette defauche 118 E/Bodemkaart van België- Verklarende tekst bij het kaartbladJauche 118 E, 1.1. Brulet, R ., 2008: Les Romains en Wallonie, Bruxelles. Cahen-Delhaye, A., 1973: Sondages dans un site d’habitat de l’âge du Fer à Orp-le-Grand, Bruxelles (Archaeologia Belgica 151). Cahen-Delhaye, A., 1974: Contribution à l’étude de la céramique d’habitat de l’âge du Fer en Hesbaye. Analyse typologique du « Tierceau » à Orp-le-Grand, Helinium XIII, 235-260 (= Archaeologia Bel gica 156). Charpy, J.-J./P. Roualet, 1991: Les Celtes en Champagne. Cinq siècles d’histoire, Epernay. Degryse, H./K. Fechner, 2001 :Archeopedologie, op zoek naar de mens in het landschap, in A. Schryvers.L. Van Impe (eds), Op het spoor van het verleden. Archeologie op de Hogesnelheidslijn, Leuven, 30-31. Dehn, R ., 1996: Ein Fürstengrab der späten Hallstattzeit von Ihringen, in S. Plouin/C. Dunning/P. Jud (eds), Trésors Celtes et Gaulois. Le Rhin supérieur entre 800 et 50 avant J.-C., Colmar, 113-118. 25 Fechner 1999; Preud’homme 2003. Similar érosion values were established at prehistorie and later sites elsewhere along the route o f the high-speed railway line. Cf. 122 Degryse/Fechner 2001; Schryvers et al. 2001. 26 For a b rief OverView, see Van Impe et al. 1997/1998, 92-99. Destexhe, G., 1987: La protohistoire en Hesbaye centrale. Du Bronze final à la romanisation, Archéologie Heshignonne 6 . Duval, A., 1989: L ’art celtique de la Gaule au Musées des Antiquités nationales, Paris, (Monographie des musées de France). Duval, P.-M., 1977: Les Celtes, Paris (L’Univers des Formes). Fechner, K., 1999: Hélécine et Orp-Jauche/Orp-le-Grand. Contribution de la pédologie à l’interprétation des fosses et des fossés de l’Age du Fer aux lieux-dits « Chapeauveau » et « LeTierceau », L’Archéologie Wallonne. Chronique 7, Namur, 6 - 8 . Eluère, Chr. 1986: L’orfèvrerie pré-romaine au musée de Cluny, La revue du Louvre et des Musées de France 36, 97-106. Fock, H ./C l. Goffioul/D. Bosquet, 2008: Les Traverses du Temps. Archéologie et TGV, Namur. Göbel, J./A. Hartmann/H.-E. Joachim/V Zedelius, 1991: Der spätkeltische Goldschatz von Niederzier, Bonner Jahrbücher 191,27-84. Hatt,J.-J., 1989: Mythes et dieux de la Gaule. I. Les grandes divinités masculines, Paris. Hautenauve, H., 2005: Les torques d’or du second Age du Fer en Europe. Techniques, typologies et symbolique, Rennes (Travaux du Laboratoire d’Anthropologie 44). Hautenauve, H ./C. Besson, 1998: Persistance de techniques d’orfèvrerie. Le torque celtique de Soucy (Aisne) et les bracelets gallo-romains de Poiré-sur-Vie (Vendée), Revue archéologique de l’Ouest 15,141150. Joachim, H.-E., 1991: The votive deposit at Niederzier, in S. Moscati/O.-H. Frey/V. Kruta/B. Raftery/M. Szabo (eds), The Celts, Milano, 532. Joachim, H.-E., 1999: Die Eburonen - Historisches und Archäologisches zu einem ausgerotteten Volks stamm caesarischer Zeit, in G. von Büren/E. Fuchs (eds), Jülich. Stadt - Territorium - Geschichte. Fest schrift zum 75jährigen Jubiläum des Jülicher Geschichtsvereins 1923 e.v.,157-170. Joachim, H.-E., 2007: Die späte Eisenzeit am Niederrhein, in M. Hegewisch, Krieg und Frieden. KeltenRömer-Germanen, Bonn/Darmstadt, 48-58. Joachim, H .-E ./V Zedelius, 1980: Ein bedeutender keltischer Verwahrfund aus Niederzier, Kr. Düren, Gymnasium 87, 205-210. Joachim, H.-E., 2007: Die Datierung der jüngerlatènezeithchen Siedlung von Niederzier-Hambach im Kreis Düren, BonnerJahrbücher 207 (2009), 33-51. Kaul, Fl., 1991: Gundestrupkedlen. Baggrund og billedverden, Kobenhavn. Kurz, G., 1995: Keltische Hort- und Gewässerfunde in Mitteleuropa. Deponierungen der Latènezeit, Stuttgart (Materialhefte zur Archäologie in Baden-Württemberg 33). Lambot, B., 2006: La Champagne et les Rèmes, inV. Kruta/M. Licka/J. Cession-Loupe (eds), Celtes: Bel ges, Boïens, Rèmes, Volques..., Morlanwelz, 222-241. Mariën, M.-E., 1961: La Période de La Tène en Belgique. Le Groue de la Haine, Bruxelles (Monographies d’Archéologie Nationale, 4). Mariën, M.-E., 1970: Le Trou de l’Ambre au Bois de Wérimont Eprave, Bruxelles (Monographie d’Archéo logie Nationale, 4). Marquart, M., 2010: Keltenland am Fluss, in M. Marquart (ed.), Keltenland am Fluss. Die Kelten im RheinMain-Gebiet, Rahden,13-18. Massart, CL, 1997: Les bijoux déposés dans les tombes à l’époque romaine en Belgique, in J. Moulin/A. Cahen-Delhaye (eds), 1997: La parure dans nos régions. De la Préhistoire au Moyen Age (Actes du Colloque de la F.A.W. Mariemont le 15 novembre 1997), Vie Archéologique 48, Mariemont, 63-80. Menez,Y./J.-Ch. Arramond, 1997: L’Habitat aristocratique fortifiée de Paule (Côtes d’Armor), Gallia 54 (1998), 119-155. Mercenier, J. & L., 1963: Marilles (Brabant). Découverte d’un fond de cabane de l’extrême finale de la Tène III au « HautTiège », Bulletin du Cercle Hesbaye-Condroz IV 54-62. 123 Mercenier, J. & L., 1965: Orp-le-Grand aux temps anciens, Bulletin du Cercle Hesbaye-CondrozV, 27-35. Müller, F., 1991: La période de LaTène. Le monde se met en mouvement, in A. Furger/F. Müller (eds.), L’Or des Helvètes. Trésors celtiques en Suisse, Zürich,71-82. Preud’homme, D., 2001: L’Habitat du lieu-dit Le Tierceau à Orp-le-Grand (Prov. Du Brabant Wallon), in C. Bellaire/J. Moulin/A. Cahen-Delhaye (eds), Guide des sites préhistoriques et protohistoriques de Wal lonie, Namur (Vie Archéologique-Numéro spécial,) Preud’homme, D. 2003: Orp-Le Tierceau, in J. Bourgeois/I. Bourgeois/B. Cherretté (eds), Bronze Age and IronAge communities in North-Westem Europe, Brussel, 253-255. Preud’homme, D./H. Fock/D. Bosquet/Cl. Goffioul, 1999a: Fouille d’un site d’habitat de l’Âge du Fer à Orp-Jauche, au lieu-dit Le Tierceau (Bt. w.), Lunula. Archaeologia ProtohistoricaVll, 62-67. Preud’homme, D./H. Fock/D. Bosquet/Cl. Goffioul, 1999b: Orp-Jauche/Orp-le-Grand. U n habitat de l’Age du Fer au lieu-dit « Le Tierceau » à Maret, L’Archéologie Wallonne. Chronique 7, 9-12. Remy, H., 1977: Les villas romaines de Jodoigne et Saint-Jean-Geest (Fouilles de J. Breuer en 1915-1916), Bruxelles (Archaeologia Belgica 195). Sas, K./H.Thoen, 2002: Schone Schijn. Romeinse juweelkunst in West-Europa - Brillance et Prestige. La joaillerie romaine en Europe occidentale, Leuven. Scheers S., 1977: Traité de numismatique celtique. IL La Gaule Belgique, Paris (Annales littéraires de l’Uni versité de Besançon 195). Schryvers, A./H. Degryse/K. Fechner/J. Bastiaens/P. Cosyns, 2001: Romeinse kuilen in Landen-Populieren, in A. Schryvers/L.Van Impe (eds), Op het spoor van het verleden. Archéologie op de Hogesnelheidslijn, Leuven, 121-141. Ugaglia, É., (ed.), 2001: L’Or de Tolosa, Toulouse. Van Impe, L./G. Creemers/R. Van Laere/S. Scheers/H. Wouters/B. Ziegaus, 1997/1998: De Keltische goudschat van Beringen (prov. Limburg), Archéologie in Vlaanderen 6, 9-132. Vouga, P, 1923: Monographie de la station publiée au nom de la Commission des Fouilles de La Tène, Leizpig. 124 T he gold hoard o f Heers Simone Scheers / Guido Creemers 1 Discovery of the hoard and site description 2 Landscape and habitation in the Late Iron Age 3 Composition of the hoard 3.1 The triskeles staters of the Eburones 3.2 The epsilon staters o f the Nervii 3.3 The P o t t in a stater of the Treveri 3.4 The gold stater ‘with star’ 3.5 The ‘eye’ stater of the Remi 3.6 The coin flan 4 Dating of the hoard 5 Late Iron Age coin finds in the Tongres/Maastricht area 6 Archaeological and historical interpretation of the hoard References Appendix 1. Descriptive list of the coins from the Heers hoard Appendix 2. Descriptive list of Late Iron Age coins (until 30 BC) from the Tongres/Maastricht area Plates I DISCOVERY OF THE HOARD AND SITE DESCRIPTION1 In autumn 1997 some trees were being cleared in an old orchard atVechmaal-Bornebeemden (munici pality of Heers) about 5 km southwest of Tongres (fig. 1), on a terrain that slopes gently toward the northwest. R af Janssen, the owner of the land, was pulling the trees out together with their roots, leaving behind holes in the ground. While digging, he spotted a glistening gold disc, which later proved to be a gold stater of the Nervii. Over the next few days he collected a dozen new coins in this way. In 1998, more trees were cleared in the eastern, most elevated section of a neighbouring piece of land, but Janssen did not come across any more finds. From 1998 to the end o f2001, he planted corn on the first field, and then established an orchard there (fig. 2). In the meantime, having contacted several amateur archaeolo gists and meeting the first author, Janssen was well aware that he had stumbled across a significant Late Iron Age coin hoard. He purchased a metal detector to make sure that he would not miss any more finds. His first intensive prospection in 1999 yielded 45 gold coins in the space of just one and a half hours. He continued to carry out prospections on a regular basis, each time finding one or more coins. After 2001, he was still finding coins on the site, always in the same zone. We would like to thank the following people for their Heesch, Luc Van Impe, Elke Wesemael, Marijke Willaert assistance in realising this study: Eddy Daniels, Jean-Marc and Ferdy Willems. Doyen, Tim Vanderbeken, Alain Vanderhoeven, Johan Van 125 Fig. 1. Topographic situation o f the Heers hoard find spot. R ed star: coin hoard; red square: R om an villa. As the above demonstrates, the coins were found dispersed across the northern zone of the two reported plots of land. The finder believed he could discern a roughly triangular distribution pattern, with sides measuring approx. 40 x 40 x 30 m. Within this pattern was a dense concentration of about 3 x 3 m, right on the boundary between the two plots (fig. 3). The fact that the coins were fairly well dispersed across the site shows that the original coin hoard had been almost entirely incorporated in the plough layer. Because the owner felt that the coins should be kept in the region, he contacted the Provinciaal Munt- en Penningkabinet at Tongres in autumn 2000. Ninety-four coins were subsequently purchased. Shortly thereafter, the owner found some more coins.The complete hoard, 158 coins in total, is currendy held in the collection of the Gallo-Roman Museum. This was clearly a coin hoard that had been disturbed by ploughing and had ended up scattered through the soil. The situation is probably identical to that of Maastricht-Amby.2 No other contempo raneous finds are known in the immediate vicinity, although the remains of a Rom an villa have been found about 300 m away. 2 L A N D S C A P E A N D H A B I T A T I O N IN T H E LATE I R O N AGE The find spot of the hoard of Heers is situated about 1500 m east of the village ofVechmaal. The sur rounding area is a typical Haspengouw landscape with fertile loess plateaus intersected by small valleys that are partially filled with colluvium (fig. 1). The site is located in the watershed of the Mombeek H which rises about 800 m to the northeast —and the Herk, which has its source in the centre of the present-day Vechmaal, about 1200 m west of the site. The local toponym ‘Bornebeemden’ is therefore a significant one. This area also features the watershed regions of the Scheldt (Herk and Mombeek) and Meuse basins (the Jeker).The region is characterised by fertile loess soils on the plateaus and slopes, with colluvial soils of considerably lower agricultural value on the valley. If we concentrate on the find spot of the coin hoard 126 Fig. 2. View o f the orchards at the site where the Heers hoard was found. (Bornebeemden/Flikkenberg vicinity; fig. 1), we observe that the subsoil on the higher parts is formed by a loess soil which gradually becomes covered with colluvium as we move towards the adjacent val ley to the west of the site. Several dozen metres to the east and south of the site is loamy, sandy ground with brown podzol-like soils o f litde agricultural value.3 In 2002 Jan Roymans and Dieter Demey (both RAAP, Netherlands) conducted a small coring survey at the site. Four core samples (PI —P4) were taken in a row along the boundary between the two plots (fig. 3).They mainly show a colluvial layer above the loess deposit. The area was also supplemented in places by soil brought in from elsewhere. At the place identified by Janssen as the central zone of the coin hoard, the original loess soil could be identified at a depth of about 35 cm beneath the present-day surface. With regard to settlement history in the Bronze and Iron Ages, we are better informed about burial practices and setdement patterns in the Kempen region and Meuse valley than in the Flaspengouw in Limburg.4 Nevertheless, contemporary archaeological inventories and regional overviews5 show that setdement must also have been fairly intensive in Haspengouw during the Metal Ages. Unfortunately, because of a dearth of studies giving a more precise dating for these setdements, we still have no clear picture of the habitation pattern during the Late Iron Age. We are much more familiar with Roman-period remains in the Haspengouw region .6 The many remnants of villas and monumental tumuli are usually explained by the presence of fertile loess soils and the proximity of the Roman town of Tongres. In recent years it has emerged that the region was also Roymans/Dijkman, this volume. For example, Centrale Archeologische Inventaris Vlaan Bayens 1959. deren; Creemers/Van Impe 2009. Bauwens-Lesenne 1968; Van Impe et al. 1997a; idem Bauwens-Lesenne 1968;Brulet 2008;M ariën 1980; Van- 1997b; Van Impe et al. 1997/1998. derhoeven 1996;Vanvinckenroye 1985. 127 Fig. 3. Distribution pattern o f the coins from the H een hoard. Dark green: densest concentration o f coins; light green: lower density o f coins; red: core sampling locations. fairly densely populated during the LaTene period .7 Since the 1960s, archaeological research at Roman villa sites in Haspengouw has brought to light several sites with traces of pre-Roman habitation. Some times we see continuity between Late La Tene and Roman setdement.8 In other instances, setdement at villa sites appears to have only begun in the Augustan period. Villa construction started in the middle of the 1st century AD. In the Meuse valley in Belgium as well, some Roman sites have been shown to have predecessors from the Late Iron Age.9 We find a comparable situation in the neighbouring region to the east,10 where the scarce information available suggests similar developments with regard to funer ary customs.11 These days ideas about setdement continuity are being confirmed by rescue excavations carried out during major infrastructural work, such as the high-speed rail fine or Distrigas/Fluxys.These excavations involve digging an enormous trench through the landscape across a great distance, with the route of the trench chosen at random in archaeological terms. The idea is to obtain a fairly ‘objective’ picture of human presence across the different periods. Systematic large-scale research in these projects enables us Corbiau 1997,224-262; Creemers/Van Impe 2009. 11 Roosens/Lux 1973;Vanvinckenroye 1987. Close et al. 1997; D e Boe 1971; De Boe 1974; De B oe/ 12 Creemers/Masson-Loodts 1999; Ervynck et al. 2000; I n ‘t Van Impe 1979; Marchai 1997;Vanderhoeven 1996,222- Ven/De Clercq 2005,183-213; Oost et al. 2001; Pauwels 223;Vanvinckenroye 1988;Vanvinckenroye 1997. et al. 2002 a/b; Wouters et al. 2002 a/b. De Boe 1987;Van Dierendonck et al. 1987. De Grooth 1987; Willems 1987. 128 13 In ’t Ven/De Clercq 2005,107-182; Opsteyn et al. 1999; Van Impe et al. 2001. Fig. 4. The coin hoard from Heers. to discover sites and off-site phenomena that have left less obvious traces than, say, Roman villas. Once again, we find that sites from the Bronze and Iron Ages occur frequently in South Limburg ,12 the eastern part of Flemish Brabant,13 and the adjacent Walloon Brabant.14 All these observations tie in with data obtained in recent years through numismatic research.15 The distribution of Celtic coins from the period up until 30 BC (fig. 6 ) indicates that the region under study was probably situated in the centre, indeed the southern core area, of Eburonean territory. Given this background, it is not surprising that a coin hoard like that of Heers should surface in the region. 3 COMPOSITION OF THE HOARD The inventory currently comprises 157 staters and an unstruck flan, which were found scattered across the field (fig. 4).The hoard may originally have contained more coins.There are five different coin types (fig. 5), with the 116 staters of the Eburones making up the majority. The hoard also contains 38 staters of the Nervii, one stater of the Treveri, one stater ‘with star’ (Bellovaci?) and one stater of the Remi. An unstruck flan was found a little outside the coin concentration, but was probably still part of the hoard. 14 Bosquet/Fock 1997; Close et al. 1997; Frébutte/Loodts 1999; Preud’honime et al. 1999. 15 Scheers 1996 a and b; Creemers/Scheers 2007; Roymans/Aarts 2009; Roymans/Scheers, this volume, 1 ff. 129 Fig. 5. Coin types represented in the gold hoard o f Heers. 1. Eburonean stater Scheers 31, class I (no. 1); 2. Nervian stater Scheers 29 (no. 138); 3. coin flan (no. 158); 4.Treveri P ottina stater (no. 157); 5. R em i eye stater (no. 156); 6. stater ‘with star’ (no. 155). 3.1 THE TRISKELES STATERS OF THE EBURONES It is hardly surprising that the Eburonean staters, all of Scheers 31, class la, make up almost 74% of the coin hoard. They were found in what is presumed to be the southern core of Eburonean territory, the region around the later Aduatuca Tungrorum. All staters belong to class la. They feature on the obverse a triskeles turning left, with one arm terminating in an elongated triangle and the two other arms in a circle. The reverse shows a horse prancing to the left, with three dots above it arranged in a triangle.16 The weights vary between 5.60 and 5.79 g. The 116 staters can be divided into two groups. The first group of 23 coins come from 11 obverse and 17 reverse dies (nos. 1-23). Traces of wear are a clear indication that the coins had been in circulation.The second group (nos. 24-116) consists of 92 staters that were struck using the same die pair. One more stater can be added to this group; it comes from the same obverse die but a different reverse die was used. These coins obviously did not circulate. It is highly unusual to find so many coins —79% of the Eburonean staters —from the same die pair in a single hoard. We are therefore almost certainly dealing with a batch of coins that had come straight from the mint. 16 La Tour 1892, pi. XXXVI 8859; Scheers 1977, pi. IX.254, type Scheers 31, class I. 130 Fig. 6. Distribution o f Late Iron Age gold coins in the Tongres/Maastricht area. The numbering o f the find sites corresponds to the num bering in Appendix 2. For the coins from Berg, see fig. 7. 3.2 THE EPSILON STATERS OF THE NERVII The hoard contains 38 ‘epsilon’ staters of Scheers 29 type, classes I and II, which are ascribed to the Nervii. They make up 24% of the coin hoard. They feature on the obverse the highly stylised traces of a human head, facing right, and on the reverse a horse with a wheel above it. Two distinct groups can be identified. The first comprises eight coins of class I ,17 while the second contains one class Ila stater without stars18 and 29 class lib staters with stars.19 Notably, the eight earliest staters are very worn (nos. 117-124), with weights below the normal average of 5.63 to 5.82 g. It is clear that these coins circu lated intensively before being buried. The class Ila coin (no. 125) is well preserved and, weighing 5.78 g, is within the average weight range. The class lib coins (nos. 126-154) on the other hand are in mint condition.This also makes them heavier, between 5.76 and 5.92 g. Moreover, the 29 coins were struck with only five obverse dies. This small number, together with the exceptional state of preservation of this group, points to the coins having come almost directly to Heers from the Nervian mint.This is clearly the most recent group of coins; they are absent from the other hoards discussed here fiFraire, Philippeville 17 La Tour 1892, pi. XXXV 8760; Scheers 1977, pi. VIII.216; Roymans/Scheers, this volume, 4, fig. 3. 18 Class Ila was previously unknown. Almost all examples that make up this class come from the hoards o f Fraire and Thuin. 19 La Tour 1892, pi. XXXV 8746, 8755; Scheers 1977, pi. VI1I.217; Roymans/Scheers, this volume, 4, fig. 3. and the three hoards of Thuin. Although the Nervii were the western neighbours of the Eburones, the number of coins that can be ascribed to them is rather small. Their presence is nevertheless significant, as Heers is the only hoard in the Eburonean territory in which Nervian staters occur. 3.3 THE POTTINA STATER OF THE TREVERI The only Treverian stater (no. 157) features on the obverse an angular eye in profile, with the iris replaced by a wheel (Scheers 30, classV).The reverse shows a horse prancing to the left, surrounded by different ornaments. Below the exergual fine is the legend P o t t i n a .20 The coin is well preserved and seems to have spent only a short time in circulation. Its weight of 5.55 g is normal for these coins.The presence of this stater is most remarkable. The Treverian staters that occur in the territory of the Eburones do not usually carry an inscription.21 P o t t i n a staters were hitherto unrepresented in this region; their distribution area in Belgium had been confined to the southern part of the Ardennes.22 3.4 THE GOLD STATER ‘W I T H STAR’ This stater (no. 155), which weighs 5.82 g, features on the obverse a stylised head and on the reverse a prancing horse between two stars (type Scheers 25, class II).23 The obverse was struck with a highly worn die, while a good die was used for the reverse. Although our knowledge of this coin series is patchy, there is no doubt that the coins with swirling stars belong to the last gold emissions. The attribution is uncertain. In the past these staters were ascribed to the Veliocasses, later to the Bellovaci.The series ‘with star’ is closely linked to the cult place of Digeon 24 and has a strong local character. These staters are rare north of the Somme. In Belgium, there have been stray finds at Arc-Ainieres (Hainaut)25 and in Lebbeke (East Flanders).26 3.5 THE ‘e y e ’ STATER OF THE REMI The only stater of the Remi (no. 156) (Scheers 30, class la) features on the obverse an angular eye in profile with an oval iris, and with three stars in front of the eye. The reverse shows a horse prancing to the left among a group of motifs; the heart-shaped m otif with the V-shaped dots above the horse is characteristic of this coin series.27 The weight of 6.02 g is lower than the average weight for the series. The obverse of the stater was struck with a die that had been re-engraved, whereas the die of the reverse was severely damaged. In Belgium these staters occur as stray finds in the southern provinces of Hainaut, Namur and Luxemburg. However, an example was also found in Waremme (Liège), not far from Heers. These coins had previously been attributed to the Treveri, but the finds are concentrated in the territory of the Remi, and in particular in the oppidum ofVieux-Laon at Saint-Thomas (France, dep. Aisne).28This site is nowadays identified as the oppidum Bibrax from Caesar’s De Bello Gallico.29 20 La Tour 1892, pi. XXXVI.8825; Scheers 1977, pi. IX.230-231. 21 La Tour 1892, pi. XXXVI.8815; Scheers 1977, pi. IX.229; type Scheers 30, class IV. 22 Scheers 1977,436, fig. 85; Loscheider 1998,136-138. 23 La Tour 1892, pi. XXIX.7235; Scheers 1977, pi.VI. 162163 : type Scheers 25, class II, pi.VI, 162. 132 24 Delestrée 1997. 25 Scheers 1977, 365 26 Willems 2010. 27 La Tour 1892, pi. XXXV.8799 var.; Scheers 1977, pl.VIII. 222. 28 Delestrée 1997,109-111. 29 Lambot/Casagrande 1997,15-29; Lambot 2002. 3-6 THÉ COIN FLAN This coin flan (no. 158) was first cast and then hammered flat. Nevertheless, the deep splits on the edges appear to suggest that this is a struck coin that was then flattened. This was a means by which a worn or foreign coin could be destroyed and withdrawn from circulation. The resulting coin flan could then be restruck. The advantage of this procedure was that it eliminated a time-consuming stage of coin pro duction, namely smelting and pouring the flan, and no traces remained of the underlying coin type. In view of the weight and the metal composition (33% gold only), the coin may have been a stater of the Eburones.30 Three other unstruck gold coin flans (see p. 148, nos. 30-32) were discovered at Berg and at HeersVechmaal, in Zouwveld, several kilometres to the west. They are stray finds and, once again, were probably coins that had been withdrawn from circulation. 4 DATING OF THE HOARD The special significance of the hoard of Heers is its mixed character. Although isolated finds of Nervian coins are no exception in the Eburonean area, so far this is the only hoard in which staters of the Eburones and the Nervii occur together. Also unique is the presence of coins from various other Belgic tribes. The blend of coins from five different tribes allows us to establish chronological links between the different emissions. The hoard of Heers belongs to the same chronological horizon as the other coin hoards, discussed in this volume, from Belgium and the Netherlands. By comparing the hoards, however, we can glean information that allows us to pinpoint the date more precisely. The Nervian staters in the hoard of Heers belong to classes I and II, as is also the case with the hoards of Fraire and Thuin. The staters in the Heers hoard, however, show a distinctive pattern. The eight class I staters have die counterparts in the coin hoards of Thuin and Fraire, but are clearly heavily worn. They were unquestionably in circulation for quite some time. The presence of such worn coins in a hoard is in fact very rare indeed. By contrast, the only class Ha coin (no. 125) is well preserved. Chronologically, it is close to the 29 staters of class lib (nos. 126-154), which are in mint condition and originate from five obverse dies. This group is particularly elegant in style and is characterised by a tight composition. It is absent altogether from the hoards of Fraire and Thuin and is here referred to as the ‘Heers variant’ of class lib .31 The mint condition of the 29 staters from Heers and the small number of obverse dies used indicate that the coins went almost directly from the Nervian mint to the hoard.Their presence also demonstrates that this hoard is younger than the other hoards with Nervian coins. Only one other hoard is known for this group of coins. Found in 1846 at Aubigny-en-Artois (France, Pas-de-Calais), it comes from the ter ritory of the Atrebates.32 The details of this last coin hoard are incomplete. It appears to have comprised only 600 to 700 Nervian staters of class Ila and Ilb.The date at which it was buried must therefore have been very close to that of the Heers hoard. The staters of the Eburones make up almost three-quarters of the coins in the hoard of Heers, which is unsurprising given the presumed location in one of the core areas of the Eburonean territory. The old est group of 23 coins is well preserved, although they were clearly in circulation for some time. The long use of heavily worn dies is evident on some coins which are scarcely legible. This practice may indicate a crisis, with large quantities o f coins needing to be struck in a short time. This haste and the surge in 30 O n the metal composition, see Appendix 1, no. 158, and Wouters, this volume, 215 fif. and Appendix 4. 31 See Roymans/Scheers, this volume, 4, fig. 3. 32 Scheers 1977,874. Roland Delmaire has also noted other coin types, but the source used is unreliable. Delmaire 1983, 62; Delmaire 1994,161. 133 production will have meant that new dies could not be engraved quickly enough. Existing ones were therefore used as long as possible, well beyond the generally acceptable stage of wear. The 92 staters of the second group (nos. 25-116), originating from the same die pair, are in mint condition and are without doubt the most recent Eburonean coins in the hoard. They make up a batch that came almost straight from the mint and in this respect resemble the most recent group of Nervian coins from this hoard. The four Eburonean staters in the hoard of Fraire (nos. Fraire 1-4) do not constitute a sound basis for comparison. Nevertheless, we are able to say that they are contemporaneous with the early group of 23 staters from the hoard of Heers.The later die-match group is absent altogether. Thus the situation here is akin to that of the Nervian staters. On the other hand, both groups of Eburonean class la staters occur in the hoard of Maastricht-Amby. That hoard also contains Eburonean class lb staters, which are absent in both the Heers and Fraire hoards, indicating that the hoard of Maastricht-Amby is younger than those of Heers and Fraire. Although the three foreign staters (Remi, P o t t in a and the coin ‘with star’) in the hoard of Heers per mit fewer comparisons, they nevertheless offer significant elements for dating. No die association could be established for the Remi stater (no. 156), making it difficult to place this coin within the series.Typologically, it belongs to the plainest class with an undecorated eye.33 It has a low weight (6.02 g), below the average of 6.10 to 6.30 g, although this may just be chance. Most of the Remi staters found in Belgium belong to this class. The coin fromWaremme on the other hand has a decorated eye.34 This type occurs frequendy in the Treveri region and in the Belgian Ardennes, which was subject to powerful Treverian influence. This geographical distinction could indicate that Remi staters reached our country by various routes. As a con sequence, it is not certain that the Remi stater arrived in Heers together with the P o t t i n a stater. Nor could a die link be traced for the P o t t in a stater of the Treveri (no. 157). Aside from the careless minting, the sharply defined details prove that this coin had been in circulation only briefly or not at all. This is also evident in the coin’s weight of 5.55 g, which is at the upper end of the normal weight curve of 5.30 to 5.60 g for this series. No other coin hoard is known in which this type is associated with coins of other Belgic tribes. The presence of a P o t t in a stater at Alesia is therefore highly significant.35 It enables us to place the start and perhaps the greater part of the emission before 52 BC.This ties in with the dating of the staters of A r d a , which were struck on theTitelberg at the end of the 50s BC .36 The strong typological connection between the staters of A r d a and P o t t in a lead us to assume that the latter coins were struck at the same mint, but at an earlier date, because of their slightly higher weight and gold content. In Belgium this type of stater is normally only found in the extreme south of the province of Luxemburg, which belonged to the territory of the Treveri.This makes Heers the most northerly find site for the P o t t i n a type. The stater ‘with swirling star’ (no. 155) unquestionably belongs to the most recent issues of this coin series. This is also attested to by its rather low weight of 5.82 g. The weight curve of this ‘swirling star’ class ranges from 5.70 to 5.85 g, contrasting with the average of 5.85 to 6.00 g for the older classes of this type.37 No die identity could be established for this stater either. The dating of this class is uncertain, but given that it is more recent, it must belong in the later years of the Gallic Wars. This is confirmed by another hoard of a mixed nature, found in 2009 at Aumale, not far from Digeon, the core area for this type. Our knowledge of this hoard is unfortunately incomplete, although it certainly contained about 30 staters with swirling stars,38 about 30 eye staters of the Suessiones39 and about 500 uniface staters with 33 Scheers 1977,409, fig. 79, type 30, class la. 163 (horse right).They can be recognized by the small 34 Scheers 1977,409, fig. 79, type 30, class Ie. eye on the obverse and by small, compact stars with short 35 Fischer/Gruel 2001, no. 678 (Camp C). standing rays on the reverse. 36 Scheers 1977,414, type 30, class VI, pi. IX, 232. See also Loscheider 1998,139-140. 37 The older classes are Scheers 25 I (horse left) and II, 161, 134 38 Scheers 1977, pl.VI.162; La Tour 1892, cf. pi. XXIX, 7234-7235. 39 Scheers 1977, pl.VII.175; La Tour 1892, pi. XXXII 8020. two Ss.40These staters with two Ss were probably the most recent coin issue of the Ambiani. Their dating is still uncertain, but they definitely belong in the later years of the Gallic Wars. The hoard of Heers is thus a witness from the time of the Gallic Wars.41 More specifically, the above chronological argumentation suggests a dating in the later years of these wars. Given the absence of the youngest issues of the Eburones and Nervii, we suggest a burial date in the mid-50s BC. 5 LATE I R O N AGE C O I N F I N D S IN T HE T R I C H T A R E A (in collaboration with Linda Bogaert) T O N GRES- MAAS The Tongres/Maastricht region has traditionally been a focus of archaeological interest, thanks chiefly to the Rom an town of Tongres, capital of the civitas Tungrorum, which featured the fertile region of the Haspengouw loess belt at its centre. The wealth of knowledge built up over time concerning Roman settlement is in stark contrast to what we know about pre-Roman settlement. The many isolated coin finds suggest that the region must have experienced fairly intensive settlement in the Late Iron Age, but almost no excavations have been carried out regarding habitation in this period. The city of Tongres itself was not founded until about 10 BC and is therefore not of pre-Roman origin. Many coins from different periods have always been found in the region, resulting in the creation of large numismatic collections in the 19th century. Since the 1980s, many of these have been incorporated in the Provinciaal Munt-en Penningkabinet, now part of the collections of the Gallo-Roman Museum in Tongres. The emphasis in these collections was on collecting Roman coins. It was not until the publica tion of Simone Scheers’ reference work, Traité de numismatique gauloise (1977), that interest was awakened in Late Iron Age coinages. This led to many reported finds of Celtic coins and to the purchase of various coins by the Provinciaal Munt- en Penningkabinet. However, a systematic inventory of coins from Limburg and neighbouring regions —both excavated coins and stray finds by metal detectorists —failed to eventuate. Numismatic research by Simone Scheers42 in particular has demonstrated that coin use was in full swing in the Late La Tène period among the population of the region under discussion and the neigh bouring part of Haspengouw. An inventory of Late Iron Age coins in the Tongres/Maastricht area in 2 0 1 0 yielded many new finds.The results are compiled in a distribution map (fig. 6 ) and table (appendix 2, p. 147-151).43What stands out is the clustering of gold and silver coins around Roman Tongres, which was founded later (c. 10 BC).This clustering is o f course partly explained by the attraction exerted by Rom an Tongres and its environs on metal detectorists. Also notable is the cluster of finds in Berg, where Celtic coins had already been found in the early 20th century (fig. 7).44 In recent decades, however, no Scheers 1977, pl.VI.156; La Tour 1892, pi. XXV, 8704. ‘oak branch with acorns’ and a ‘curled snake’. It is almost Virtually the same conclusion had already been reached certainly a stater o f the Nervii, type Scheers 29 class I in the albeit less detailed reports on the hoard o f Heers. (6 g.), which eventually ended up in the Gallo-Roman See Creemers/Scheers 2007; Roymans 2004; Scheers/ Museum by way o f the D e Schaetzen collection: Scheers Creemers 2001; idem 2002a and b. et at. 1991, 30. See also Smeesters 1974. The gold stater Scheers 1996; supplemented by several new finds in o f the ‘M orini’, De La Tour 8717, which he also reports, Duurland 2000,15-16. appears to be a uniface stater type Scheers 24-1, which This inventory o f finds, made primarily by metal detec was also acquired by the Gallo-Roman Museum (see torists working in the area, was compiled by the provin Scheers et at. 1991, 27). According to Willy Vanvinck- cial archaeologist, Linda Bogaert. enroye’s notes, a silver Annarovecci coin was found by a Including Huybrigts 1914,125-126. This is a gold stater m ember o f Peuskens’ family. ‘o f more than 7 grammes’, found along the ‘O ude Steenstraat’. O ne side depicts a horse, the other a ‘faucille’, an 135 Fig. 7. Distribution o f Late Iron Age gold coins in the vicinity o f Berg, near Tongres. The numbering o f the find sites corresponds to the numbering in Appendix 2. less t-han 20-odd gold and silver coins have been collected.45 In the Late Iron Age, Berg must have had a supralocal function, perhaps as a cult site. The area immediately west ofTongres has also regularly yielded coin finds from the middle of the 1st century BC. In general, we see a clear predominance of Scheers 31 Eburonean staters in the Tongres region, although we also regularly encounter staters of foreign origin, as well as several silver rainbow cups of the triquetrum type that are probably also an Eburonean emission.46 Also notable is the fact that the Late Iron Age fortification at Kanne-Caestert47 on the Meuse river has so far not yielded a single coin, which suggests that it was only used in times of crisis and that for the rest there was no permanent settlement there. Caesar’s claim that he had practically wiped out the Eburones appears to be contradicted by the emis sion of silver A n n a r o v e c i coins, which were struck in this area in the period 50-30 BC. An elite with the capacity to issue coinage must still have been operating there shordy after Caesar’s conquests. It is prob ably no coincidence that most coins have been found in the immediate vicinity of AduatucaTungrorum, which was founded a little later.Various Bochum-type rainbow cups with a high copper content, which belong to this period, were also found in this area. It is these elite figures and their descendants who will have laid the foundation for the ethnogenesis of the Tungri and of the later expansion of Roman T o n g re s .48 45 To date, most finds have not been published.There is a brief report in Van Heesch 2005. Cf. also the maps in this article. 46 See the discussion in Roymans/Dijkman, this volume, 171 ff. 136 47 Roosens 1975 and 1976;Vanvinckenroye 2001;Verhoeven 2008. 48 O n the A nnaroveci coins: T hirion 1967; Scheers 1996; Roymans/Aarts 2009. Fig. 8. Distribution o f silver AN NAROVECI coins in theTongeren/M aastricht area.The numbering o f the find sites corresponds to the num bering in Appendix 2. The lack o f archaeological investigations into Late Iron Age sites in the region makes it impossible to place the stray coin finds in their contexts. The inventorisation and study of other significant find catego ries in this region, such as glass bracelets, also leaves much to be desired.49We are left with the impression that these finds tend to be concentrated in the Meuse region. In several instances it seems that we can speak of continuity between the pre-Roman and Roman periods at villa sites.50 With a few exceptions, such as the recently excavated traces of byre houses and four- and six-post granaries at Kesselt (Lanaken),51 the picture we have is largely fragmentary and the finds have not been sufficiently studied to demonstrate their usefulness in the context of this study. Further research is therefore needed to gain a clear picture of settlement during the Late Iron Age in Haspengouw (Limburg) and neighbouring regions. 49 For a first attempt at an inventory for the area east o f Tongres, see Duurland 2000, 15. 2002. 51 Pauwels et al. 2010. 50 Including Creemers/Van Impe 2009; Demey/Roymans 137 6 ARCHAEOLOGICAL OF THE AND HISTORICAL INTERPRETATION HOARD On the basis of their clustered distribution, their highly homogeneous composition and die analyses, the coins found at Heers can be viewed as the remains of a single coin hoard, which has been ploughed up in recent times. Given that it seems to be a one-off deposition at what was probably an uninhabited site, a Versteckdepot would appear to be the most plausible interpretation. There is nothing to suggest that it was a ritual deposition, although there is no certainty on this point. That would require a study of the archaeological context by means of a control excavation. Several springs in the vicinity (800 and 1500 m away; see above) are much too far away for this to have been a spring deposition. O f interest is the presence of the remains of a Roman villa in the immediate vicinity (300 m away). It remains to be seen whether this is the successor to a native settlement dating back to the Late Iron Age. The coin hoard of Heers should be dated to somewhere in the mid-50s BC. If there is a historical event behind the burial of this hoard, the Eburonean revolt in the winter of 54/53 BC seems to be the obvious candidate.52 The Eburones, Nervii andTreveri were all party to Ambiorix’s revolt and its aftermath. This dating ties in seamlessly with the chronological position of the Heers hoard, between those of Fraire/Thuin on the one hand and Amby on the other. In any event, the gold hoard of Heers will become an important point of reference for the study of native coin circulation and social relationships in the region during the latest part of the La Tene period. REFERENCES Bamps, C., 1886: Aperçu sur mes découvertes de monnaies et d’autres antiquités antérieures au moyenâge faites dans le nord, l’ouest et le centre de la province de Limbourg belge, Bulletin de la Section Scientifique et Littéraire de la Société Mélophiles de Hasselt, 23, 73-160. Bamps, C., 1887: Aperçu sur les découvertes d’antiquités antérieures à la domination romaine,faite dans le Limbourg belge, Hasselt (reprint). Bamps, C., 1889: Aperçu sur les découvertes de monnaies et d’autres antiquités antérieures au moyen âge faites dans la province du Limbourg belge, Bulletin de la Section Scientfque et Littéraire de la Société Mélophiles de Hasselt, 25,195-258. Bauwens-Lesenne, M., 19 6 8 : Bibliografisch repertorium van de oudheidkundige vondsten in Limburg, behoudens Tongeren-Koninksem, Brussel (Oudheidkundige répertoria Reeks A: Bibliografische répertoria VIII). Bayens, L., 1959: Bodemkaart van België.Verklarende tekst bij het kaartblad Borgloon 106 E, Brussel. Bosquet, D./H. Fock, 19 9 7 :Waremme. Vestiges de l’Âge du Fer au lieu dit ‘La Costale’, in M .-H. Corbiau (ed.), Le patrimoine archéologique de Wallonie, Namur, 230-232. Brulet, R., 2008: Les romains en Wallonie, Bruxelles. Chalon, R., 1862: Curiosités numismatiques. Monnaies, médailles et jetons rares et inédites. Quatrième article, Revue belge de numismatique et de sigillographie 18,106-122. Close, F./M. Gustin/J.-P. Marchai, 1997: Bassenge, Eben-Emael. U n sauvetage archéologique et la décou verte d’occupations anciennes à ‘Int’ les deux voyages’, dans la carrière CBR du Romont, in M.-H. Corbiau (ed.), Le patrimoine archéologique de Wallonie, Namur, 345-347. Corbiau, H. (ed.), 1997: Le patrimoine archéologique de Wallonie, Namur. Creemers, G./I. Masson-Loodts, 1999: Grafveld uit de metaaltijden, agrarische ontginning en resten van een holle weg te ‘s Gravenvoeren (Voeren, Lb.), Lunula. Archaeologia protohistorica, VII, 13-14. 52 See the discussion in Roymans/Scheers, this volume, 1 ff. 138 Creemers, G./S. Scheers, 2007: Wichtige Fundstücke als Zeugnisse der Eburonen in Nordostbelgien, in M. Hegewisch (ed.), Krieg und Frieden. Kelten, Romer, Germanen, Begleitbuch zur Ausstellung im Rheinischen Landesmuseum Bonn, 169-174. Creemers, G./L.Van Impe, 2009: Eigenbilzen (Belgique, Limbourg). Une famille aristocratique de l’âge du Fer contrôlant la région, in J. Cession-Louppe, Les Celtes aux Racines de l’Europe. Actes de Colloque tenu au Parlement de la Communautéfrançaise de Belgique et au Musée royal de Mariemont les 20 et 21 octobre 2006, Morlanwelz (Monographies du Musée Mariemont 18), 115-122. De Boe, G., 1971: De Romeinse villa op de Meerberg te Val-Meer (Limburg), Louvain (Acta Archaeologica Lovaniensa, 4). De Boe, G., 1974: Haccourt I. Vestiges d’habitat pré-romain et premières périodes de la villa romaine, Bruxelles (Archaeologia Belgica 168). De Boe, G., 1987: Bewoning rond de villa te Neerharen-Rekem (B.), in P. Stuart /M.E.Th. de Grooth (eds), Langs de weg, exhibition catalogue, Heerlen/Maastricht, 51-54. De Boe, G./L. van Impe, 1979: Nederzetting uit de Ijzertijd en Romeinse villa te Rosmeer, Bruxelles (Archaeo logia Belgica 216). De Grooth, M. E. Th., 1987: Villa rustica. Het boerenbedrijf in het Rijn/Maasgebied, in P. Stuart./ M.E.Th. de Grooth (eds), Langs de weg, exhibition catalogue, Heerlen/Maastricht, 81-86. Delestrée, L.P., 1997: Les monnayages d’or de la Gaule Belgique dérivés du statère à flan large, Revue numismatique 152, Paris, 91-120. Delmaire, R ., 1983: Corpus des trésors monétaires antiques de la France. IL Nord —Pas-de-Calais 62, Paris. Delmaire, R ., 1994: Carte archéologique de la Gaule. 62/1. Le Pas-de-Calais, Paris 161. De Loë, 1937: Musées royaux d’art et d’histoire de Bruxelles. Belgique ancienne. Catalogue descriptif et raisonné. III. La période romaine, Bruxelles Demey, D./J. Roymans, 2 0 0 2 : Beleidsmatige archeologische inventarisatie. Gemeenten Borgloon en Heers, Amsterdam (RAAP-Rapport 812). de Saulcy, E, 1868: Mémoires et dissertations. Lettre à M.A. de Longpérier sur la numismatique gauloise. XXIX. Révision des dix premières lettres, Revue Numismatique, 405-418. Doyen, J.-M., 1987: Les subdivisions ‘au segments de cercles’ du type BN 8030: état de la question, in Mélanges Colbert de Beaulieu, 313-330. Doyen, J.-M., 2005: Monnaies gauloises de sites et trésors monétaires de Gaule Belgique. Trouvailles isolées et dépôts (1995-2005) (première partie), Bulletin du Cercle d’Etudes Numismatiques, 42/3. Duurland, M., 2 0 0 0 : Romeinse vindplaatsen in het landelijke gebied tussen Tongeren en Maastricht. Een inventa risatie en periodisering, Amsterdam (unpublished MA-thesis University of Amsterdam. Ervynck,A./D. Pauwels/A.Vanderhoeven/G.Vynckier, 2000: Romeinse nederzettingen op het Distrigastraject tussen Oupeye en Lanaken (Limburg), Romeinendag 6 , 41-42. Fischer, B./C. Gruel, 2001: Catalogue des monnaies gauloises, in M. Reddé/S. von Schnurbein (eds), Alésia. Fouilles et recherchesfranco-allemandes sur les travaux militaires romains autour du Mont-Auxois (19911997), 2 (Mémoires de l’Académie des Inscriptions et Belles-Lettres 22), 21-39, pl. 1-28. Frebutte, C h./I. Loodts, 1999: Structures de l’âge du fer à Slins et Fexhe-Slins (Juprelle, Lg.), Lunula. Archaeologia protohistorica 1, 55-61. Huybrigts, F., 1 9 1 4 : Compte-rendu des fouilles et acquisitions faites durant l’année 1913 —1914, Bulletin de la Société scientifique et littéraire du Limbourg 32,125-126. Huybrigts, F., 1923: Monnaies germaniques, Annales du XXVe Congrès international de la Fédération archéologique et historique de Belgique, I. Tongres, la ville antique, sa voirie primitive, ses relations commerciales et industrielles, Tongres, 1923, 75-93. In ’tVen, I./W De Clercq (eds), 2005: Een lijn door het landschap. Archeologie en het vTn-project 1991 — 1998, Brussel (Archeologie in Vlaanderen, Monografie 5) . 139 In Wen, I./W.Wouters/T. Debruyne/B. Cooremans, 2 0 0 5 : Heers/Vechmaal -St. Martinusstraat 2 (Oost), in I. In ’tVen/W. Declercq (eds), Een lijn door het landschap. Archeologie en het vTn-project 1997 - 1998, Brussel (Archeologie in Vlaanderen Monografie 5), 1,195-200. Lambot, B./P. Casagrande, 1997: Une monnaie d’Ebusus sur l’oppidum de “Vieux-Laon” à Saint-Tho mas (Aisne). Nouvelles données archéologiques sur Bibrax et le camp de Mauchamp (Cesar B.G. II), Bulletin de la Société archéologique champenoise, 90,15-29. Lambot, B., 2002: L’argent perdu des Remi d’Acy-Romance, in P. Meniel/B. Lambot (eds), Découvertes récentes de l’âge dufer dans le massif des Ardennes et ses marges. Repas des vivants et nourriture pour les morts en Gaule. Actes du 25e colloque international de l’AFEAF, Charleville-Mézières 2001, Reims, 125-142. La Tour, H. De, 1892: Atlas de monnaies gauloises, Paris. Lesenne, M., 1973: Une monnaie gauloise trouvée à Tongres, Archeologie. Kroniek, Nationaal Centrum voor oudheidkundige navorsingen in België, 32. Loscheider, R., 1998: Untersuchungen zum spatlatènezeitlichen Münzwesen des Trevererlandes, Archeologia Mosellana 3, 63-225. Marchai, O., 1997: Oupeye, Haccourt. Une occupation de l’Âge du Fer sous le site de la villa romaine, in M.-H. Corbiau (ed.), Le patrimoine archéologique de Wallonie, Namur, 247-249. Mariën, M.E., 19 8 0 : Belgica Antiqua. De stempel van Rome, Antwerpen. Oost,T./L.Van Impe/J. Bastiaens, 2001: Nederzettingen uit de late bronstijd en de ijzertijd in GingelomMolenveld, in A. Schryvers/L.Van Impe (eds), Op het spoor van het verleden. Archeologie op de hogesnel heidslijn, tentoonstellingscatalogus, 147-170. Opsteyn, L./I. Roovers/M. Lodewijckx/F. Kumps, 1999: Archeologisch onderzoek op de aardgasleiding te Neerlanden (Vl.-Bt.), Lunula. Archaeologia protohistorica 4, 71-73. Paquay,J., 1934: Tongeren Voorheen. Geschiedkundige schets,Tongeren. Pauwels, D./A.Vanderhoeven/G.Vynckier, 2002a: Lafelt (Riemst). Nederzetting uit de Ijzertijd en bijge bouwen van een Romeinse villa, in A.Vanderhoeven/G. Creemers (eds), Archeologische kroniek Limburg 1999, Limburg - Het Oude Land van Loon 81, 297-300. Pauwels, D./A.Vanderhoeven/G.Vynckier, 2002b:Val-Meer (Riemst). Sporen uit de Ijzertijd, Romeinse nederzetting en Laat-Romeinse graven, in A.Vanderhoeven/G. Creemers (eds), Archeologische kro niek Limburg 1999, Limburg —Het Oude Land van Loon 81j 311-314. Pauwels, D./A.Vanderhoeven/G.Vynckier, 2010: Nederzettingen uit de late ijzertijd en de Romeinse tijd te Kesselt, Romeinendag, Louvain-la-Neuve, 29-33. Preud’homme, D./H. Fock/T.D. Bosquet/C. Goffioul, 1999: Fouille d’un site d’habitat de l’Âge du Fer à Orp-Jauche, au heu dit Le Tierceau (Bt w.), Lunula. Archaeologia protohistorica, 1, 62-67. Preud’homme, D./H. Fock/T.D. Bosquet/C. Goffioul, 1999: Fouille d’un site d’habitat de l’Âge du Fer à Orp-Jauche, au lieu dit Le Tierceau (Bt w.), Lunula. Archaeologia protohistorica 7, 62-67. Roosens, H., 1975: Oude versterking te Kanne-Caster, Archaeologia Belgica, 177, 32-36. Roosens, H., 1976: De oude versterking te Caster, Archaeologia Belgica, 186, 54-58. Roosens, H./G.V Lux, 1973: Grafveld met Gallo-Romeinse tumulus te Bedingen, Brussel (Archaeologia Belgica 147). Roymans, N., 2001:The Lower Rhine triquetrum coinages and the ethnogenesis of the Batavi, in Th. Grünewald (ed.), Germania inferior. Besiedlung, Gesellschaft und Wirtschaft an der Grenze der rômischgermanischen Welt, Berlin/New York, 93-145. Roymans, N., 2004: Ethnie identity and impérial power. The Batavians in the Roman empire, Amsterdam (Amsterdam Archaeological Studies 10). Roymans, N./J. Aarts, 2009: Coin use in a dynamic frontier région. Late Iron Age coinages in the Lower Rhine area,fournal ofArchaeology in the Low Countries 1-1, 5-26. Roymans, N/W. Dijkman, 2010: De Keltische goud- en zilverschat van Amby, gemeente Maastricht, Utrecht. 140 Scheers, S., 1969: Les monnaies de la Gaule inspirées de celles de la République romaine, Leuven (KUL,Werken op het gebied van de geschiedenis en de filologie, Reeks 5, 6 ). Scheers, S., 1977: Traité de numismatique gauloise. IL La Gaule belgique, Paris. Scheers, S., 1996a: Frappe et circulation monétaire sur le territoire de la future civitas Tungrorum, Revue belge de numismatique et de sigillographie 142, Bruxelles, 5-51. Scheers, S., 1996b: Les statères bifaces du type Lummen-Niederzier: un monnayage éburon antérieur à la conquête romaine, in M. Lodewijckx (ed.), Archaeological and historical aspects ofWest-European societies, Leuven (Acta Archaeologica Lovaniensia. Monographiae, 8), 87-94. Scheers, S./G. Creemers, 2001: U n trésor de statères gaulois trouvé à Heers (Limbourg, Belgique), Prehistoria 2 0 0 0 ,1,174-179. Scheers, S./G. Creemers, 2002a: U n trésor monétaire des Eburons à Heers (Limbourg, Belgique), Lunula. Archaeologia protohistorica 10, Bruxelles, 80-85. Scheers, S./G. Creemers, 2002b: Een muntschat van de Eburonen te Heers (Limburg, België), Archeologie in Limburg 91, 2-6. Scheers S./J.Van Heesch/R.Van Laere, 1991: Uitmuntend goud. Catalogus gouden munten van het Munt- en Penningkabinet van de Provincie Limburg, Tongeren, Brussel. Smeesters J., 1974: Gouden Gallische munt te Berg, Limburg 53, 4,132. Thirion, M., 1962: Le trésor de Fraire. Monnaies gauloises en potin, Revue belge de numismatique et de sigillographie 108, 67-112. Thirion, M., 1964: Keltische staters te Sint-Joris-Winge, Meer Schoonheid 3, 23-24. (Thirion, M.), 1964: Musées Tongres, Musée gallo-romain, Bulletin du Cercle d’Etudes Numismatiques 1/2, 34-35. Thirion, M., 1967: Monnaies gauloises à la légende ANNAROVECI, Bulletin du Cercle d’Etudes Numis matiques 4/1,11-15. Thirion, M.,,1972: Une monnaie gauloise trouvée àTongres, Bulletin du Cercle d’Etudes Numismatiques 9,47. Tourneur,V, 1907-1908: L’origine desTongrois et une monnaie d’Anarovecos, Gazette Numismatique 12, 33-46. Tourneur,V, 1946: La Belgique pendant l’occupation militaire romaine (49-27 av.J.-C.), Latomus 5,175180. Vanderhoeven,A., 1996:The earliest urbanisation in Northern Gaul. Some implications of recent research in Tongres, in N. Roymans (ed.), From the sword to the plough. Three studies on the earliest romanisation of Northern Gaul, Amsterdam (Amsterdam Archaeological Studies 1), 198-260. Vanderhoeven, A./G.Vynckier/A. Ervynck/B. Cooremans, 1993: Het oudheidkundig bodemonderzoek aan de Kielenstraat te Tongeren (prov. Limburg). Interimverslag 1990-1993. Deel 1. De voor-Flavische bewoning, Archeologie in Vlaanderen 2, 89-146. Van de Weerd, H., 1914: De Civitas Tungrorum, Antwerpen (Verhandelingen van de Algemeene Katholieke Vlaamse Hoogeschooluitbreiding, 176/177). Van Dierendonck, R.M./LJ.F. Swinkels/WJ.H. Willems, 1987: Rijke hereboeren uit Maasbracht, in P. Stuart/ M.E.Th. de Grooth (eds), Langs de weg, tentoonstellingscatalogus, Heerlen/Maastricht, 62-67. Van Heesch, J., 2005: Celtic coins and religious deposits in Belgium, in C. Haselgrove/D. Wigg (eds), Ritual and Iron Age coinage in North-West Europe (Studiën zu Fundmünzen der Antike, 20), BerlinDurham, 247-263. Van ïmpe, L./G. Creemers/S. Scheers/R.Van Laere, 1997a: De Keltische goudschat van Beringen (Lb.), Lunula. Archeologia protohistorica, 5, 21-23. Van Impe, L./G. Creemers/S. Scheers/R.Van Laere/H.Wouters/L. Maes, 1997b: Goden / Macht. Krijgers / Kracht. Keltische gouddepots, Tentoonstellingscatalogus, Tongeren. Van Impe, L./G. Creemers/R. Van Laere/S. Scheers/H. Wouters/B. Ziegaus, 1997/1998: De Keltische goudschat van Beringen (prov. Limburg, B.), Archeologie in Vlaanderen 6, 9-132. Van Impe, L./A. Schryvers/M.Van Strydonck, 2001: Drie ijzertijdkuilen in Landen, in A. Schryvers/L. Van Impe (eds), Op het spoor van het verleden. Archeologie op de hogesnelheidslijn, exhibition catalogue, s i , 143-145. Vannérus, J., 1948: Les deniers de Charles le Chauve (840-877) aux légende DE ou IN FISCO, Revue belge de numismatique et de sigillographie 94, 77-100. Vanvinckenroye,W., 1985: Tongeren Romeinse stad, Tielt. Vanvinckenroye ,W, 1987: Onderzoek van de Gallo-Romeinse tumulus van Gutschoven, Tongeren (Publicaties van het Gallo-Romeins Museum Tongeren 35). Vanvinckenroye, W., 1988: De Romeinse villa op de Sassenbroekberg te Broekom, Tongeren (Publicaties van het Gallo-Romeins Museum Tongeren 38). Vanvinckenroye, W., 1997: De Romeinse villa in het ‘Middelpadveld’ te Vechmaal (Heers), Limburg-Het Oude Land van Loon 76, 179-192. Vanvinckenroye, W., 2001: Über Atuatuca.Casar und Ambiorix, in M. Lodewijckx (ed.): Belgian Archaeology in a European setting, Leuven (Acta Archaeologica Lovaniensia Monographiae 13), 63-68. Verhoeven, M.P.F., 2008: Studieopdracht naar een archeologische evaluatie van het plateau van Caestert (Riemst, Provincie Limburg), Weesp (RAAP-rapport 1769). Wouters, W./I. In ’tVen/T. Debruyne, 2002a: Muizen (Gingelom). Sporen uit de Ijzertijd op het Distrigas-traject, in G. Creemers/A. Vanderhoeven (eds), Archeologische kroniek Limburg 1998, Limburg BE- Het Oude Land van Loon 81,235. W outers,W /I. In ’tVen/T. Debruyne, 2002b:Vechmaal (Heers). Greppels uit de Ijzertijd en vroegRomeinse periode op het Distrigas-traject, in G. Creemers/A. Vanderhoeven (eds), Archeologische kroniek Limburg 1998, Limburg —Het Oude Land van Loon 81, 261-262. Willems, R, 2010: Un statère du type Scheers 25, cl. I trouvé à Lebbeke enfin publié, Bulletin du Cercle d’études numismatiques, 47, 3, 310-311. Willems,WJ.H., 1987: De grote villa vanVoerendaal, in P. Stuart/M.E.Th. de Grooth (eds), Langs de weg, tentoonstellingscatalogus, Heerlen/Maastricht, 46-50. 142 APPENDIX I. DESCRIPTIVE HEERS HOARD. LIST OF THE COINS FROM THE The numbering corresponds to the numbering of the coins in Plates 1-16. O = obverse die; R. = reverse die. The coin composition was identified by means of surface analysis (X-ray fluorescence analysis). See Wouters, this volume, Appendix 4. All coins are in the collection of the Gallo Roman Museum, Tongres. weight (g) die nos. analysis % references Au Ag Cu g o ld staters type Scheers 31, class la, Eburones 1 5.60 0 3 -R 3 2 5.70 0 6 -R 8 3 5.66 0 6 -R 9 4 5.41 0 6 -R 9 33 46 21 5 5.69 0 6- R 9 34 47 19 6 5.75 0 6 -R 9 7 5.64 0 7 -R 9 8 5.76 0 8 - R 10 9 5.50 010 - R 13 10 5.75 010 — R 14 11 5.70 0 11 — R 15 36 38 25 12 5.72 0 11 - R 15 13 5.71 0 12 - R 16 37 40 23 14 5.71 0 1 2 - R 17 15 5.82 012 - R 18 16 5.70 0 13 - R 19 17 5.78 0 13 - R 21 18 5.73 014 - R 18 19 5.80 0 1 4 - R 22 35 44 21 20 5.47 0 14 - R 23 21 5.68 0 1 5 - R 24 22 5.70 0 1 5 - R 25 23 5.75 0 1 6 - R 26 24 5.81 0 1 8 - R 28 41 25 33 25 5.65 0 1 8 - R 30 37 31 32 26 5.61 0 1 8 - R 30 27 5.69 0 1 8 - R 30 28 5.71 0 1 8 - R 30 29 5.62 0 18 - R 30 38 44 18 30 5.74 0 1 8 - R 30 35 40 25 31 5.62 0 1 8 - R 30 32 5.67 0 1 8 - R 30 38 33 29 33 5.72 0 18 - R 30 34 5.63 0 1 8 - R 30 35 5.74 0 1 8 - R 30 36 5.56 0 1 8 - R 30 37 5.61 0 1 8 - R 30 38 5.60 0 1 8 - R 30 37 30 32 143 39 5.74 0 1 8 - R 30 40 5.72 0 18 -R 30 41 5.67 0 18 -R 30 42 5.59 0 18- R 30 43 5.71 0 18- R 30 44 5.64 0 18- R 30 45 5.79 0 1 8 - R 30 46 5.73 47 48 36 25 39 35 35 30 0 18 -R 30 37 38 25 5.71 0 18 -R 30 32 46 22 5.69 0 18- R 30 49 5.61 0 18 -R 30 50 5.46 0 18 -R 30 51 5.67 0 18 -R 30 52 5.63 0 18 -R 30 38 34 28 53 5.68 0 18 - R 30 54 5.88 0 1 8 - R 30 32 34 34 55 5.57 0 1 8 - R 30 56 5.76 0 18 - R 30 57 5.68 0 18 - R 30 58 5.56 0 18 - R 30 59 5.68 0 1 8 - R 30 60 5.69 0 18 - R 30 37 44 19 61 5.77 0 18 - R 30 62 5.72 0 1 8 - R 30 63 5.68 0 1 8 - R 30 64 5.68 0 18 - R 30 65 5.67 0 18 - R 30 66 5.73 0 18 -R 30 67 5.73 0 18 - R 30 68 5.68 0 18 - R 30 69 5.71 0 18 - R 30 70 5.63 0 18 - R 30 71 4.55 0 18 - R 30 72 5.45 0 18 - R 30 73 5.68 0 18 -R 30 74 5.82 0 1 8 - R 30 75 5.62 0 18 - R 30 76 5.66 0 18 -R 30 77 5.39 0 1 8 - R 30 78 5.70 0 18 -R 30 79 5.71 0 1 8 - R 30 80 5.62 0 18 - R 30 81 5.62 0 1 8 -R 30 5.73 5.73 5.70 0 18 - R 30 5.73 5.64 fragmentary 37 38 25 41 25 34 0 1 8 - R 30 41 37 22 0 18 - R 30 38 40 22 0 1 8 -R 30 36 38 26 0 18 -R 30 38 42 20 87 5.65 0 1 8 - R 30 88 5.67 0 1 8 - R 30 89 5.57 0 1 8 - R 30 90 5.74 0 1 8 - R 30 91 5.70 0 1 8 - R 30 92 5.66 0 1 8 - R 30 93 5.66 0 18-R 30 94 5.60 95 5.38 96 5.66 0 1 8 - R 30 97 5.55 0 1 8 - R 30 98 5.69 0 1 8 - R 30 99 5.74 100 5.72 101 5.71 0 1 8 - R 30 102 5.65 0 1 8 - R 30 103 5.67 104 5.72 105 5.79 0 1 8 - R 30 106 5.69 0 1 8 - R 30 42 39 19 0 1 8 - R 30 37 39 24 0 1 8 - R 30 40 38 22 0 1 8 - R 30 37 39 24 0 1 8 - R 30 38 33 29 0 1 8 - R 30 38 36 26 0 1 8 - R 30 37 34 29 51 38 11 53 26 21 107 5.66 0 1 8 - R 30 108 5.72 0 1 8 - R 30 109 5.61 0 1 8 - R 30 110 5.70 0 1 8 - R 30 111 5.66 0 1 8 - R 30 112 5.60 0 1 8 - R 30 113 5.69 0 1 8 - R 30 114 5.62 0 1 8 - R 30 115 5.59 0 1 8 - R 30 116 5.65 0 1 8 - R 30 gold staters type Scheers 29, class la, Nervii 117 5.68 0 3 -R 1 118 5.83 0 10 - R 17 119 5.74 0 23 - R 42 120 5.82 0 27 - R 47 121 5.65 0 42 - R 63 122 5.76 0 43 - R 64 gold staters type Scheers 29, class lb, Nervii 123 5.65 0 45 - R 66 124 5.63 0 45 - R 68 gold stater type Scheers 29, class lia, Nervii 125 5.78 0 57 - R 93 gold staters type Scheers 29, class lib, Nervii 126 5.91 0 71 - R 123 127 5.82 0 7 2 - R 124 128 5.76 0 7 2 - R 125 129 5.86 0 7 3 - R 126 130 5.92 0 7 3 - R 126 145 55 31 14 57 30 13 56 35 S 66 23 11 63 30 7 51 25 24 51 40 9 5.55 41 52 7 5.68 33 42 25 131 5.87 0 73 - R 127 132 5.84 0 7 3 - R 127 133 5.84 0 7 3 - R 127 134 5.91 0 73 - R 128 135 5.90 0 7 3 - R 128 136 5.90 0 73-R128 137 5.84 0 74-R129 138 5.81 0 75 - R 130 139 5.89 0 7 5 - R 131 140 5.89 0 7 5 - R 131 141 5.85 0 75-R131 142 5.90 0 75-R131 143 5.92 0 75 - R 131 144 5.84 0 75 - R 131 145 5.90 0 75 - R 131 146 5.94 0 7 5 - R 131 147 5.88 0 75-R131 148 5.82 0 7 5 - R 132 149 5.92 0 7 5 -R 132 150 5.99 0 75-R132 151 5.92 0 75 - R 133 152 5.92 0 7 5 -R 133 153 5.84 0 7 5 -R 134 154 5.92 0 7 5 -R 134 gold stater 'with star', type Scheers 25, class II, Bellovaci 155 5.82 gold 'eye'stater, type Scheers 30, class 1, Remi 156 6.02 gold stater type Scheers 30, class IV, Treveri 157 gold coin flan 158 146 A P P E N D I X 2. ( U N T I L 3 0 C. D E S C R I P T I V E BC) IN THE LIST OF LATE I R O N AGE T O N G E R E N - M A A ST R I C H T C O IN S A R E A . 53 The numbering corresponds to the numbering in the distribution map in figures 6, 7 and 8. Located finds a. Gold and silver coins 1 Berg (Tongeren),Tomberg. Roman villa? Stater Ambiani uniface, Scheers 24 (gold; 5.19 g). Collec tion Gallo-Romeins Museum PM P-R 6217; (Thirion) 1964,34; Smeesters 1974,132; Scheers 1977, 352, no. 82; Vanvinckenroye unpublished information (archeologische vondstenkartering Limburg); Scheers/Van Heesch/Van Laere 1991, 27, no. 11; Scheers 1996a, 27, note 60; van Heesch 2005, 257. 2 Berg (Tongeren),Tomberg. Roman villa? DURNACOS AUSCRO, LT 1892, pi. XVII 5762 (silver). Scheers 1969, 7, no. 86; Duurland 2000,15-16 and fig.4.2. 3 ‘s Herenelderen (Tongeren), In het RijckerVeld. Context unknown. Quarter stater Remi?, Scheers 533 variant, Doyen class VI (gold, plated). Unpublished find Patrick Schuermans. Scheers 1977, pi. XIX; Doyen 1987. 4 Berg (Tongeren), reservoire,Trappenberg. Settlement and/or cult place? Quinarius Sequani Q.DOCI SAM.F, LT 1892, pi. XVI 5405-5411 (silver; 1.70 g). Find Gerard Messiaen. Scheers 1996a, 28, note 73; van Heesch 2005,257. 5 Berg (Tongeren). Setdement and/or cult place? Quinarius Sequani? Q.DOCI SAM F, LT 1892, pi. XVI 5405-5411 (silver; 1.81 g). Unpublished find Peter Munsters. Information Peter Munsters. 6 Berg (Tongeren). Setdement and/or cult place? Quinarius Ubii, LT 1892, pi. XXXVIII 9396, Scheers 57 class I (silver; 1.81 g). Unpublished find Peter Munsters. Information Peter Munsters. 7 Berg (Tongeren). Settlement and/or cult place? Stater Nervii class la, Scheers 29 (gold). Unpublished find Rob, Groen. Information Nico Roymans. 8 Berg (Tongeren). Setdement and/or cult place? Stater Nervii class la, Scheers 29 (gold; 6.01 g). Col lection Gallo-Romeins Museum PM P-R 6209; Huybrigts 1914, 125-126; Paquay 1934, 68; infor mation Thirion 1960; Bauwens-Lesenne 1968,14; Smeesters 1974,132, Vanvinckenroye unpublished information (archeologische vondstenkartering Limburg); Scheers/van Heesch/Van Laere 1991, 30, no. 14; Scheers 1996a, 27, note 58; van Heesch 2005, 257. 9 Berg (Tongeren). Settlement and/or cult place? Stater Eburones class I, Scheers 31 (gold; 5.67 g). Col lection Gallo-Romeins Museum Tongeren GRM 10908; unpublished find Peter Munsters. 10 Berg (Tongeren). Setdement and/or cult place? Stater Eburones class la, Scheers 31 (gold; 5.81 g). Collection Gallo-Romeins Museum Tongeren GRM 10935; unpublished find Peter Munsters. 11 Berg (Tongeren). Setdement and/or cult place? Stater Eburones class la, Scheers 31 (gold; plated; 2.33 g). Collection Ferdy Willems; unpublished find Peter Munsters. 12 Bolder (Riemst). Context unknown. Stater Eburones class la, Scheers 31 (gold). Unpublished find Benny Emons. Information Tim Vanderbeken, ZOLAD+. 13 Jesseren (Borgloon). Context unknown. Stater Eburones class la, Scheers 31 (gold; 5.40 g). Unpub lished find D. Hermans 2009. Information Simone Scheers. 14 Rijkhoven (Bilzen). Context unknown. Stater Eburones class lb, Scheers 31 (gold; 5.56 g). Unpub lished find Luc Meyers 1998. Information Simone Scheers. 15 Berg (Tongeren), reservoire. Settlement and/or cult place? Stater Eburones class la, Scheers 31 (gold; 5.40 g). Unpublished find Gerard Messiaen 1993; unpublished information W.Vanvinckenroye (arche- 53 The coin types “Scheers”, “LT” and “triquetrum ” refer to 2004. Scheers 1977, La Tour (de) 1892 and Roymans 2001 and 147 ologische vondstenkartering Limburg);Van Heesch 2005, 257. 16 Berg (Tongeren). Setdement and/or cult place? Stater Eburones class la, Scheers 31 (gold; 5.41 g). Unpublished find Gerard Messiaen 1993; unpublished information W.Vanvinckenroye (archeologis che vondstenkartering Limburg);Van Heesch 2005,257. Y] Rienist, Maastrichtersteenweg. Context unknown. Stater Eburones class la, Scheers 31 (gold, 5.40 g). Unpublished find Laurent Ualgasi. Information Tim Vanderbeken, ZOLAD+. 18 Berg (Tongeren). Setdement and/or cult place? Stater Eburones class I, Scheers 31 (gold; 5.32 g). Unpublished find Hans Hansen. Information Simone Scheers. 19 Horpmaal (Heers), Keiberg. Rural setdement? Stater Eburones class II, Scheers 31 (gold; 5.28 g). Unpublished find. Information Luc Meyers and Simone Scheers. 20 Vechmaal (Heers), Zouwveld. Roman villa. Stater Eburones class II, Scheers 31 (gold). Unpublished. Information Hans Hansen en Luc Meyers. 21 Berg (Tongeren). Setdement and/or cult place? Stater Eburones class II, Scheers 31 (gold; 5.12 g). Collection Gallo-Romeins Museum PMPK 93/284. 22 Berg (Tongeren). Setdement and/or cult place? Stater Eburones class II, Scheers 31 (gold; 5.19 g). Collection Gallo-Romeins Museum PMPK 93/285. 23 Berg (Tongeren). Setdement and/or cult place? Stater Eburones class II, Scheers 31 (gold, 5.19g). Collection P. Delvaux. Find Dirk Billiau. Doyen 2005, 177-178, fig. 2. 24 Berg (Tongeren). Setdement and/or cult place ? Stater Eburones class II, Scheers 31 (gold, 5.41g). Find Dirk Billiau. Unpublished find. 25 Vechmaal (Heers), southwest of Zouwveld. Context unknown. Stater Eburones class II, Scheers 31 (gold). Unpublished find Bart Smets. 26 Vechmaal (Heers), Rigoosberg. Roman rural setdement/ villa? Stater Eburones class II, Scheers 31 (gold). Unpublished find Patrick Schuermans. 27 Roclenge-sur-Geer (prov. Liège). Roman villa. Stater Eburones class II, Scheers 31 (gold). Unpub lished find Robert Teuchie. 28 Gors-Opleeuw (Borgloon), Mellenstraat. Stater Eburones class II, Scheers 31 (gold). Unpublished find Jan Helsen. 29 Lauw (Tongeren), Sleiberg. Roman villa. Biface stater Eburones type Lummen-Niederzier (gold; 6.39 g). Unpublished find Geert Daniëls. Scheers 1996b. 30 Vechmaal (Heers), Zouwveld. Roman villa. Blank coin (gold; 6.00 g). Collection Gallo-Romeins Muse um Tongeren GRM 11495. Unpublished find. Information D. Hermans. Information Simone Scheers. 31 Vechmaal (Heers), Zouwveld. Roman villa. Blank coin (gold; 5.47 g). Collection Gallo-Romeins Museum Tongeren GRM 9262. Unpublished find R af Janssen. 32 Berg (Tongeren). Setdement and/or cult place? Blank coin (gold; 5.54 g). Find Dirk Billiau. Informa tion Simone Scheers. van Heesch 2005, 257. 33 Berg (Tongeren), reservoire. Settlement and/or cult place? Blank coin (gold; 5.39 g). Unpublished find Gerard Messiaen, van Heesch 2005, 257. 34 Berg (Tongeren). Settlement and/or cult place? Triquetrum (silver, 5.46 g). Unpublished find Peter Munsters. 35 Berg (Tongeren). Setdement and/or cult place?Triquetrum variant a (silver). Unpublished find Peter Munsters. 36 Berg (Tongeren). Setdement and/or cult place?Triquetrum variant a (silver). Unpublished find Peter Munsters. 37 Berg (Tongeren), Op de Heurkensberg. Setdement and/or cult place? Triquetrum variant a (silver). Unpublished find Luc Meyers. 38 Maastricht, Dousberg. Context unknown. South German rainbow cup, LT 9423 (gold, 7.37 g). Find H. Hansen. Information Simone Scheers. Roymans/Dijkman 2010, bijlage 1, no. 24. 148 39 Maastricht, Randwyck. Rural settlement. Stater Eburones class II, Scheers 31 (gold). Find C. Peusens andT. Poels. Roymans 2004, Appendix 4.2., Roymans/Dijkman 2010, bijlage 1, no. 12. 40 Maastricht, Amby, Heihof. Stater Eburones class II, Scheers 31 (gold; 4.96 g). Find P. Derks. Roymans/ Dijkman 2010, bijlage 1, no. 19. 41 Maastricht, Amby, Heihof. Triquetrum variant c (silver; 5.15 g). Find P. Derks. Roymans/Dijkman 2010, bijlage 1, no. 19. 42 Maastricht, Amby, Bodemsweg. Triquetrum variant c (silver). Find F. Schröder. Roymans/Dijkman 2010, bijlage 1, no. 20. 43 Vechmaal, Bornebeemden (Heers). Coin hoard: 116 staters class la Eburones, 8 staters class I Nervii, 30 staters class II Nervii, 1 gold stater Treveri, 1 gold stater Remi, 1 gold stater VeHocasses, 1 gold coin blank. Find R af Janssen. This publication. 44 Maastricht, Amby, Hagenstraat. Coin hoard: 78 silver rainbow staters, 21 staters class la Eburones, 19 staters lb Eburones. Roymans/Dijkman 2010, this publication. 45 Berg (Tongeren). Settlement and/or cult place? Quinarius Annaroveci, Scheers 58 (silver; 1.64 g). Unpublished find Peter Munsters. 46 Berg (Tongeren). Settlement and/or cult place? Quinarius, Scheers 58 (silver). Collection GalloRomeins Museum GRM 10906. Unpublished find Peter Munsters. 47 Eben-Emael (Bassenge - prov. Liège). Context unknown. Quinarius ANNAROVECI, Scheers 58 (silver). Scheers 1977. 48 Rutten (Tongeren), Strijdmakkersstraat. Context unknown. Quinarius ANNAROVECI, Scheers 58 (silver; 1.40 g). Unpublished find Alain Schouterden. Information Johan van Heesch (collection Ferdy Willems). 49 Berg (Tongeren). Settlement and/or cult place? Quinarius ANNAROVECI, Scheers 58 (silver). Unpublished find Luc Meyers. 50 Berg (Tongeren), Trappersberg. Settlement and/or cult place? Quinarius ANNAROVECI, Scheers 58 (silver). Unpublished find Peuskens. Unpublished information WVanvinckenroye (archeologische vondstenkartering Limburg). 51 Between Koninksem and Rutten (Tongeren). Context unknown. Quinarius ANNAROVECI, Scheers 58 (silver), van de Weerd 1914, 8-9. 52 Between Koninksem and Rutten (Tongeren). Context unknown. Quinarius ANNAROVECI, Scheers 58 (silver), van de Weerd 1914, 8-9. 53 Between Koninksem and Rutten (Tongeren). Context unknown. Quinarius ANNAROVECI, Scheers 58 (silver), van de Weerd 1914, 8-9. 54 Between Koninksem and Rutten (Tongeren). Context unknown. Quinarius ANNAROVECI, Scheers 58 (silver), van de Weerd 1914, 8-9. 55 Between Koninksem and Rutten (Tongeren). Context unknown. Quinarius ANNAROVECI, Scheers 58 (silver), van de Weerd 1914, 8-9. 56 Tongeren. Context unknown. Quinarius ANNAROVECI, Scheers 58 (silver; 1.40 g). Unpublished find. Information HugoVanhoudt; information S. Scheers. 57 Tongeren. Context unknown. Quinarius ANNAROVECI, Scheers 58 (silver; 1.58 g). Old collection Constant Bamps, currently KMPK Brussels. Bamps, 1889, 249 ; Tourneur, 1907-1908, 39-41; Tour neur, 1946,179, III; Scheers 1977, 509, no. 2. 58 Tongeren, excavations by de Renesse-Breidbach. Context unknown. Quinarius ANNAROVECI, Scheers 58 (silver; 1.16 g). Paris, Bibliothèque Nationale, no. 8893 (collection F. de Saulcy). de Saulcy 1868, 408-409; Bamps, 1889, 250-251; Tourneur, 1907-1908, 39, 2;Thirion 1967,14, no. 2; Scheers 1977, 509, no. 2. 59 Tongeren, excavations by de Renesse-Breidbach. Context unknown. Quinarius ANNAROVECI, Scheers 58 (silver; 1.12 g). Paris, Bibliothèque Nationale, no. 8894 (collection F. de Saulcy). Chalon 149 1862,106-107; de Saulcy 1868, 408;Thirion 1967,14, no. 3; Scheers 1977, 509, no. 2. 60 Maastricht, Amby, Zonhof. Context unknown. Quinarius ANNAROVECI, Scheers 58 (silver). Roymans/Dijkman 2010, 10, site 21, fig. 3, d. b. Bronze and cast bronze coins 61 Tongeren, near Bilzersteenweg, within Roman walls. Context unknown. Remi R EM O /REM O , LT 1892, pi. XXXII 8040, Scheers 146 (bronze; 2.66 g). Collection M.P. Jadoulle Tongeren; informa tion Thirion 1960 (KMPK Brussels);Thirion 1972, 47; Lesenne 1973, 32; Scheers 1977, 632, no. 56; Scheers 1996a, 28, note 68. 62 Tongeren, Kielenstraat. Excavation Roman city centre, supervised by Alain Vanderhoeven. Rem i with running man, LT 1892, pi. XXII 8124, Scheers 191 (cast bronze; 2.39 g). TO 90-4/12/90-66-16-574; information Johan van Heesch; Scheers 1996a, 28, note 68. 63 Tongeren, Kielenstraat. Excavation Roman city centre, supervised by Alain Vanderhoeven. Nervii rameau, Scheers 190-1 /II (bronze; 2.66 g).TO 88-13/l/88-192-3-12-185;Vanderhoeven 1993,92. 64 Berg (Tongeren). Settlement and/or cult place? Nervii rameau-A, Scheers 190-IV (cast bronze; 4.74 g). Unpublished find Peter Munsters, no. 29. 65 Koninksem (Tongeren). Context unknown. Nervii rameau-A, Scheers 190-IV (cast bronze; 2.54 g). Unpublished find Gerard Messiaen. Scheers 1996a, 28, note 69. 66 Tongeren, Bilzersteenweg. Context unknown. Nervii rameau-A, Scheers 190-IV (cast bronze; 5.13 g). Collection Gallo-Romeins Museum de Schaetzen no. 4509; information M. Thirion 1960 (KMPK Brussels). 67 Berg (Tongeren). Setdement and/or cult place? Nervii quarter-stater with segments, Scheers 152-V (bronze; 1.23 g). Collection Gallo-Romeins Museum PMP 90/43; Scheers/van Heesch/Van Laere 1991, 32, no. 16; Scheers 1996a, 28, note 61. 68 Berg (Tongeren). Setdement and/or cult place?Triquetrum type Bochum (bronze). Unpublished find Peter Munsters. 69 Berg (Tongeren), Op de Heurkensberg. Settlement and/or cult place? Triquetrum type Bochum (bronze; 3.02 g). Unpublished find Luc Meyers. Information S. Scheers. 70 Rukkelingen-Loon (Heers), Sterstraat. Context unknown. Triquetrum type Bochum (bronze). Unpublished find Winand Bruninx. 71 Helshoven (Heers). Context unknown.Triquetrum type Bochum (bronze; 4.40 g). Unpublished find Patrick Schuermans. 72 Tongeren, Caesarlaan. Context unknown. Triquetrum type Bochum (bronze; 3.22 g). Unpublished find R.Van Caelenberghe. Information Simone Scheers. 73 Tongeren, Hondstraat. Excavation Roman town house 1989. Triquetrum type Bochum (bronze; 5.37g). Scheers/van Heesch/Van Laere 1991, 38, no. 22. 74 Riksingen (Tongeren), Nieuwe Steenweg. Context unknown.Triquetrum type Bochum (bronze; 4.50 g ). Collection Gallo-Romeins Museum PMPK 97/54. Unpublished find Robert Teuchie. Not-located finds a. Gold and silver coins 75 Tongeren. Context unknown. Stater Ambiani uniface, Scheers 24 (gold). Huybrigts 1923,74 ;Thirion 1964, 23. 76 Tongeren. Context unknown. Quinarius Sequani Q.DOCI SAM F, LT 1892, pi. XVI 5405-5411 (silver; 1.53 g). Gallo-Romeins Museum, collection de Schaetzen no. 2746. 77 Koninksem (Tongeren), in the vicinity. Context unknown. Stater Treveri class IV, Scheers 30-IV (gold; 5.86 g). KMPK Brussels collection C. Bamps; Bamps 1889, 25, 247 and 256, pi. Ill, 1; Scheers 1977, 424, no. 6; Scheers 1996a, 27, note 59. 150 78 Tongeren. Context unknown. Quinarius Ubii LT 1892, pi. XXXVIII 9396, Scheers 57-1 (silver; 1.60 g). Collection Léon Naveau de Bommershoven; Bamps 1889, 25, 256, pi. Ill, 3. 79 Tongeren. Context unknown. Stater Eburones class la, Scheers 31 (gold; plated; 3.55 g). Collection Gallo-Romeins Museum PMP 91/01; Scheers 1977, 439-440; Scheers/van Heesch/Van Laere 1991, 35, no. 19; Roymans 2004, appendix 4.2. 80 Tongeren or immediate vicinity. Context unknown. Stater Eburones class la, Scheers 31 (gold; 5.49 g). Collection Gallo-Romeins Museum PMP 91/205; Scheers 1977, p. 439-440; Scheers/van Heesch/ Van Laere 1991, 36, no, 20; Roymans 2004, Appendix 4.2. 81 Heers, direction of Oreye. Context unknown. Stater Eburones class la, Scheers 31 (gold; 5.61 g). Information P. Delvaux Brussels; Doyen 2005,183. 82 Neerharen (Lanaken). Context unknown. Stater Eburones class II, Scheers 31 (gold; 5.30 g). Unpub lished find. Information Luc Meyers. 83 Vechmaal (Heers). Context unknown. Quinarius ANNAROVECI, Scheers 58 (silver; 1.6 g). Unpub lished find. Information Marc Billiau and Hans Hansen. b. Bronze and cast bronze coins 84 Tongeren? Context unknown. Ambiani with bull and horseman (bronze; 2.02 g). Collection GalloRomeins Museum (collection de Schaetzen); findspot Tongeren according to information Thirion 1960 (KMPK Brussels). 85 Tongeren. Context unknown. R end stater with eye, Scheers 30/3 (bronze; 4.49 g). KMPK Brussels old collection C. Bamps; Bamps 1889, 25, 247; Scheers 1977, 431, no. 30/3, ex. 2. 86 Berg? (Tongeren). Settlement and/or cult place? Nervii quarter-stater with segments, Scheers 152-V (bronze; 1.23 g). Collection Gallo-Romeins Museum PMP 91/207; Scheers/van Heesch/Van Laere 1991, 33, no. 17; Scheers 1996a, 28, note 61. 87 Tongeren. Context Unknown. Nervii VERCIO, Scheers 145 (bronze; 5.57 g). Collection GalloRomeins, Museum old fund; information Thirion 1960 (KMPK Brussels). 88 Tongeren. Context unknown. Nervii VIROS, LT 8772, Scheers 29a (bronze; 5.06 g). Collection Gallo-Romeins Museum, de Schaetzen no. 2748; probably this coin is mentioned in (Thirion) 1964, 34-35; Scheers 1977, 407, no. 14; Scheers 1996a, 28, note 62. 89 Tongeren. Context unknown. Nervii VARTICE, Scheers 190-11 (bronze; 2.16 g). Collection de Schaetzen no. 4883. Unpublished find baron de Schaetzen 1938; information M. Thirion (KMPK Brussels). 90 Tongeren. Context unknown. Nervii probably VARTICE, Scheers 190-II (bronze; 3.00 g). Collec tion de Schaetzen (from collection Christaens-Peeters); probably coin mentioned in (Thirion) 1964, 35; Scheers 1996a, 28, note 64. 91 Tongeren. Context unknown. Nervii rameau-A, Scheers 190-IV (cast bronze). Find Jan Spits. (Thiri on) 1964, 35; Thirion 1962, 95 and ill. 75, fig. 2; Scheers 1977, 746, no. 6 6 ; Scheers 1996a, 28, note 69; information M .Thirion 1960 (KMPK Brussels). 92 Tongeren. Context unknown. Senones, LT 7417 (cast bronze; 4.19 g). Doyen 2005,187, fig. 2. 93 Tongeren? Context unknown. Treveri ARDA, Scheers 30a-I, LT 8839 (bronze). Collection GalloRomeins Museum, de Schaetzen no. 2754. (Thirion) 1964, 35. 94 Tongeren? Context unknown. Treveri ARDA, Scheers 30a- III(?), LT 8839 (bronze). Collection Gallo-Romeins Museum, de Schaetzen no. 2747. (Thirion) 1964, 35. 95 Tongeren? Context unknown.Treveri A. HIRCTVS, Scheers 162, LT 9235 (bronze). Collection GalloRomeins Museum. (Thirion) 1964, 35. PLATES AU coins are depicted on the scale 2:1. The numbering of the coins corresponds to the numbering in Appendix 1. 153 HEERS, PLATE I 1 / f i j f s B f \ I m l 1 yl i 10 154 HEERS PLATE 2 HEERS, PLATE 3 HEERS PLATE 4 157 HEERS, 158 PLATE 5 HEERS, PLATE 6 159 H E E RS, PLATE 7 HEERS, PLATE 8 l6 l HEERS, IÓ2 PLATE 9 HEERS, PLATE 10 163 HEERS, 164 PLATE II HEERS, PLATE I 2 165 HEERS, PLATE I 3 HEERS, PLATE I 4 167 HEERS, PLATE I 5 149 i68 150 HEERS, PLATE I 6 169 T he gold and silver hoard o f Maastricht-Amby Nico Roy mans/Wim Dijkman1 1 2 3 4 5 6 Discovery and excavation o f the hoard Composition of the hoard and origin of the coins 2.1 The gold staters o f the Eburones 2.2 The silver rainbow staters: an Eburonean emission? The dating of the hoard Historical context: the revolt o f the Eburones and their contacts with transrhenish Germani Late Iron Age habitation in the Maastricht area Conclusions References Appendix 1. List of sites with Late Iron Age material in the municipality of Maastricht Appendix 2. Descriptive list of the coins from the Amby hoard Plates I D I S C O V E R Y A N D E X C A V A T I O N OF T H E H O A R D In the spring of 2008, in a field in the village of Amby northeast of Maastricht, amateur metal detectorist Paul Curfs made the discovery of a lifetime: a dozen gold and silver Celtic coins from the 1st century BC. A short time later, similar coins were found at the same site by other detectorists. These were reported in exemplary fashion to the municipal archaeological service of Maastricht, which contacted the first author. Given that the coins were discovered just a few dozen metres apart, we suspected at once that this was a hoard that had been disturbed by ploughing. Because it was possible that part of the hoard still lay buried, and to prevent gold diggers from plundering the coins, we decided to press ahead immediately with a small-scale excavation. There were two key questions we needed to answer: were the coins really part of a hoard and what could we learn about the find’s archaeological context? The investigation took place from 13 to 15 October 2008 and was conducted jointly by the archaeological service of the city o f Maastricht and the Archaeological Centre at the Vrije Universiteit in Amsterdam .12 The field where the coins were found lay between Hagenstraat and Bodemsweg on the eastern side of the village of Amby, at an elevation of about 55 m +NAP on a terrace of the River Meuse (fig. 2). The field, sloping upward to the east, consists of fertile loess soil. The excavation, which is discussed below, unearthed several sherds of late medieval pottery from the topsoil, an indication that this site has undoubtedly been under cultivation since the Late Middle Ages. 1 2 The authors would like to thank Bert Brouwenstijn and Soeters, Anne Brakman and W im Dijkman o f the muni Jaap Fokkema.VU University Amsterdam, who took care cipality o f Maastricht. We would like to thank Paul Curls, o f the illustrations. Richard Speets and Ed Heitzer, who were responsible for The field team was led by H enk Hiddink and Karel-Jan metal detection. Kerckhaert (both from VU-HBS) supported by Gilbert Fig. 1. Overview o f the Aniby hoard. Foto Restaura. The investigation explored a section of the field measuring about 12 x 31 m at the location where the stray coins were found. First of all, a crane was used to remove the approx. 40 cm layer of topsoil layer by layer. The surface uncovered in this way was continually searched with a metal detector. This process unearthed a number of stray dispersed coins at a depth of about 30 cm beneath ground level (fig. 3 ), which is probably the maximum reach of most metal detectors. Next, immediately beneath the topsoil, an area was levelled at a depth of about 50 cm below ground (level). Here we came across a heavy concentration of coins (fig. 2) and so decided to excavate an area of about 1.5 x 1.5 m by hand with a trowel. This involved laying a profile line across the centre of the concentration and excavating one half. The profile revealed that the soil at this level had recently been disturbed by a field cultivator, which had broken up the structure of the soil. The coins lay m two sub concentrations within a one-metre radius, but it was clearly evident from the cut that this was because the field cultivator had struck the concentration and pulled it apart. This is probably what pushed some of the coins up into the present-day topsoil. An area just below the disturbed area was then levelled at approx. 70 cm beneath the surface. Here too, no traces were found of the pit in which the hoard was originally buried. Presumably, this feature had been completely disrupted by the field cultivator in recent times. The X andY coordinates were established for all excavated coins, together with a height measure ment. This information was then used to create a distribution map for the coins in both the horizontal and vertical planes (fig. 3 ).The map showed that almost the entire coin concentration has been excavated, with the outer boundary being reached on all sides. We may conclude from this data that virtually the entire hoard has been recovered and that its original location has also been found. We also established, however, that strictly speaking the coins were no longer in situ. The hoard must o rig in ally have been buried in a pit that was approx. 70 cm deep. It would have been buried in either an earthenware pot or a pouch made of organic material. Given that no pottery sherds 172 Fig. 2. The find site o f the coin hoard at Amby-Hagenstraat. were found during the investigation, we may deduce that the coins were wrapped in a leather or cloth pouch. Amateur archaeologists conducted an intensive search of the field in the immediate vicinity of the hoard but failed to unearth other finds pointing to habitation in the Late Iron Age. It seems that the hoard was not buried within a settlement or cemetery; if that were the case, we would certainly have come across large numbers o f pottery sherds and fragments of glass LaTene bracelets.3 Nor is burial at a cult place a possibility, as we would then expect a larger variety of metal objects in particular. This is because a cult place is by definition a site where ritual activities with a ritual focus are repeatedly carried out .4 There is no evidence at all for this at Amby; the location appears to have been used only once as a place of deposition. Unlike the hoards of Echt and Niederzier for instance, all evidence points to the coins having been buried at a site that was uninhabited at the time.This does not mean that the hoard was buried at random, however. The location may have been significant, perhaps chosen because it featured 3 Cf. the study o f the Echt coin hoard: Roym ans/H iddink 4 Derks 1998,133. 2006. See also Roymans/Verniers 2009. 173 MAASTRICHT-AMBY 2008 319.579 Fig. 3. Map o f the excavation at Amby-Hagenstraat showing distribution o f the coins found in the horizontal plane (below) and vertical plane (above). Drawn by K.-J. Kerckhaert (VU-HBS). A find spot o f coin; B height in metres +NAP a tree or was located in a sacred wood. We cannot say why the coins were buried there. Two main inter pretations present themselves: the coins were buried to keep them safe (Versieckdepot in German) until they could be dug up again later, or they were a votive deposit intended for a supernatural power. In this discussion of whether the deposition was ritual in nature, we must guard against a simplistic distinction between profane and religious. For example, a hoard may be placed temporarily under the protection of a supernatural power to keep it safe.5 5 Nick 2005,145. 174 Fig. 4. Gold Eburonean staters from the Amby hoard. Foto Restaura. 2 C O M P O S I T I O N OF T H E H O A R D A N D O R I G I N OF THE C O I N S Eighty-one coins were collected during the excavation. If we add to this the 38 coins found by private individuals before and after the excavation, we arrive at a total of 119 coins. We are able to identify two distinct groups: 41 gold coins of the Scheers 31 type attributed to the Eburones and 78 silver ‘rainbow cups’ of the triquetrum type originating in the Rhine region. 2.1 THE GOLD STATERS OF THE EBURONES One-third of the coin hoard comprises staters of the Scheers 31 type (fig. 4). The central motif on the obverse is a triskeles surrounded by a zigzag line, a cross and a series of dots (fig. 5). The reverse shows a stylised horse facing left within a configuration of dots and circles. Scheers identifies two subtypes — classes I and II - within this coin group.6 The coins in class II are characterised by a cruder engraving of both the horse and the triskeles and by the presence of a cross above the horse’s back. Class II is clearly the later group because of its slightly lighter weight and higher copper content.7 All coins from the Amby hoard belong to class I. However, there are two variants within this class, which can be distinguished on the basis of the stylisation of the image on the reverse. We refer to them 6 Scheers 1977, 439 ff. 7 Roymans 2004,23 and Table 4.1. 175 Fig. 5. Overview o f the typochronology o f the FORRER 399/400 (HESSEN) SCHEERS 30-4 (TREVERI) A Eburonean staters and their prototypes. here as variants la and lb (fig. 5). Class la features three circles above the horse, SCHEERS31 (EBURONES) whereas class lb has a row o f six circles, placing it in an intermediate position stylistically between classes la and II. This sequence is confirmed by the slightly lower average weight of the class lb coins compared with the class la coins in the Amby hoard. Both variants are equally represented at Amby, with 21 and 20 specimens respectively. The triskeles staters are closely related typologically to some gold coinages from the Middle Rhine region (fig. 5). The triskeles m otif on the obverse was prob ably inspired by the gold ‘rainbow cups’ of the triquetrum type, which originated from the area east of Bonn. The reverse is almost an exact copy o f the Scheers 30-IV gold stater, attrib uted to the Treveri.*8 Table 1 gives us some idea of the weight of the Amby coins. The type la coins have an average weight of 5.70 g and the type lb coins 5.65 g.This places them within the values of the currendy known class I coins, which range from 5.85 to 5.50 g. class number average weight la 21 5.70 g lb 20 5.65 g total 41 Table 1. Typology and average weight o f the Scheers 31 type staters from the Amby hoard. The reddish colour alone of the coins tells us that they are made of poor-quality gold, to which a con siderable quantity of silver and especially copper has been added. The composition of some class I coins found elsewhere has recently been analysed. They were shown to comprise about 50% copper, with both the gold and silver content fluctuating at around 25%.9 The composition of almost all the Eburonean staters from Amby has been tested using X-ray fluorescence analysis. This has produced higher values for gold and especially silver (see Appendix 2).We should point out, however, that this was a surface analysis and that the proportion of copper is significandy underrepresented as a result of corrosion. We should also bear in mind considerable variation in coin composition, sometimes even within a group of die-match coins, as has been established in the Heers hoard.10 8 For Rhineland rainbow cups o f the triquetrum type, see Roymans 2004,67-101. For a recent distribution map of Scheers 30-IV staters, see Loscheider 1998,134-135. Fig. 6. Distribution o f gold staters o f the Scheers 31 type, attributed to the Eburones.The numbering refers to Roymans/Scheers, this volume, 40 ff., Appendix 3. a 1-5 specimens; b >20 specimens; c hoard find A closer analysis of the coin images gives us an idea of the number of dies used for the 40 gold staters from Amby (p. 35, Appendix l ) .11 The coins in class la are highly heterogeneous in composition, with a total o f 11 obverse and 17 reverse dies used for 21 coins. For the 20 coins in class lb, on the other hand, only three obverse and three reverse dies were used. This second group of coins is therefore more homogeneous in composition and closer to the source (i.e. the mint), whereas the first group circulated for somewhat longer and became more mixed as a result.12 Thus the gold staters consist of two different sets of coins: a number of stray class la coins that were scraped together and a fairly homogeneous set of class lb coins. Once again, this indicates that the second group is the later variant. In an absolute sense, however, the differences in time of minting will have been minor, given the negligible differences in weight and in traces of wear between coins from the two groups. The distribution area for the coins spans Central Belgium, the southeastern Netherlands as far as the Rhine and the neighbouring part of the German Lower Rhine region (fig. 6 ). We can discern various 9 Roymans 2004, 38, Table 4.2. The results were obtained by means o f neutron activation analysis, which also cov ers the core o f the coin. 10 Wouters, this volume. 11 We wish to thank Simone Scheers for her assistance with the die analysis o f the Amby coins. 12 We observe a similar phenomenon with many other gold hoards, including the Heers hoard. 177 Fig. 7. Overview o f die pairings o f Scheers 31 staters from the Amby hoard with coins from other find sites. concentrations within this area, which could suggest that there were different regional emissions. The distribution of die-linked coins, however, shows that we are dealing with a single circulation area (fig. 7). For the class la coins from Amby, die links are known with coins from the southeastern Nether lands (Empel, Houten, Randwijk, Oirschot, Weert), Belgium (Asse, Berg, Molenbeek-Wersbeek, Heers, Braives, Waremme, Jauche), France (Boviolles) and Germany (Inden). For coins in class lb, die links are known from Empel, Randwijk and Waremme. There is no certainty about where these coins were struck but the present distribution pattern points to the Tongres/Maastricht region, which would make it a local coin type. Based on its low weight and gold content, and its association with other late coins in the Heers and Fraire hoards (see below), this coin type is dated to the m id-lst century BC. Because of this dating and the distribution of these coins across the southeastern Netherlands and Central Belgium, they are fairly generally ascribed to the Eburones reported on by Caesar.13 As we will demonstrate below, there are really no arguments that would support an alternative attribution. 13 Cf. Scheers 1977, 441; idem 1996, 32; Heinrichs 1999; Roymans 2004,43-44; Creemers/Scheers 2007. i 78 Fig. 8. Silver rainbow cups o f the triquetrum type from the Amby hoard. Foto Restaura. 2.2 SILVER RAINBOW STATERS: AN EBURONEAN EMISSION? Two thirds of the Amby hoard are Rhineland ‘rainbow cups’ (fig. 8). Like the Eburonean staters, these feature a triskeles or triquetrum inside a laurel wreath on the convex obverse, and a combination of point circles and double circles arranged in pyramid form and surrounded by a zigzag fine. This coin type has its origins in the German Middle Rhine region east of Bonn, where they were struck as gold staters, presumably in the oppidum on the Diinsberg.14They are known as gold staters of the Mardorf group, named after the hoard find of the same name. The triquetrum coins discussed here are the silver successors to that gold series. The silver rainbow cups from Amby can be broken down into two groups in terms of the image on the concave reverse of the coins. Firstly, we have a group of 24 coins with a standard image, labelled variant a. Secondly, there is a group of 50 coins bearing an additional mark on the reverse in the form of a ring, a torque or a human head (?). These additional marks are indicated by a letter code.15A total of six variants are known from Amby: b, c, i, k, t and u (fig. 9), one of which (variant u) had never been seen before. Earlier research has shown that we can distinguish between predominantly silver coins bear ing early additional marks that rarely if ever persisted in mid-Augustan contexts in the Rhineland and 14 Cf. Roymans 2004; Schulze-Forster 2002,113. Roym ans/H iddink 2006. O n the possible meaning of 15 Roymans 2004, 72-73 (fig. 6.3), with supplements in the additional marks, see Roymans 2004, 90-91. 179 Fig. 9. The additional mark variants o f the silver triquetrum Lith-group coins present in the Amby hoard. Classification after Roymans 2004, fig. 6.3, with supplements. predominantly copper coins with late additional marks that occur frequently in Augustan contexts.16We can assign all coins with additional marks from Amby to this first group, which we call the ‘Lith group’.17 The coins are of a silver-copper alloy. The traces of copper oxide on uncleaned coins point to the admixture of a substantial amount of copper, as does the reddish lustre on some specimens.18 Likewise, the yellowish lustre on parts of some coins indicates the admixture of gold. This picture is confirmed by the metal analysis of Lith-group coins found elsewhere. These have a silver/gold content of between 40 and 60%, whereas the remaining part (about half) consists of copper.19The composition of twelve coins from Amby was tested using the X-ray fluorescence method (see Appendix 2). This revealed a copper content fairly constantly below 10% and high percentages of silver in particular, but we must remember that this was a surface analysis only, which can give a highly distorted picture (see above). The coins range in weight from 5 to 6.4 g.20Table 2 shows the average weight for each variant bear ing additional marks. We see that Lith-group coins with additional marks are slightly heavier than those without such marks (variant a). It is not clear whether this tells us anything about the chronology or ori gin of these coins. In general, coins in the Lith group vary considerably in weight. A number of heavier specimens are known from the Dutch river region, while several conspicuously light coins made mainly of bronze have been reported from the Diinsberg.21 16 Roymans 2004,70; Schulze-Forster 2002,122-128. 17 Only variant k has known specimens from later Augustan 19 20 contexts, such as the Bochum hoard. Cf. Roymans 2004, fig. 6.16. However, these are bronze and appear to be o f later date. 18 See for example nos. 106 and 107, both variant i specimens. 180 Roymans 2004, 73 and table 6.2. Incomplete coins or coins with parts o f their surface mis sing have not been included. 21 Roymans/Van der Sanden 1980, 205 ff. and fig. 10; Schulze-Forster 2002, 122. variant number average weight u 3 6.00 g c 9 5.92 g k 2 5.83 g b 24 5.77 g t 8 5.75 g i 4 5.66 g 5.59 g a 24 indet. 4 total 78 Table 2. Typology (based on additional mark variant) and average weight o f the silver triquetrum coins from the Amby hoard. The distribution patterns for silver triquetrum coins (figs 10 to 13) point to close relationships between tribal groups on either side of the Rhine. But how exactly we should interpret these relationships is open to debate. A good place to start is the general distribution map of Lith-group coins bearing the earliest additional marks (p. 16, fig. 11). We can distinguish three clusters: the oppidum on the Diinsberg in Hessen in Germany and the regions around Kessel/Lith and Maastricht in the southern Netherlands. Assuming that the prototypes of the silver triquetrum coins (the gold triquetrum staters of the Mardorf group) were probably struck on the Diinsberg, this would seem to suggest a westward flow of coins from the Diinsberg region to the Rhine/M euse region. This does not mean, however, that all the silver coins in the Lith group were also struck on the Diinsberg. An alternative interpretation is that coin production itself shifted westwards. This is supported by the substantial number of coins with early additional marks from the Kessel/Lith and Maastricht regions, as well as by the presence there of several additional mark variants that are unknown from the Diinsberg.22We also note that the coins of variants b, c and i from the Diinsberg differ by virtue of their extremely low weight and high copper content.23 In the future, a systematic study o f the coin dies used for each additional mark variant will yield more information about the nature and intensity of the relationships between the regions east and west of the Rhine.24 For the present, we feel that the most plausible scenario involves production of Lith-group silver coins in the western Rhine region, and the Dutch river region in particular. This would mean that here too we are dealing with a ‘local’ coin type, struck by a subgroup of the Eburones. Especially the u and h variants. For the latter, see R oy- the Fraire hoard (Scheers/Creemers, this volume, 47 ff. mans 2004, fig. 6.13. and 60, no. 8) and one from Lith (Roymans 2004, 191, Schulze-Forster 2002,122-123. no. 147), whereas no examples are yet known from the For example, the reverse on coins 82-85 o f variant b eastern R hine region. from Amby (Plate 9) has a die link with a coin from l8l Fig. 10. Distribution o f silver triquetrum coins o f variant b. After Roymans 2004, fig. 6.7, with supplements. 1 Kessel (3 specimens); 2 Lith, the Bergen; 3 Maren (2 specimens); 4 Alem; 5 Empel, De W erf (2 specimens); 6 Deil (2 speci mens), Hooiveld; 7 Est, Rijs and Ooyen (3 specimens); 8 Randwijk, Hokkerden; 9 Rum pt; 10 Ittervoort; 11 Werl; 12 Zimmern; 13 Diinsberg (copper); 14 Heuchelheim (2 specimens); 15 Fraire; Amby-Hagenstraat (23 specimens) Fig. 11. Distribution o f silver triquetrum coins o f variant c. After Roymans 2004, fig. 6.8, with supplements. 1 Kessel (3 specimens); 2 Echteld, Kanaal; 3 Echteld, vicinity; 4 Empel, D e Werf; 5 Rum pt; 6 Gellicum; 7 Est, Tieflaar; 8 Deil, Hooiveld; 9 Amby-Heihof; 10Tongeren-Berg; 11 Werl; 12 Dünsberg (3 specimens, copper); 13 Lausanne; 14 Amby-Bodemsweg; 15 Echt; 16 Riksingen; Amby-Hagenstraat (9 specimens) l82 Fig. 12. Distribution o f silver triquetrum coins o f variant i. After Roymans 2006, fig. 12, with supplement. 1 Dunsberg (4 specimens); 2 Echt (2 specimens); Amby-Hagenstraat (4 specimens) Fig. 13. Distribution o f silver triquetrum coins o f variant t. After Roymans 2006, fig. 10, with supplement. 1 Empel (2 specimens); 2 Kessel; 3 Maren (=Kessel); 4 Dünsberg; 5 Echt; Amby-Hagenstraat (8 specimens); 6 MöhneseeKörbecke; 7 Ickern 183 THE 3 DATING OF THE HOARD Under the current dating system, staters of the Scheers 31 type are regarded as the last series of gold coins from Northern Gaul, struck in the m id-lst century and in particular in the 50s BC .25 This dating is based on the metrological attributes (an extremely low gold content and weight) and the late position ing of the coin group in the general typochronology for Gallo-Belgic gold staters. The coin images are derived from prototypes (fig. 5) that can be placed towards the end of LT D1 and the beginning of LT D2, roughly between 90 and 60 BC .26 Also significant for the dating of the Scheers 31 staters is their association with a POTTINA stater in the Heers hoard .27 In the Northern Gallic region, gold staters bearing a legend are dated to the m id-lst century.28 The presence of a POTTINA stater in Alesia (where Caesar inflicted a decisive defeat onVercingetorix and his Gallic army in 52 BC )29 means that this coin will have been minted before 52 BC and it is likely that the coin from Heers was also added to the hoard before 52 BC. The fact that the Amby hoard contains both class la and lb Eburonean staters is relevant for the relative dating of the hoard. It suggests that the hoard is o f a slightly later date than the Fraire and Heers hoards, in which only class la is represented.30 However, the fact that the Heers and Amby hoards contain no Scheers 31 staters of class II31 suggests that both were buried before the emission o f class II staters began. This will have taken place before 50 BC, since there are few indications in the period after Caesars conquest of a continuation of native gold emissions in Northern Gaul.32 As for silver triquetrum coins, the Amby hoard confirms the picture of a significant chronological difference between a series of early variants of additional marks (the Lith group) which had virtually ceased circulating in the mid-Augustan era, and a series of later variants (the Bochum group) which were still in full circulation then. The association with gold staters of the Scheers 31-1 type in both the Amby and Fraire hoards33 suggests that production of Lith-group coins must have started earlier than originally assumed —shordy before or during Caesar’s conquest rather than in the first two decades thereafter.34 Half of the Echt hoard (fig. 14) featuring exclusively triquetrum coins consists o f silver specimens of the Lith group with early additional marks that are also present in Amby. The other half, however, is made up of coins with a higher copper content and additional marks which are absent from Amby and may be of a slighdy later date.35 We may therefore conclude that the Amby hoard was buried shordy before 52 BC. The gold hoards of Fraire, Heers and Amby can be placed in a chronological sequence, with Amby as the latest.36 Signifi25 Haselgrove (1999) dates the coins to the second half o f 31 his phase 3 (90-60 BC). However, the current dating information argues for placing them at the beginning of his phase 4 (60-20 BC). absence in the Amby hoard relevant to the chronology. 32 26 See the detailed argumentation in Roymans 2004, 38 ff. According to Schulze-Forster (2002, 113) the gold Staters o f class II are well represented in the M aastricht/ Tongres region (see pi 40-41, Appendix 3), making their Haselgrove 1999,149 ff.; Roymans 1990,140 ff.; Loscheider 1998,156-157. 33 The four triquetrum staters from the Fraire hoard were rainbow cups o f the triquetrum type (Mardorf group) initially published as being composed o f electrum. It now circulated until the second quarter o f the 1st century and emerges, however, that they are silver coins. O ne coin in the Stieldorf hoard they are associated with Treveri even has additional mark variant (fa) o f the Lith group staters Scheers 30 type, the prototype o f the reverse of and is a die match with four coins (nos 82 to 85) from the Eburonean staters. Haselgrove (1999,139) places the Scheers 30-IV staters in LaTene D2a (90-60 BC). the Amby hoard. 34 For the original dating, see Roymans/Van der Sanden 35 36 Roym ans/H iddink 2006. 27 Scheers/Creemers, this volume. 28 Loscheider 1998,151; Haselgrove 1999,150. 29 See Fischer 2001, 38, no. 678. The coin was recendy found in ‘Camp C ’. 30 For the Fraire hoard, see Scheers/Creemers, this volume. 184 1980,212; Roymans 2004,76-82. For the relative chronology o f the hoards o f the FraireAmby horizon, see also Roymans/Scheers, this volume, 7 ff. Fig. 14. Hoard find o f silver rainbow cups from Echt. Foto Restaura. cant too is the conclusion that silver triquetrum staters of the Lith group with early additional marks are contemporaneous with the gold Scheers 31-1 staters. The Echt hoard appears to have been buried somewhat later than the Amby hoard. rainbow 31-la 31-lb 1. Fraire 4 2. Heers 116 - 3. Amby 21 20 31-11 - stater Nervii-I 4 - 78 - Nervii-ll 47 38 8 30 - Table 3. Relative chronology o f the gold hoards o f Fraire, Heers and Amby. triquetrum coins Lith group Bochum group a b c t If Fraire 3 1 - - - Amby 24 24 9 8 3 4 Echt 4 1 1 - 2 / m d 2 - - 41 1 2 8 - k Scheers 31-1 4 - Table 4. Composition according to additional mark variant o f the triquetrum staters in the Fraire, Amby and Echt hoards. i *5 H I S T O R I C A L C O N T E X T : T H E R E V O L T OF T H E AND THEIR CONTACTS W ITH TRANSRHENISH 4 EBURONES GERMANI Celtic numismatics has a rich tradition of linking numismatic data to historically documented events. A well-known example is Simone Scheers’ theory that links a series of stylistically related gold emissions from Northern Gaul to the period of Caesar’s Gallic Wars (58-51 BC). Nowadays we are rather more critical about such historical associations.37 The increase in the dating evidence has revealed a need to adjust Scheers’ dating model. Many coin series have been shown to have started earlier than originally assumed. Against this background, attributions of coin emissions to the period of the Roman conquest must now be properly substantiated. The above question is also relevant when it comes to interpreting the Amby hoard. And yet, given the above dating of the hoard, it does seem legitimate in this case to establish a connection with the Caesarean conquest. The particular focus here is Caesar’s description of the Eburonean revolt of 54/53 and the networks of alliances forged by the Eburonean leader Ambiorix. He is said to have received sup port during the revolt from the Suebi and probably the anti-Roman faction of the Ubii, both of whom inhabited the east bank o f the Rhine .38 We can perhaps conceive of this Germanic support as having assumed the form of mobile groups o f cavalry who crossed the Rhine .39 After their victory over a Roman army, the Eburones were struck by merciless Roman punitive campaigns in 53 and 51, which amounted in fact to an attempt at genocide. The presence in the Amby hoard of a group of silver rainbow cups points to contacts between the Eburones and Germanic groups on the eastern side of the Rhine. We should think here of alliances between tribal leaders, also involving payments in the form o f large numbers o f gold or silver coins. There are two possible interpretations. According to the first, the triquetrum staters of the Lith group are viewed as having been struck in the eastern Rhine region, probably on the Dunsberg. Ambiorix’s Germanic allies would then have brought the coins with them, a hypothesis developed by Heinrichs.40 This hypothesis is improbable, however, if we assume that the Fraire hoard —and hence the start of circulation for silver rainbow cupsE dates to before the Eburonean revolt of 54.41 We therefore posit a second and in our view more plausible interpretation in which the Lith-group staters were struck by a subgroup of the Eburones themselves, who had probably taken over this coin series from related Ger mani from the eastern Rhine region before the Roman conquest. There is clear evidence (see above) that some coins were produced on Eburonean territory itself, particularly in the Kessel/Lith and/or Maastricht/Tongres regions. According to this interpretation, the support that Germani from the eastern Rhine region gave to the Eburonean revolt should not be regarded as a one-off incident, but should be viewed in the context of pre-existing political relations between the two parties. Small-scale migrations to Eburonean territory by grou'ps from across the Rhine may also have been a factor here. The earlier date for Lith-group triquetrum staters means that we need to reconsider the historical context into which they have been placed. That context was the ethnogenesis of the Batavian commu nity in the post-Caesarean period between 50 and 15 BC .42 Now that we know that Lith-group coins 37 Cf. Loscheider 1998,147 ff; Haselgrove 1999,119. rones from Germani from east o f the R hine would lead 38 Caesar, BG 5.27; 6.5. Cf. Heinrichs 1999. us to expect a coin flow in the opposite direction —an eastwards flow as a ‘reward’ for their support. 39 Cf. Caesar, BG 6.35 and 4.9 on the raid by 2000 Sugambrian cavalrymen who crossed the R hine in 53 to 41 T he Fraire hoard can be dated to c. 60 BC; the minting plunder Eburonean territory, and the large cavalry force o f the coin types present in the hoard may have begun o f Tencteri and Usipetes who invaded Eburonean terri some time earlier. See the synthesis in Roymans/Scheers, tory in 55. 40 Heinrichs 1999. However, assistance granted to the Ebu 186 this volume, 1 ff. 42 Roymans/Van der Sanden 1980; Roymans 2004,96. were already in circulation at the time of Caesar’s conquests, we must understand them primarily in an Eburonean context. Moreover, we find clusters not only in the Dutch river region but in the Maastricht area as well. However, it was above all in the Batavian river region that later rainbow cups of the Bochum group continued to circulate and perhaps also in part be minted in the second half of the 1st century BC. There are good arguments for assuming that the Amby hoard was buried shortly before 52 BC. It then seems obvious to establish a link with the crisis surrounding the Eburonean revolt of 54-53 BC and subsequent Roman punitive campaigns. In such a situation, it is not difficult to imagine individuals safeguarding their mobile'wealth in the form of Versteckdepots. 5 LATE I R O N AGE H A B I T A T I O N IN THE M A A S T R I C H T AREA In the Maastricht area, there have in recent decades been numerous reported finds of material from the Late Iron Age. We present an outline below, with a particular focus on Celtic coins and fragments of glass LaTene bracelets. In this way we seek to shed fight on habitation in the final stages of the Iron Age. The information presented here is also reproduced on a distribution m a p (fig. 15). The present-day Meuse valley around Maastricht owes its origins to the tectonic uplift of South Lim burg, the more or less contemporaneous incision of the Meuse and the alternating climates of ice ages and warm periods. It was the interplay o f these factors that gave the river valley its fundamental shape. On both sides o f the Meuse there are sedimentary deposits of old river terraces that determine the character o f the landscape. During the ice ages of the Pleistocene, fine soil particles were blown in by the wind and deposited as loess sediments over large areas. The loess soil was highly susceptible to erosion and was easily washed away when rain fell on vegetation-poor slopes. The loess was then deposited at the foot of these slopes, resulting in a layer of colluvium sometimes several metres thick. The Geomorphological Map of the Netherlands43 tells us that the find site of the Amby hoard is located on the terrace o f Eisden-Lanklaar.This terrace dates from the penultimate Ice Age, known as the Saafian, which occurred between 225,000 to 130,000 years BP. As the find site is situated several hundred metres from the somewhat flatter edge o f the terrace, no colluvium has been detected. There have been various reports in recent decades o f stray Celtic coin finds in the Maastricht munici pality. A gold stater o f the Scheers 31 type (class II), ascribed to the Eburones, was discovered at Randwyck in the 1980s (fig. 15, site 12).The same type of stater (also class II) was discovered in Amby-Heihof in 2008, as well as a silver ‘rainbow cup’ o f the triquetrum type (site 19, fig. 16-a,b). Another silver rain bow cup of that type was found at Amby-Bodemsweg (site 20, fig. 16-c). A gold rainbow cup of South ern German origin was unearthed in 2007 on the Dausberg on the western periphery of Maastricht (site 24, outside the map sheet of fig. 15). A silver quinarius of the Scheers 58 type featuring the legend ANNAROVECI on both sides was found near the Zonhof tennis courts in Amby in the 1980s (site 21, fig. 16-d).This coin type can be dated to the third quarter of the 1st century BC and can probably be ascribed to the Tungri, who were emerging at that time .44The youngest Celtic coins from Maastricht are made of bronze and originate from the Augustan era; these are several AVAVCIA coins of the Scheers 217 type, with a swastika m otif on the obverse and a horse on the reverse.45 A class I coin was found in the Stokstraatkwartier (site 7), with the legend AVAVCIA above the horse, while two class II coins were 43 Geomorfologische kaart van Nederland, Staring Centrum 2004).We assume that the Eburones ceased to exist as an 1989. autonomous political community as a result o f Caesar’s 44 Roymans/Aarts 2009, 16, and fig. 12. The ethnogenesis o f the Tungri probably occurred in the second half o f the 1st century BC (cf. Vanderhoeven/Vanderhoeven punidve campaigns. 45 Cf. the recent synthesis o f this coin type in Aarts/Roymans 2009. I87 Fig. 15. Distribution o f Late Iron Age find sites in and around Maastricht. The num bering o f the find sites corresponds to the numbering in Appendix 1. An asterisk marks the location o f the Amby hoard. In red: route o f the R om an road from Bavai to Cologne and the extent o f the R om an vicus o f Maastricht. also discovered, one in the Stokstraatkwartier (site 6 ) and the other on the right bank of the Meuse just south ofWyck (site 11).We may assume that the AVAVCIA coins from the Maastricht city centre belong to the first stage of the Roman vicus that developed where the Bavai-Cologne road crossed the Meuse. Glass LaTene bracelets are another diagnostic artefact of the Late Iron Age. A total o f 31 fragments have been found within the municipality of Maastricht, distributed across 14 find sites (figs 17 and 18).46 46 Dijkman 1990. Fig. 16. Several stray Celtic coin finds from Amby. Scale 2:1. a-b Amby-Heihof; c Amby-Bodemsweg; d Am by-Zonhof Fig. 17. Fragments o f Late Iron Age glass bracelets from the municipality o f Maastricht. 189 Fig. 18. Fragments o f Late Iron Age glass bracelets from Maastricht. Scale 1:1. The numbering refers to the find sites in figure 2 and Appendix 1. As this form of arm jewellery occurs exclusively as grave goods in women’s graves, it will have been worn by women. The cobalt-blue glass variants, often decorated with applied trails of yellow glass paste, predominate in this region. Specimens with a 5-ribbed profile are far more dominant than ones with a D-shaped profile. These bracelets circulated from the later 3rd century BC until the Augustan period, with the wide 5-ribbed type apparently belonging to the early variant and the D-shaped type to the relatively late variants.47 As is also the case elsewhere in the Netherlands, most bracelet fragments are 47 Roymans/Verniers 2009. 190 Fig. 19. Earthenware spout in the shape o f a cow’s or deer’s head, found at Hotel Derlon in Maastricht. Beginning o f 1st century AD. Scale 1:1. stray surface finds. In some cases they have been discovered in secondary contexts in Merovingian graves on the Vrijthof (site 3), in the Pandhof of the St Servatius church (site 2) and in the Rothem cemetery (site 16).48 The largest concentration of bracelet fragments was found in Randwyck (sites 12 and 13), where intensive prospection by B. Knippels in the 1980s uncovered about twenty specimens.49 Only three examples are known from Amby and Rothem. It is problematic that the archaeological context of a large proportion of the find sites for Late Iron Age glass bracelets and coins cannot be pinpointed more closely. However, several locations have also yielded other find material, in particular pottery, which shows that there was definitely a settlement 48 Braat 1956, fig. 17, f. 49 Dijkman et al. 1986,187-189; Dijkman 1990, table. Fig. 20. Spoon-bow fibulae from the Late Iron Age from the Meuse at the St Servatius bridge in Maastricht. Scale 1:1. there.50 In the Maastricht city centre, the Stokstraatkwartier and the St Servatius church and its environs have yielded stray sherds of handmade pottery that can be dated to the Late Iron Age or earliest Roman period. One special object is a spout in the shape of a cow’s or deer’s head (fig. 19), found alongside the remains of a cobbled road discovered near Hotel Derlon (site 9) under a culture layer from the second quarter of the 1st century AD.51 This material may well belong to the earliest phase of the Rom an vicus of Maastricht. It is still unclear whether a native pre-Augustan settlement preceded the vicus, but vari ous fragments of La Tene bracelets and five La Tene D2 spoon-bow fibulae of the Kessel type from the Meuse at the St Servatius bridge (fig. 20) do point in that direction .52 A further indication is a piece of Dijkman 1989, table 1. from the same workshop. For the distribution and dating Panhuysen 1984,36. o f this type o f fibula, see Roymans 2004, fig. 7.10. Four fibulae are virtually identical and perhaps originate 192 wood recovered during the excavation of the inner garden (pandtuin) of the Onze Lieve Vrouwebasiliek (site 8 ). It is dated dendrochronologically to shortly after 59 BC ± 6 , attesting to building activity in that vicinity in the 1st century BC .53 A large concentration of pottery belonging to the final phase of the Iron Age was found in Koebroek (site 13) on the east bank of the Meuse. It comes from the fill of a channel that can be identified as a former arm of the Meuse. This ceramic assemblage can be dated typochronologically to between 150 BC and 50 AD, which is confirmed by two C l 4-datings.54Various glass bracelet fragments were discovered in the immediate vicinity of this channel. Until recently we had no information at all about the pre-Roman funeral ritual in the Maastricht region. However, two clusters o f cremation burials from the Middle and/or Late Iron Age were recendy excavated at Itteren in the north of the municipality of Maastricht (outside the map sheet in fig. 15). A cluster o f four graves (one containing pieces of a glass La Tene bracelet) lay within two rectangular ditched enclosures; the other cluster of about 2 0 graves was made up of simple grave pits without bound ary ditches.55 In the Maastricht region, a total of 23 find sites are now known for material from the Late Iron Age. These habitation features are very unequally distributed and the distribution m a p (fig. 15) no doubt presents an incomplete picture of habitation at that time. We can see three notable concentrations: in the old city centre of Maastricht (Stokstraatkwartier), on the eastern bank of the Meuse in Randwyck and in Amby. The picture that we now have suggests fairly dispersed habitation in the form of small settle ments. The only salient feature is the presence o f a hill fort at Caestert south of Maastricht, which can be dated to the Late Iron Age.56 A Roman settlement with centre functions developed at Maastricht from the second decade BC. A surprising number of Celtic coins have been found east of the village of Amby in recent years, as well as several scattered glass bracelet fragments. Although settlement features from the Early Iron Age and a Late Bronze Age urnfield have recently been discovered at Ambyerveld,57 this area lacks any distinct settlement features from the Late Iron Age. 6 CONCLUSIONS The gold and silver hoard of 119 coins discovered in the village of Amby must have been buried on an uninhabited site shortly before 52 BC. The hoard comprises two different groups of coins, gold staters o f the Scheers 31 type and silver rainbow cups of the triquetrum type. In the fight of their dating and distribution, both coin groups may be regarded as Eburonean emissions, although they do appear to have been struck by different subgroups. Given the considerable wealth that the hoard represents, its owner will have belonged to the upper echelon of the Eburonean community, and the coins must have been acquired in several larger sets. The dating to the period of the Roman conquest suggests that this was probably a Versteckdepot, concealed during a time of crisis with the intention of later recovery. This interpretation does not rule out the possibility that religious ideas and/or ritual actions were associated with the hoard’s burial at this particular location. The Amby hoard constitutes an important archaeological document for several reasons: a. It is the first Late Iron Age gold hoard found on Dutch soil. b. The late dating of the hoard argues for a fink to Caesar’s Gallic conquests and in particular to the Eburonean revolt of 54/53 BC. 53 Unpublished dating R IN G 1996. 54 Panhuysen et al. 1990, 216. GrN-16299: 2030 ± 70 BP 55 and G rN -16300:2180 ± 120 BP. preparing a publication. 56 57 Verhoeven 2008. Van der M ark/Schorn 2008. Oral communication by Adrie Tol, A R C H O L, who is 193 c. The coin hoard, and in particular the presence of silver rainbow cups, confirms the existence of close contacts between Eburones and Germanic groups on the eastern bank of the Rhine. d. The earlier dating of the silver rainbow cups of the Lith group (c. 60-50 BC instead of 50-30 BC) has major implications for historical interpretation.This coin series began as an Eburonean emission and not as a Batavian one, as previously supposed. e. The Amby hoard lends strength to the idea that the Maastricht/Tongres region was the main political core of the Eburonean polity. R E F E R E N C E S Aarts, J./N . Roymans, 2009: Tribal emission or imperial coinage? Ideas about the production and circula tion of the so-called AVAVCIA-coinages in the Rhineland, in J. Van Heesch/I. Heeren (eds), Coinage in the Iron Age. Essays in honour of Simone Scheers, London, 1-17. Braat, W.C., 1956: Le cimetière mérovingien de Rothem, commune de Meerssen, in Oudheidkundige Mededelingen uit het Rijksmuseum van Oudheden, N.R. 37, 68-81. Creemers, G./S. Scheers, 2007: Wichtige Fundstücke als Zeugnisse der Eburones in Nordostbelgien, in Krieg und Frieden. Kelten, Römer, Germanen, Bonn, 169-174. Derks,T, 1998: Gods, temples and ritual practices. The transformation of religious ideas and values in Roman Gaul, Amsterdam (Amsterdam Archaeological Studies 2). Dijkman,W , et al., 1986: IJzertijdvondsten uit Maastricht-Randwyck, Archeologie in Limburg, 30,187-189. Dijkman, W , 1989: Een vindplaats uit de Ijzertijd te Maastricht-Randwyck, Amersfoort (Nederlandse Arche ologische Rapporten 8 ). Dijkman, W , 1990: Glazen armband fragmenten uit de late Ijzertijd in Maastricht, Archeologie in Limburg 45, 309-310. Fischer, B., 2001: Les monnaies gauloises: enseignements et questions, in M. Reddé/S. von Schnurbein (eds), Fouilles et recherches Franco-Allemandes sur les travaux militairs Romains autour du Mont-Auxois (1991-1997), 2,12-68. Fitzpatrick, A.P., 2005: Gifts for the golden gods: Iron Age hoards of torques and coins, in C. Haselgrove/D. Wigg-Wolf (eds), Iron Age coinage and ritual practices, Mainz (Studien zu Fundmünzen der Antike, 20), 157-182. Gobi, J./A. Hartmann/H.E. Joachim/V. Zedelius, 1991: Der spätkeltische Goldschatz von Niederzier, BonnerJahrbücher 191,27-84. Haselgrove, C., 1999: The development of Iron Age coinage in Belgic Gaul, The Numismatic Chronicle 159,111-168. Heinrichs, J., 1999: Zur Verwicklung ubischer Gruppen in den Ambiorix-Aufstand d.J. 54 v.Chr. Eburonische und ubische Münzen im Hortfund Fraire-2, Zeitschrift für Papyrologie und Epigraphik 127, 275-293. Loscheider, R ., 1998: Untersuchungen zum spadatènezeitlichen Münzwesen des Trevererlandes, Archaeologia Mosellana 3, 61-226. Nick, M., 2005: Am Ende des Regenbogens... Ein Interpretationsversuch von Hortfunden mit keltischen Goldmünzen, in C. Haselgrove/D. Wigg-Wolf (eds), Iron Age coinage and ritual practices, Mainz (Studien zu Fundmünzen der Antike, 20), 115-155. Panhuysen,T.A.S.M., 1984, Maastricht staat op zijn verleden. Vierkant Maastricht, Maastricht. Panhuysen, T./P. Boyens/W. Dijkman/R. Hulst, 1990: Opgravingen door het Gemeentelijk Oudhei dkundig Bodemonderzoek, Maastricht (GOBM), in H. Stoepker (ed.), Archeologische kroniek van Limburg over 1988 en 1989, in Publications de la Société Archéologique et Historique dans le Limbourg 126, 214-230. 194 Roymans, N., 1990: Tribal societies in Northern Gaul. An anthropological perspective, Amsterdam (C ingula 12). Roymans, N., 2004: Ethnic identity and imperial power The Batavians in the Early Roman Empire, Amsterdam (Amsterdam Archaeological Studies, 10). Roymans, N./W. van der Sanden, 1980: Celtic coins from the Netherlands and their archaeological con text, Berichten van de Rijksdienst voor het Oudheidkundig Bodemonderzoek 30,173-254. Roymans, N./T. Derks (eds), 1994: De tempel van Empel. Een Herculesheiligdom in het woongebied van de Bataven, 's-Hertogenbosch. Roymans, N ./H . Hiddink, 2006: Ein Hortfund spatlatènezeitlicher“Regenbogenschiisselchen“ der Nor dgruppe aus Echt (Niederlande), Archäologisches Korrespondenzblatt 36, 225-236. Roymans, N./J. Aarts, 2009: Coin use in a dynamic frontier region. Late Iron Age coinages in the Lower Rhine area,Journal of Archaeology in the Low Countries 1, 5-26. Roymans, N./L. Verniers, 2009: Glazen La Tène-armbanden in het gebied van de Nederrijn. Typologie, chronologie en sociale interpretatie, Archeobrief 13, no. 4, 22-31. Scheers, S., 1977: Traité de numismatique celtique II. La Gaule belgique, Paris. Scheers, S., 1996: Frappe et circulation monétaire sur le territoire de la future civitas Tungrorum, Revue beige de numismatique et de sigillographie 142, 5-51. Schulze-Förster, J., 2002: Die latènezeitliche Funde vom Dünsberg, Marburg (PhD dissertation University o f Marburg). Stead, I.M., 1991:The Snettisham treasure: excavations in 1990, Antiquity 65, 447-464. Vanderhoeven, A./M . Vanderhoeven, 2004: Confrontation in archaeology. Aspects of Roman m ilitary presence in Tongeren, in F.Vermeulen/K. Sas/W. Dhaeze (eds), Archaeology in confrontation. Aspects of Roman military presence in the Northwest, Ghent (Archaeological Reports Ghent University 2), 143-154. Van der Mark, R ./E . Schorn, 2008: Maastricht-Ambyerveld. Inventariserend veldonderzoek door middel van proefsleuven (BAAC rapport A-07.0030). Van Impe, L., et al. 1997/1998: De Keltische goudschat van Beringen (prov. Limburg), Archeologie in Vlaanderen 6,1-124. Verhoeven, M.P.F., 2008: Studieopdracht naar een archeologische evaluatie van het plateau van Caestert (Riemst, prov. Limburg), Weesp (RAAP-Rapport 1769). 195 A P P E N D I X I. LIST M A T E R I A L IN T HE OF SITES W I T H M U N I C I P A L I T Y LATE OF I R O N AGE M A A S T R I C H T . The numbering corresponds to the numbering in the distribution map in fig. 15. Unless otherwise stated, all fragments of LaTène bracelets are of cobalt-blue glass. 1 Caberg. Klinkers quarry. Glass bracelet fragment. 5-ribbed (7a). Settlement. Unpublished. 2 St Servatius church. Merovingian cemetery. Glass bracelet fragment. 5-ribbed (7a?). Stray find. Dijkman 1990, table, no. 22. 3 Vrijthof. Merovingian cemetery. Glass bracelet fragment. 5-ribbed (7a). Stray find. Unpublished. 4 Tongersestraat. Glass bracelet fragment. Stray find. 4-ribbed (7c?), light green. Dijkman 1990, table, no. 28. 5 Jodenstraat. Glass bracelet fragment. 5-ribbed (7a). Stray find. Dijkman 1991, fig. 6,1. 6 Havenstraat. Bronze AVAVCIA coin, class II, and pottery. Setdement. Unpublished. 7 Onze Lieve Vrouweplein. Mabro-terrein. Bronze AVAVCIA coin, class I. Unpublished. Two glass bracelet fragments. 2 x 5-ribbed (7a). Stray finds. Dijkman 1990, table, nos. 16,17. 8 Onze Lieve Vrouwebasiliek. Inner garden. Wood with dendrochronological dating to the Late Iron Age. Stray find. Unpublished. 9 Plankstraat. Derlon. Path and pottery. Setdement. Panhuysen 1984, 36. 10 Meuse river at St Servatius bridge. Five bronze spoon-bow fibulae o f the Kessel type, La Tène D2. Found in 1934 by G. Sibilla. Fig. 20. Probably a hoard find. Roymans 2004, fig. 7.10 (distribution map). 11 Céramique. Bronze AVAVCIA coin, class II. Stray find. Unpublished. 12 Randwyck. Gold stater Scheers 31-11. Found in the 1980s by C. Peusens andT. Poels. Roymans 2004, Appendix 4.2. Five glass bracelet fragments: 4 x 5-ribbed (7a) and lx D-shaped profile (3a). lx fight blue with decoration in yellow glass paste. Stray finds. Dijkman et al. 1986, table, nos. 1-3,12. 13 Koebroek. Pottery from channel fill. Unpublished. 16 glass bracelet fragments, lx D-shaped profile (3a), 1 x 3-ribbed (6 a), 2 x 3 or 5-ribbed, lx 4-ribbed (11), 8 x 5-ribbed (7a), including 3x decoration with yellow glass paste (7b), lx 7-ribbed. Stray finds. Dijkman et al. 1986, table, nos. 5-11. Dijkman 1990, table, nos. 13,14,18-21, 23,25 and 26. 14 Vogelzang.Two glass bracelet fragments, lx D-shaped profile (3a), lx 5-ribbed (7a), light green. Stray finds. Dijkman et al., 1986, table, no. 4. Dijkman 1990, table, no. 24. 15 De Heeg. Glass bracelet fragment, lx 3-ribbed (6 a). Stray find. Dijkman 1990, table, no. 27. 16 Rothem. Merovingian cemetery. Glass bracelet fragment. 5-ribbed (7a). Stray find. Braat 1956, fig. 17, grave 16, f. 17 Rothem. Geuldal. Glass bracelet fragment. 5-ribbed (7a). Stray find. Unpublished. 18 Amby. Ambyerveld. Glass bracelet fragment. 3-ribbed (8 a). Stray find. Dijkman 1990, table, no. 15. 19 Amby. Heihof. Gold stater Scheers 31-11 (4.96 g, fig. 16) and silver rainbow cup of triquetrum type, variant c (5.15 g, fig. 16). Detector finds P. Derks. Coll. Derks. Unpublished. 20 Amby. Bodemsweg. Silver rainbow cup of the triquetrum type, variant c (fig. 16). Detector find F. Schröder. Schröder collection. Unpublished. 21 Amby. Zonhof. Silver quinarius Scheers 58 type with legend ANNAROVECI (fig. 16). Detector find P. Stassen. Stassen collection. Unpublished. 22 Amby. Hagenstraat. Coin hoard. Hoard find. This publication. 23 Itteren. Emmaus. Ditches and two clusters of cremation graves from the Middle and Late Iron Age. A fragment of a glass La Tène bracelet comes from a grave. Unpublished. Oral communication from Adrie Toi, ARCHOL bv. 24 Maastricht. Dousberg. Gold southern German rainbow cup LT 9423. Weight 7.37 g. Unpublished. Detector find H. Hansen 2007. Information from S. Scheers, Leuven. 196 A P P E N D I X 2. D E S C R I P T I V E LIST OF C O I N S F R O M T HE A M B Y HOARD The numbering corresponds to the numbering of the coins in Plates 1-12.0 = obverse die; R = reverse die. The coin composition was identified by means of surface analysis (X-ray fluorescence analysis; Wout ers, this volume, Appendix 5). No. class/ die no. variant weight analysis % collection (g) Au Ag Cu stater Scheers 31 type, Eburones 1 la 0 17- R 27 5.54 30 54 16 GDB Maastricht, no. 1 2 la 0 6-R8 5.54 28 56 16 GDB Maastricht, no. 2 3 la 0 9 - R 12 5.58 31 50 19 GDB Maastricht, no. 3 4 la 0 8 - R 11 5.66 5 la 0 3-R4 5.74 28 52 20 GDB Maastricht, no. 5 P. Curfs 6 la 010 - R 14 5.81 27 53 20 GDB Maastricht, no. 6 7 la 0 10- R 14 5.74 26 50 24 GDB Maastricht, no. 7 8 la 0 13 - R 20 5.70 30 47 23 GDB Maastricht, no. 8 9 la 0 6-R9 5.68 29 53 18 GDB Maastricht, no. 9 10 la 0 4-R5 5.83 11 la 0 18- R 28 5.68 29 47 24 GDB Maastricht, no. 11 12 la 0 18- R 30 5.58 29 37 34 GDB Maastricht, no. 12 13 la 0 18- R 30 5.71 30 42 28 GDB Maastricht, no. 13 14 la 0 18- R 30 5.71 27 42 31 GDB Maastricht, no. 14 15 la 0 1 3 -R 18 5.73 27 48 25 GDB Maastricht, no. 15 16 la 018 - R 29 5.64 17 la 015 - R 24 5.72 33 39 28 18 la 0 6- R9 5.82 29 40 31 GDB Maastricht no. 18 19 la 016 - R 26 5.77 29 54 17 GDB Maastricht, no. 19 20 la 0 3 -R 3 5.75 34 52 14 GDB Maastricht, no. 20 21 lb '0 19 - R 31 5.66 29 46 25 GDB Maastricht, no. 21 22 lb 0 19 - R 31 5.73 28 41 31 GDB Maastricht, no. 22 23 lb 0 19 - R 31 5.69 24 lb 0 19- R 31 5.69 30 41 29 GDB Maastricht, no. 24 25 lb 0 19- R 31 5.74 P. Curfs 26 lb 0 19- R 31 5.58 E. Heitzer 27 lb 0 1 9 -R 31 5.62 28 lb 0 19 - R 31 5.65 32 45 23 GDB Maastricht, no. 28 29 lb 0 19 - R 31 5.67 30 45 25 GDB Maastricht, no. 29 30 lb 0 1 9 -R 31 5.60 P. Heydelaar 31 lb 0 20 - R 32 5.66 R Curfs 32 lb 0 20 - R 32 5.71 31 40 29 GDB Maastricht, no. 32 33 lb 0 20 - R 32 5.54 32 49 19 GDB Maastricht, no. 33 34 lb 0 20 - R 32 5.62 35 lb 0 20 - R 32 5.68 36 lb 0 21 - R 33 5.66 P. Heydelaar GDB Maastricht, no. 16 GDB Maastricht, no. 17 P. Curfs E. Heitzer R. Speets R. Speets 28 51 21 GDB Maastricht no. 36 37 lb 0 21 - R 33 5.65 29 53 18 GDB Maastricht, no. 37 38 lb 0 21 - R 33 5.63 28 50 22 GDB Maastricht, no. 38 197 115 lb 0 20 - R 32 5.58 R. Speets 117 la 0 5-R7 5.7 P. Curfs 119 lb 0 1 9 -R 31 5.62 P. Curfs rainbow stater, triquetrum type, Eburones 39 a 5.67 40 a 5.87 14 82 4 GDB Maastricht no. 39 GDB Maastricht, no. 40 41 a 5.80 GDB Maastricht, no. 41 42 a 5.74 GDB Maastricht, no. 42 43 a 5.30 GDB Maastricht, no. 43 44 a 5.67 GDB Maastricht, no. 44 45 a 5.57 GDB Maastricht no. 45 46 a 5.20 GDB Maastricht, no. 46 47 a 5.64 GDB Maastricht, no. 47 48 a 5.73 GDB Maastricht, no. 48 49 a 6.01 50 a 5.48 51 a 5.58 GDB Maastricht, no. 51 52 a 5.59 GDB Maastricht, no. 52 53 a 5.77 GDB Maastricht, no. 53 54 a 6.15 P. Heydelaar 55 a 5.32 P. Curfs P. Curfs 13 81 6 GDB Maastricht, no. 49 GDB Maastricht, no. 50 56 a 5.49 57 a 5.55 E. Heitzer 58 a (4.87) E. Heitzer 59 a 5.75 E. Heitzer 60 a 5.25 R. Speets 61 a 5.56 R. Speets 62 a 5.56 P Heydelaar 63 b 5.74 GDB Maastricht, no. 63 GDB Maastricht, no. 64 64 b 5.60 65 b 5.74 66 b 5.04 67 b 6.19 GDB Maastricht, no. 67 GDB Maastricht, no. 68 GDB Maastricht, no. 65 13 84 3 GDB Maastricht, no. 66 68 b 6.14 69 b 5.60 70 b 5.90 71 b 6.18 72 b (4.90) GDB Maastricht, no. 72 73 b 4.43 GDB Maastricht, no. 73 74 b 5.65 GDB Maastricht, no. 74 75 b 5.60 GDB Maastricht no. 75 76 b 5.88 P. Curfs 77 b (5.18) P. Curfs 78 b 4.41 E. Heitzer 79 b 5.49 R. Speets 80 b 5.60 R. Speets 81 b (4.61) P. Curfs 82 b 6.10 GDB Maastricht, no. 82 198 GDB Maastricht, no. 69 13 82 5 GDB Maastricht, no. 70 GDB Maastricht, no. 71 83 b 5.78 84 b 5.87 85 b (4.94) 86 c 6.18 87 c 88 14 77 9 14 82 4 6.40 14 70 16 GDB Maastricht, no. 87 c 6.16 12 83 5 GDB Maastricht, no. 88 89 c 6.06 90 c 6.05 GDB Maastricht, no. 90 91 c 5.78 GDB Maastricht, no. 91 92 c 5.63 GDB Maastricht, no. 92 93 c 6.04 GDB Maastricht, no. 93 94 c (4.94) R. Speets 95 t 5.78 96 t 6.31 97 t 5.91 98 t 5.67 GDB Maastricht, no. 98 99 t 5.40 GDB Maastricht, no. 99 100 t 5.81 GDB Maastricht, no. 100 101 t 5.67 GDB Maastricht, no. 101 102 t 5.42 E. Heitzer 103 U 6.15 GDB Maastricht, no. 103 104 U 6.15 GDB Maastricht, no. 104 105 U 5.71 P. Curfs 106 i 5.67 GDB Maastricht, no. 106 107 i 5.59 GDB Maastricht, no. 107 108 i 5.71 GDB Maastricht, no. 108 109 i (4.53) 21 71 8 GDB Maastricht, no. 109 110 k 5.83 12 82 6 GDB Maastricht, no. 110 111 indet. (5.06) P. Curfs 112 indet. fragm. E. Heitzer 113 indet. fragm. R. Speets 114 a 5.67 R. Speets 1

Log In

Late Iron Age Gold Hoards from the Low Countries and the Caesarian Conquest of Northern Gaul, AAS 18 / Atuatuca 3, Publications of the Gallo-Roman Museum of Tongeren, Amsterdam, 2012. COMPLETE EDITION

Late Iron Age Gold Hoards from the Low Countries and the Caesarian Conquest of Northern Gaul, AAS 18 / Atuatuca 3, Publications of the Gallo-Roman Museum of Tongeren, Amsterdam, 2012. COMPLETE EDITION

Related Papers

RELATED PAPERS