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
116
b
5.95
P Curfs
118
k
GDB Maastricht, no. 83
GDB Maastricht, no. 84
GDB Maastricht, no. 85
GDB Maastricht, no. 86
GDB Maastricht, no. 89
13
80
7
GDB Maastricht, no. 95
GDB Maastricht, no. 96
14
80
6
GDB Maastricht, no. 97
P. Curfs
199
PLATES
AH coins are depicted on the scale 2! 1. The numbering of the coins corresponds to the numbering in
Appendix 2.
201
M A A S T R I C H T - A M B Y,
A \b>
'X. \
if t t v '
202
‘ *
Y
r
PLATE
I
M A A S T R I C H T - A M B Y,
PLATE
2
203
M A A S T R I C H T - A M B Y,
204
PLATE
3
M A A S T R I C H T - A M B Y,
PLATE
4
205
M A A S T R I C H T - A M B Y,
206
PLATE
5
PLATE
6
SSDw
M A A S T R I C H T - A M B Y,
207
M A A S T R I C H T - A M B Y,
208
PLATE
7
M A A S T R I C H T - A M B Y,
PLATE
8
209
M A A S T R I C H T - A M B Y,
210
PLATE
9
M A A S T R I C H T - A M B Y,
PLATE
10
211
M A A S T R I C H T - A M B Y,
212
PLATE
II
M A A S T R I C H T - A M B Y,
PLATE
12
213
Analytical investigation o f Late Iron Age gold hoards from the Low
Countries
Helena Wouters
1
2
3
4
5
Introduction
Analytical methodology
Techniques for determining the composition and textural properties of alloys. Some comments
Results
Interpreting variation in the alloy composition of coinages
6 Fragments o f gold torques
7 Conclusions
References
Appendix 1. EDXRF
Appendix 2. EDXRF
Appendix 3. EDXRF
Appendix 4. EDXRF
Appendix 5. EDXRF
Appendix 6 . EDXRF
I
analysis results for the Fraire hoard
analysis results for the gold finds of Orp-le-Grand
analysis results for the Thuin-1 hoard
analysis results for the hoard of Heers
analysis results for the hoard of Maastricht-Amby
analysis results for single coins from museums and private collections
INTRODUCTION
This study describes and interprets the results obtained from an analysis of the Late Iron Age gold coinag
es of the Fraire/Amby horizon with the objective o f gaining knowledge about their metal composition.
In particular, the coins from the hoards of Amby (Maastricht), Heers, Thuin-1 and Orp-le-Grand will
be studied. Their composition will be related to that of a number of coins from other public and private
collections, which were also analysed in the context of this study. The final aim of this investigation is to
obtain a better understanding o f the metal composition of the gold coinages of the Fraire/Amby horizon.
Within the study of Late Iron Age gold hoards found in the Low Countries in recent decades, a
material approach to coins suggests itself. Since typological and metallurgical developments of ancient
coinages are often closely finked, chemical analyses of coins may play a key role in understanding chro
nologies and relationships with contemporary coinages from neighboring areas.
With respect to terminology, coinages are often labelled as gold, silver or copper objects, but they
in fact rarely consist of a single pure element. In reality, gold artefacts and in particular coins are often
composed of a ternary Au-Ag-Cu alloy1 with significant traces of other elements such as antimony, tin
and lead. The purity of gold coins can be roughly determined by their colour. It has been shown that in
1
Indications with the chemical symbols: Au (gold), Ag
(silver), C u (copper).
215
0,00
Fig. 1. Ternary diagram o f the Au-Ag-Cu alloy system showing the colours o f the various alloys.
British coinages copper and silver were often added to electrum (the naturally occurring alloy o f gold
and silver) in order to make it more closely resemble pure gold.2 The ternary diagram in Figure 1 dem
onstrates the colours of the various Au-Ag-Cu alloys.3
The coins were produced by striking a metal blank or flan between two dies. W hen large groups of
coins are found together in hoards, it is a challenging task for numismatists to spot die links between coins.
The study of Roymans and Scheers in this volume gives the seriation schemes of coin dies represented
in the hoards of the Fraire/Amby horizon and ascribed to the Nervii and the Eburones.The analytical
investigation presented below will retain the classification and numbering of individual coins given by
the authors of previous contributions. The variation in alloy composition of the different classes of the
coinages will also be evaluated.
2
ANALYTICAL
METHODOLOGY
Most of previous metallurgical studies of Late Iron Age gold coinages4 are based on the major element
composition of the coins. Some studies also include the determination of trace element and platinumgroup element (PGE) content and Pb isotope signatures in an attempt to differentiate the metal sources.5
Gold coins are mostly alloyed with silver and copper. X-ray analysis therefore focused on the determina2
Northover 1992,235-299.
3
Bendall 2003,23; modified from Northover 1992.
coins from Manching in southern Bavaria (Gebhard et
4
Including British Gallo-Belgic (Cowell et al. 1987;
al. 1999) and from the Martberg on the Lower Moselle
Cowell/Bum ett 1988; Northover 1992), Central Gaul
(Bendall 2003 and 2008), and a study by Castelin (1985).
ish (Gondonneau et al. 1996; Nieto/Barrandon 2002)
2 IÖ
and western Germanic coinages (Hartmann 1976), gold
tion o f Au, Ag and Cu, representing over 9 9 % o f the coin compositions. The following analytical tech
niques were applied for the study of the coins.
EDXRF
Non-destructive energy dispersive X-ray fluorescence analyses (EDXRF) were carried out using mobile
ArtTAX 6 equipment.7 W hen assessing the potential of this technique, it is important to be aware of the
reduced penetration o f X-rays in objects made of gold and silver. This means that results relate to the
composition of the surface layer, which is generally not representative of the interior of the ancient
object. Indeed, about 90% of the fluorescent rays result from the first 10 to 20 pm of the surface.8 Despite
its resistance to corrosion, a remarkable difference in gold alloy composition is measured between the
surface and the core, as has been mentioned in previous work .9 The results should be interpreted with
caution, as will be explained in section 3. Energy dispersive X-ray fluorescence analyses were performed
according to the following parameters: W-anode X-ray tube, 50 kV, 600 pA, air, N i 25pm filter, 200 pm
collimator and a counting livetime of 2 0 0 s.
SEM-EDX
Energy dispersive X-ray analyses with the scanning electron microscope (SEM-EDX ) 10 are performed
by placing the whole coin in the sample chamber and analysing the surface properties or by embedding
samples in resin and polishing the surface to reveal the cross-section.The latter process was carried out on
20 selected coins and two jewellery fragments. The choice was based on the presence o f existing damage
or sometimes deep splits on the edges (appeared after flattening) or on its fragmentary state, so that they
could be sampled without risking their integrity. Samples of approximately 1 m m 2 were removed using
a fine jeweller’s saw. After being embedded in an epoxy resin,11 samples were ground flat and polished
down to 1 pm grade using silicon carbide abrasive and diamond pastes.
Quantitative measurements were performed using a standard ZAF procedure.Various NIST (National
Institute of Standards and Technology, USA) gold-silver (SRM 481 wires A till F) and gold-copper stand
ards (SRM 482 wires A tol F) were used to validate the quantitative settings.
The conditions for the measurements were a 35° take-off angle, a working distance o f 15 mm, an
accelerating potential of 20 kV, a beam current o f 1.2 nA and a counting livetime of 200s.
3
TECHNIQUES
FOR
AND TEXTURAL
MENTS
DETERMINING
PROPERTIES
OF
THE
COMPOSITION
ALLOYS.
SOME
COM
Table 1 presents the results of different non-destructive techniques applied to the same coin. Significant
variation can be concluded. The results show that EDXRF analyses, as performed within this project,
are in fairly good agreement with those of PIXE analyses. This data might suggest a similar agreement
between the results of SEM-EDX (surface) and neutron activation analyses (relating to average results for
5
Bendall 2003 and 2008.
8
Wouters 1990, 2.14-2.16.
s
BrukerAXS.
9
Wouters 1998,64-68 ; Wouters 2002,37-48.
The coins o f the hoard ofThuin-1 were analysed by Luc
10 JE O L JSM 6300 SEM, Oxford Instruments INCA EDX
Maes using a conventional ED X R F (Tracor Northern):
file IRPA.2L/20-DI: 81/2235.
with Si(Li) detector.
11 EpoFix resin by STRUERS A/S, DK.
217
the whole coin). In what follows, it will be explained that the results need to be interpreted with caution
as surface and core material can differ considerably.
object
analytical technique
Au
Ag
Cu
Fraire, F3, class la; Scheers/
Creemers, this volume
SEM-EDX, on the surface (LUC Diepenbeek)
32 ±2
19±4
49 ±6
PIXE, LARN, Namur
44
33
23
PIXE, IRHT, Orléans
59
20
31
Neutron activation analysis, Philips, Eindhoven
36
24
40
EDXRF, KIK, Brussels
44 ± 3
33 ±4
23 ±3
SEM-EDX, on the surface (LUC, Diepenbeek)
27 ±1
7± 1
66 ±2
PIXE, LARN, Namur
37
17
46
PIXE, IRHT, Orléans
48
13
39
Neutron activation analysis, Philips, Eindhoven
23
14
62
EDXRF, KIK, Brussels
37 ± 1
24 ±5
39 ±4
PMPK-R6218, Vervoz, class II
Table 1. Results o f different non-destructive techniques applied to two coins o f Scheers 31 type (Eburones).1:
f
!
A typical feature of ternary Au-Ag-Cu alloys used for coins is their heterogeneity, as nicely illustrated in
the backscattered electron images of the cross-sections of two coins (a gold Eburones stater and a silver
triquetrum stater) from the Amby hoard (fig. 2).These metallic segregations could be explained by the
binary and ternary phase diagram1213 for the Au-Ag-Cu system (fig. 3a,b) .14 Since copper and silver at
higher concentrations in the melt are not soluble in each other, an eutectic texture is produced during
cooling (called a zebra texture). Depending on the initial composition of the mix (whether the original
alloy mix is positioned on the Cu or the Ag side of the eutectic), the alloy will either nucleate silver- or
copper-rich primary phases. These dendrite phases grow until the remaining melt reaches the eutectic
composition (71.9% Ag). At this point the melt freezes, thereby producing intricate patterns of interlayer
silver- and copper-rich plates (eutectic texture) interstitial to the original dendrites. The BSE image of
the silver rainbow stater shows that each dendrite forms the core of an alloy crystal with the eutectic
mixture on the periphery. In the BSE image of the gold stater, the crystal grain boundaries are well pre
sented by the highlighted zones of enriched Au-Ag phases.
The eutectic mix content, relative to the dendrite crystals, depends on how close the original mixture
was to the eutectic composition, as can be derived from the ternary phase diagram of the Cu-Ag-Au
system (fig. 3b) .15 For example, for an alloy consisting of 30 % o f Au, the eutectic point will be at 45
% Ag and 25 % Cu. The closer the original mixture is to the eutectic one, the more the alloy will be
12 Analysis results from the laboratories o f LUC (Diepen-
metallic components, while a ternary system, which is
beek), LARN (Namur), IR H T (Orléans) and Philips
usually more complex, consists o f three metals. A phase
(Eindhoven) are reported in Scheers et al. 1991,34-37.
is a homogeneous chemical composition and uniform
13 A phase diagram is a diagram with axes o f temperature
material, describing one component in a metallic system.
and composition describing the different phases that may
14 Scott 1991,129.
occur in an alloy with a change in either composition
15 Scott 1991,134.
or temperature. A binary phase diagram consists o f two
218
Fig. 2. Backscattered image o f a cross-section o f two coins from the Amby hoard (a gold Eburones stater, A20 (left) and a silver
rainbow stater, A110 (right), SEM-EDX). In the image o f the gold stater, the dark and light grey zones represent respectively
A u-C u and Au-Ag metallurgical phases. In the image o f the silver stater, dark coloured C u-rich dendrite phases are surrounded
by a zebra phase with a eutectic texture formed by internally grown Cu and Ag-rich zone.
enriched with phases that have an eutectic texture. The phase differences observed in the BSE images in
Figure 2 can therefore be explained by the position in the ternary diagram as marked in Figure 3, right
(composition of the interior is presented).
The sometimes high values for the standard deviations obtained in Table 1 are inherently caused by
these complex textures. The heterogeneity resulting from the exsolution16 o f silver and copper-rich phas
es implies the need for a large electron beam to obtain data that is representative of the original alloy mix.
A large beam will provide information on the bulk composition. However, due to the technical limita
tions of the SEM technique, electron beams can be no larger than 50 pm. At least five measurements are
therefore required from each cross-section from which the average result and the corresponding stand
ard deviation are calculated. Results obtained from coins characterised by larger growths of Ag or Cu
dendrite-rich phases show rather high standard deviations compared to coins whose bulk composition
is closer to the eutectic mixture and which therefore show a limited growth of primary dendrite phases.
Another comment relates to the validity of the analytical methods of both XRF and SEM-EDX, and
therefore to whether the surface and near-surface analyses are representative of the whole coin. Because
the coins lay buried for about 2 0 0 0 years, both silver and/or copper may have been depleted from their
surfaces, thereby increasing the proportion of the more resistant gold and/or silver in the coin’s outer
layer. This corrosion can severely influence the surface analysis results, especially in cases where the
depleted layer is extremely thick, sometimes up to 100 pm. The leaching and diffusion of oxidised copper
during the corrosion process of a silver coin can cause severe changes in interior to surface compositions,
for example, an interior Ag concentration of 74 ± 3 wt%, with an Ag content of 96 ± 2 wt% on the
surface.17 The composition of both the surface layer and the interior part of the alloy therefore needs to
be taken into consideration.
16 A term used in mineralogy, a process by which an initially homogeneous solid solution separates into at least
removal o f any materials.
17 Wouters 1998, 66; Linke et al. 2006, 69.
two different crystalline minerals w ithout the addition or
219
Fig. 3. Phase diagrams with indication o f the different metallurgical phases that may occur in gold alloys with a change in either
composition or temperature. Left: binary phase diagram o f Cu-Ag system. Right: ternary phase diagram o f Cu-Ag-Au system
with indication o f the composition o f two examples: a gold Eburones stater, A20 (yellow) and a silver rainbow stater, A l l 0 (red).
The degree of alteration observed on the coins represented in the hoards of the Fraire/Amby horizon
is illustrated by the quantitative analyses results of Au, Ag and Cu obtained at the surface and interior
parts of two different coins, given in Table 2. The results indicate an important difference in copper and
silver content on the surface compared to the cross-section, which means that caution is also advised
for this study using absolute values. Consequently, it is extremely important in comparative studies not
to mix surface and interior analysis results but to use each for its own purpose. The non-destructive
surface-analysis results (without sampling) generally serve to screen and determine indicatively the overall
composition of the alloys and to establish different alloy series within the coins. The biggest advantage of
this type of analysis is that an endless number of objects can be reached without sampling, thus resulting
in higher statistical relevance for possible conclusions. Analysis of the interior, on the other hand, reveals
valuable complementary information; this can only be performed on carefully selected artefacts.
From the above observations we can conclude that the calculated standard deviations depend either
on the heterogeneity of the alloy and/or the degree of corrosion.
220
object
analytical technique
zone
Au
Afl
Cu
Scheers 31-1 stater, Eburones, Amby, A 20
EDXRF
surface
34 ± 9
52 ±3
14 ± 11
SEM-EDX
cross-section
28 ±2
35 ±3
37 ±2
EDXRF
surface
12 ± 2
82 ±1
6±1
SEM-EDX
cross-section
16 ±1
63 ±6
21 ±5
silver rainbow stater, Amby, A 110
Table 2. Quantitative analysis results o f Au, Ag and C u at the surface and interior part for a gold Eburones stater and a silver
rainbow stater.18
4
RESULTS
The surface of 154 selected coins belonging to the Late Iron Age gold hoards of Heers, Fraire, Amby,
Thuin-1 and Orp-le-Grand was subjected to primary non-destructive analysis using EDXRF. The
numismatic classification of the coins19 as triskeles staters of the Eburones (type Scheers 31, class I and
II), epsilon staters of the Nervii (type Scheers 29, class I and II), a POTTINA stater of the Treveri (type
Scheers 30, class V), an eye stater of the Rem i (type Scheers 30, class I), a gold stater with star of the
Bellovaci (type Scheers 25, class II) and Rhineland rainbow staters of the triquetrum type was maintained
in order to unequivocally evaluate the variation in metal composition within the Fraire/Amby horizon.
The results are presented in Table 3 as average values, together with their standard deviation. Con
centrations are expressed in weight %. The number of coins analysed within each numismatic group is
indicated (column n) as well as the minimum and maximum value if more than one coin is represented.
The individual results can be found in the appendices to this chapter, where they are grouped per hoard,
as well as in the descriptive list o f coinages at the end o f each chapter on the corresponding hoard in
this volume.
The measured variations in gold alloy compositions are summarised in a triangular diagram (fig. 4).
This ternary graph shows a general trend (marked as a red fine) of decreasing Au content, which stops
before the Scheers 30-V stater series and remains relatively constant as the copper content increases.
Such a trend would follow the same pattern observed elsewhere20 and in particular for the development
o f the provincial-Roman 21 and Central Gaulish coinages.22 This pattern is considered to be the result
of the progressive debasement of previous gold coinages through the addition of an approximately 2 : 1
mixture of Ag and Cu.
For the POTTINA staters (Scheers 30-V) and triskeles staters (Scheers 31), there was a change from
the previous debasement trend to a progressive increase in copper content (green line).This cannot be
explained in terms of debasement, as the gold content remains relatively stable, while more copper is
added at the expense of silver. Bendall stated that this change can only be due to the use of separate
supplies of refined gold, copper and silver, since further debasement of the alloy would lead to lower
Au contents.23 Additionally, it can be argued that a shortfall of silver might have occurred because of
larger production volumes o f Eburonean class II staters and the shared use of silver material for the
All measurements were performed at the Royal Institute
for Cultural Heritage, Brussels.
21 Cowell et al. 1987; Cowell/Burnett 1988; Northover
1992.
As given in the previous contributions in this volume by
22 Nieto/Barrandon 2002; Bendall 2003,2008.
Scheers/Roymans, .1 ff.
23 Bendall 2008.
Castelin 1985.
221
Au
type Scheers
hoard
n
(mean
+ SD)
(min)
Cu
Ag
(max)
(mean
+ SD)
(min)
(max)
(mean
+ SD)
(min)
(max)
26
24 ±6
25-11, Bellovaci ?
Heers
1
51 ±3
25 ±3
29-la, Nervii
Fraire
1
49 ±1
41 ± 1
10 ±1
38 ±1
11 ± 1
29-la, Nervii
Heers
1
51 ±1
29-la, Nervii
Thuin-I
27
55 ±3
49
60
29 ±3
24
35
16 ± 4
10
29-lb, Nervii
Thuin-I
13
55 ±3
49
60
30 ±4
21
36
15 + 6
9
29
29-1la, Nervii
Thuin-I
13
53 ±3
50
59
25 ±4
18
31
22 ±6
11
31
29-llb, Nervii
Thuin-I
16
56 + 2
52
60
29 ±3
24
34
15 ± 5
10
24
53
66
29 ±4
23
35
12±5
7
21
14
34
18
39
18
31
3
16
29-llb, Nervii
Heers
6
59 ±5
30-la, Remi
Heers
1
51 ±1
40 ±1
9± 1
30-V, Treveri
Heers
1
41 ±1
52 ±2
7±3
31-la, Eburones
Amby
17
29 ±2
31-la, Eburones
Fraire
1
44 ± 3
31-la, Eburones
Heers
31
37 ±2
31-la, Eburones
Orp-leGrand
1
44 + 3
31-lb, Eburones
Amby
10
30 ±2
flan
Heers
1
33 ±1
rainbow stater, triquetrum
Amby
12
14 + 2
rainbow stater, triquetrum
Fraire
1
10 ±1
26
35
48 ±6
37
55
33 ±4
32
42
37 ±6
23 ±3
25
47
16 ± 3
28
32
46 ±5
21
79 ±5
85 ±1
26 ±5
40 ±4
40
53
42 ±5
12
23 ±6
24 ±4
25 ±5
70
84
7±4
5+ 1
Table 3: Average non-destructive E D X RF analysis results obtained for coins belonging to the gold hoards o f Heers, Amby,
Thuin-1, Fraire and Orp-le-Grand. Concentrations are expressed in weight %. n = number o f coins analysed within each
numismatic group.
ANNAROVECI quinarii.24 This deficit was probably compensated in turn by reducing the proportion
of silver in the alloy.
From Table 3 and Figure 4 we can primarily deduce that the composition o f the coins from the
same class remains the same, irrespective of the hoard they belong to. Further, it can be noted that the
Nervian staters are situated more towards the end of the red fine, whereas the composition o f the Eburonean staters is around the green fine, which means they contain additional copper. More details will
be discussed in section 5.
Since the metal compositions obtained on the surface are not representative of the whole coin, per
mission was obtained, for a few selected coins, to either make a cross-section at the edge of the coin or to
remove a very small sample. Table 4 gives the results (average values of at least five different measurements
and standard deviations) for the selected coins. Concentrations are expressed in weight %.
With regard to composition variations within different classes, similar conclusions can be reached
from the core metal, in particular the fact that triskeles staters of the Eburones have a higher copper con
tent than the epsilon staters of the Nervii.This is important for the following discussion about the metal
composition of the Late Iron Age gold coinages of the Fraire/Amby horizon. It is especially interesting
to note that preliminary application of the more accessible EDXRF analysis, despite its being a surface
analysis technique, can make a real contribution to identifying tendencies provided relative values are
used.
4 Scheers, private communication.
222
• Scheers 25-11
@ Scheers 29-la
• Scheers 29-lb
• Scheers 29-lla
• Scheers 29-llb
O Scheers 30-1
■
Scheers 30-V
A
A
Scheers 31-la
Scheers 31-lb
• rainbow stater
Cu
Fig.4.Ternary Au-A g-Cu plot showing the average non-destructive E D X R F analysis results obtained for the coins belonging to
the gold hoards o f Fleers,Amby,Thuin-l, Fraire and Orp-le-Grand.Two different gold debasement types are indicated: a debase
m ent with a 2:1 mixture ofA g and Cu (red line) and an increasing Cu with constant Au content (green line).
In addition to coins from the gold hoards, 17 additional single coins from museums and private col
lections were selected for analysis in order to establish adequate comparison studies of the Late Iron
Age gold coinages of the Fraire/Amby horizon. The individual results of the surface EDXRF analysis
are given in Appendix 6 . Analysis o f the core metal by sampling was only permitted for two coins. The
results are integrated into Table 4.
223
coin
type Scheers
Au
Ag
Cu
H 129, Heers
29-llb, Nervii
53
24
23
H 131, Heers
29-llb, Nervii
55
28
17
H 157, Heers
30-V, Treveri
36
48
16
H 40, Heers
31-la, Eburones
31
30
39
H 42, Heers
31-la, Eburones
33
35
32
H 46, Heers
31-la, Eburones
36
33
31
GRIM 10908, Berg
31-la, Eburones
33±3
34 ±4
33 ±4
A 3, Amby
31-la, Eburones
33 + 1
44 ± 1
23 ±1
A 7, Amby
31-la, Eburones
33 ±3
33 ±7
34 ±5
A 8, Amby
31-la, Eburones
37 ±3
33 ±5
30 ±3
A 17, Amby
31-la, Eburones
38 ±2
28 ± 1
34 ±2
A l 8, Amby
31-la, Eburones
33 ±1
30 ±2
37 ±2
A 20,