Vol 1

Roman Lead Sealings
VOLUME I
MICHAEL CHARLES WILLIAM STILL
SUBMITTED FOR TIlE DEGREE OF PILD.

SEPTEMBER 1995
UNIVERSITY COLLEGE LONDON

INSTITUTE OF ARCHAEOLOGY
(L n")
'3 1.
ABSTRACT
This thesis is based on a catalogue of c. 1800 records, covering over 2000

examples of Roman lead sealings, many previously unpublished. The catalogue is
provided with indices of inscriptions and of anepigraphic designs, and subsidiary indices
of places, military units, private individuals and emperors mentioned on the scalings.
The main part of the thesis commences with a history of the use of lead sealings
outside of the Roman period, which is followed by a new typology (the first since
c.1900) which puts special emphasis on the use of form as a guide to dating.
The next group of chapters examine the evidence for use of the different
categories of scalings, i.e. Imperial, Official, Taxation, Provincial, Civic, Military and
Miscellaneous. This includes evidence from impressions, form, texture of reverse,
association with findspot and any literary references which may help. The next chapter
compares distances travelled by similar scalings and looks at the widespread distribution
of identical scalings of which the origin is unknown.
The first statistical chapter covers imperial sealings. These can be assigned to
certain periods and can thus be subjected to the type of analysis usually reserved for
coins. The second statistical chapter looks at the division of categories of scalings within
each province. The scalings in each category within each province are calculated as
percentages of the provincial total and are then compared with an adjusted percentage
for that category in the whole of the empire.
The final chapter is based on the iconography found in the impressions on the
scalings. This includes the styles of imperial portraits, deities, animals, inanimate objects,
designs which may come from outside of the empire, similar impressions on other items,
epigraphic styles and possible examples of matrices.
2
TABLE OF CONTENTS
Volume I
Title Page 1
Abstract 2
Table of Contents 3
List of Maps, Figures, Tables and Illustrations 10
Acknowledgements 17
Introduction 19

List of findspots to accompany maps 23
1. The Use of Lead Sealings before and after the Roman Period 26
Pre-Roman use of lead sealings 26
Post-Roman use of lead scalings 28
2. Typology of Roman Lead Scalings 35
Type 1: Two-sided with different sized obverse and 36

reverse resulting in flange around flan.
Type 2: Two-sided with various shapes of flan 38
Type 3: Two-sided with squared-off appearance due 40
to having been cut from a lead bar
Type 4: Two-sided consisting of two plates, joined 41
by a strip, clipped together
3

Type 5: One-sided with swelling on blank reverse 42

Type 6: One-sided with flat reverse 45

Type 7: One-sided rectangular mould-formed bars 47

Type 8: One-sided with central nipple on reverse 48

Type 9: One-sided with pedestal-foot projection on reverse 49

Type 10: One-sided consisting of two plates 50

Experimental manufacture of lead sealings by the author 50

Notes 52

3. Imperial Sealings - The Evidence for their Use 54

Evidence from impression 54

Evidence from form and reverse 59

Evidence from association with findspot 61

Evidence from written sources 66

Examples dated by context 67

Conclusions 67

Notes 69

4. Official Scalings - The Evidence for their Use 72




Conclusions 78

Notes 79

5 Taxation Scalings - The Evidence for their Use 80




4


Conclusions 83

Notes 84

6. Provincial Scalings - The Evidence for their Use 85






Additional example 89

Conclusions 89

Notes 90

7. Civic Sealings - The Evidence for their Use 91






Conclusions 96

Notes 97

8. Military Scalings - The Evidence for their Use 98





The absence of evidence 114

5

Conclusions 117

Notes 118

9. Miscellaneous Sealings - The Evidence for their Use 122




Evidence from association with item sealed 129



Conclusions 131

Notes 133

10. Scalings as evidence for the distribution of goods 135

Conclusions 141

Notes 142

II. Imperial Scalings - The Statistical evidence 143

Scalings assigned to emperors by inscription 146

Scalings assigned to emperors by portrait identification 150

Comparison of scalings assigned to emperors by inscription and portrait 153

Conclusions 157

Notes 159

12. Provincial Distribution of Scalings - The Statistical evidence 160

Conclusions 193

13. The Iconography of Roman Lead Scalings 194

Imperial Portraits 194

Religion/Mythology 196
6

Human figures 204

Animals 204

Combinations 205

Inanimate objects 206

Designs possibly from outside of the empire 207

Similar impressions on other items 208

Epigraphic styles 208

Possible examples of matrices 209

Notes 213

14. Conclusions 214

Abbreviations 217

Bibliography 223
Volume II

Title Page to Volume II 237

Introduction to Catalogue 238

Catalogue (followed by fold-out key to Catalogue sections) 241

Introduction to Main Indices 607
Main Indices 609

Main Epigraphic Index 609
Main Index of designs (from both epigraphic and anepigraphic examples) 667
7

Subsidiary indices 715

Index to Categories within the Catalogue 715
Quick Reference Guide to Classification and Provenance of Sealings 715

Index of sites in order of Province 721

Military Units mentioned on Lead Sealings 724

Names of Provinces and their Inhabitants 726

Names of Towns/Cities and their Inhabitants 726
Nomina of individuals and Families (except members of Imperial Family) 727
Cognomina of Individuals (except members of the imperial Family) 730
Members of Imperial Families (where mentioned by name) 734
Members of Imperial Families identified from portraits alone 734
List of previously unpublished examples 738
List of previously published examples for which I have offered 739

new readings, interpretations etc. from personal inspection of
examples/photographs and analogy with known examples.

Gazetteer of sites 730
8

Introduction to Illustrations 768

Illustrations of scalings 769
Supplementary Material (bound in at back of Volume II)
Still, M.C.W., 1993, 'Opening up Imperial lead sealings', Journal of Roman

Archaeology 6, 403-408.
Still, M.C.W., 1994, 'Some Roman lead scalings from Arzus', in Draganov, D. (ed),
Settlement Life in Ancient Thrace - Ilird International Symposium "Cabyle ", Jambol,
389-395.
Still, M.C.W., 1995, 'Parallels for the Roman lead sealing from Smyrna found at Ickham,
Kent', Archaeologia ('antiana cxiv (1994), 347-3 56.
CORRIGENDA
Volume I
Page 58, line 4: delete 'has' at the end of the line and insert it at the end of line 5.
p.88, line 6: for 'manslo' read 'mutalio'.
p.120, note 16: for 'Janet' read 'Jarrett'.
p.203, bottom line: for '[1764]' read '[1765]'.
p.230, top line: for 'Jarret' read 'Jarrett'.
9
LIST OF MAPS, FIGURES, TABLES AND ILLUSTRATIONS
Maps

Map of empire showing findspots 21
(including inset I of Pannonia & inset 2 of Moesia Inferior and Thrace)
Map of Britain showing findspots (inset 3 of map of empire) 22
Figures

Fig. I Typology of Roman Lead Sealings 53
Fig.2 Actual numbers of Imperial scalings found in each province 143

(there are no exact figures for Numidia)
Fig.3 Imperial scalings as a percentage of the total of all scalings in their 144
respective provinces compared with the total of imperial sealings as a
percentage of the whole empire total (including and excluding Gallia Lugdunensis)
Fig.4 Actual numbers of Imperial scalings identified by inscription 146
Fig.5 Imperial scalings per thousand identified by inscription 149
Fig.6 Imperial scalings per thousand identified from portrait alone 152

Fig.7 Imperial scalings per thousand identified from 153
(i) inscription and (ii) portrait alone
10
Fig.8 imperial sealings per thousand identified from (i) inscription and 156
(ii) portrait alone, the latter omitting examples bearing any other
inscription or multiple busts
Fig.9 Percentages of sealings in different categories for whole 161

empire including Gallia Lugdunensis
Fig. 10 Percentages of sealings in different categories for whole 163

empire excluding Gallia Lugdunensis.
Fig. 11 Percentages of scalings in different categories in Africa 166

Proconsularis (unshaded) compared with category total for the whole
empire expressed as a percentage of the whole empire total for those two
categories (shaded).
Fig. 12 Percentages of scalings in different categories in Asia (unshaded) 167

compared with the category total for the whole empire expressed as a
percentage of the whole empire total for those two categories (shaded).
Fig. 13 Percentages of scalings in different categories in Britannia (unshaded) 169

percentage of the whole empire total for those twelve categories (shaded).
Fig. 14 Percentages of scalings in different categories in Dalmatia (unshaded) 171

percentage of the whole empire total for those two categories (shaded)
Fig. 15 Percentages of scalings in different categories in Gallia Belgica 172

(unshaded) compared with the category total for the whole empire expressed
as a percentage of the whole empire total for those five categories (shaded).
II
Fig. 16 Percentages of sealings in different categories in Gallia Lugdunensis 174
(unshaded) compared with the category total for the whole empire expressed as a
percentage of the whole empire total for those six categories (shaded).
Fig. 17 Percentages of scalings in different categories in Germania Superior 176

percentage of the whole empire total for those three categories (shaded).
Fig. 18 Percentages of sealings in different categories in Italia (unshaded) 178

percentage of the whole empire total for those five categories (shaded).
Fig. 19 Percentages of scalings in different categories in Moesia Inferior 180

percentage of the whole empire total for those six categories (shaded).
Fig.20 Percentages of scalings in different categories in Noricum (unshaded) 182

Fig.2 I Percentages of scalings in different categories in Numidia (unshaded) 183

Fig.22 Percentages of sealings in different categories in Pannonia Superior 185

as a percentage of the whole empire total for those four categories (shaded).
12
Fig 23 Percentages of sealings in different categories in Pannonia - unspecified 187
as a percentage of the whole empire total for those four categories (shaded).
Fig.24 Percentages of sealings in different categories in Syria (unshaded) 189

percentage of the whole empire total for those four categories (shaded).
Fig.25 Percentages of sealings in different categories in Thracia (unshaded) 190

percentage of the whole empire total for those four categories (shaded).
Fig 26 Percentages of scalings in different categories from unknown 192

provinces (unshaded) compared with the category total for the whole
empire expressed as a percentage of the whole empire total for those
four categories (shaded).
Tables
Table I. Figures and Records used to calculate 'Imperial scalings per 148
thousand identified by inscription'
Table II. Figures and Records used to calculate 'Imperial scalings per 151
thousand identified from portrait alone'
Table III. Figures and Records used to calculate 'Imperial sealings per 155
thousand identified from (I) inscription and (ii) portrait alone, the latter omitting
examples bearing any other inscription or multiple busts'
13
Table IV. [Percentages of sealings from] Whole empire including sealings 160
from Gallia Lugdunensis (Total 2304 examples)
Table V. [Percentages of scalings from] Whole empire excluding sealings 162

from Gallia Lugdunensis (Total 1262 examples)
Illustrations of sealings
Imperial - Asia (0003-0005); Britannia (0007A-001 8); Gallia Belgica (00035) 769
Imperial - Gallia Lugdunensis (0042-0069); Italia (0111) 770

Imperial - Moesia Inferior (0134-0 178) 771
Imperial - Noricum (0182); Pannonia Superior (0189); Pannonia (0202A-02 12) 772

Imperial - Parinonia (0213-0225) 773
Imperial - Pannonia (0226-0227); Thracia (0230-0241) 774

Imperial - Thracia (0242-0252) 775

Imperial - Thracia (0253-0267) 776
Imperial - Thracia (0268-0270); Unknown Provinces (0271-0273) 777

Imperial - Unknown Provinces (0274-0279) 778
14
Official - Britannia (0284); Moesia Inferior (0287A); Pannonia Superior 779
(0288-0289; Pannonia (0290-0292A)
Provincial - GaHia Belgica (0314), Thracia (0320-0321) 780

Civic - Gallia Belgica (0329); Moesia Inferior (0338-0351), Pannonia (0366);
Syria (0367); Thracia (0368)

Civic - Thracia (0372-0384) 781
Civic - Thracia (0386-0387); Unknown Province (0388) 782

Legionary - (3allia Lugdunensis (0429)
Ala - Britannia (0444)
Cohort - Britannia (0457-0479)
Beneficiarii Consularis - Britannia (0524)

Miscellaneous - Asia (0657-0667) 783
Miscellaneous - Asia (0668-0675); Britannia (0705-0709) 784

Miscellaneous - Britannia (0710-0731) 785
Miscellaneous - Britannia (0739-0779); Germania Inferior (151 1-1512); 786

Moesia Inferior (1583-1647)

Miscellaneous - Pannonia (1685-1699) 787

Miscellaneous - Pannonia (1702-1714) 788

Miscellaneous - Thracia (1721A-1733) 789
IS

Miscellaneous - Thracia (1740-1757) 790

Miscellaneous - Thracia (1758-1769) 791

Miscellaneous - Unknown Provinces (1775-1 792A) 792
Miscellaneous - Unknown Provinces (1794-1809); App 1. 13 - Italia 793
16
ACKNOWLEDGEM ENTS
For allowing me to view their museum collections I would like to thank Dr. G.
Atanasov of Silistra Historical Museum; Dott ssa M. Bertinetti of the Museo Nazionale
Romano; Dr. R. Bland of the Department of Coins and Medals, British Museum; Mr.
B F. Cook of the Department of Greek and Roman Antiquities, British Museum; Sophie
Descamps of the Département des Antiquités Grecques, Etrusques et Romains, Musée
du Louvre; Dr. D. Draganov of Jambol Historical Museum; Dr. F. Grew of the Museum
of London; Dr. J. Hall of the Museum of London; Miss C. Johns of the Department of
Prehistoric and Romano-British Antiquities, British Museum; Mr. A. MacGregor of the
Department of Antiquities, Ashmolean Museum; Endre Téth of the Hungarian National
Museum; Michael Vickers of the Department of Antiquities, Ashmolean Museum.
For their general help and information, often considerable, I would like to thank
Miss D. Boteva; Mr. J. Bradshaw; Dr. J-C. Chenet of the Centre d'histoire et de
civilisation de Byzance, College de France; Dr. Amanda Claridge of the British School at
Rome; Dr. J. Clayton Fant of the University of Akron, Ohio; Dr. A. Detsicas, editor of
Archaeo/ogia Canliana; Mr. D. Dimitrov of Pernik; Dr. M. Feugére, C.N R.S.; Dr. V.
Gerasimova-Tomova of the National Historical Museum, Sofia; Mr. John Goodall; Dr.
M. Henig of the Institute of Archaeology, Oxford; Dr. J. Humphrey, editor of the
Journal of Roman Archaeology; Dr. I. Jordanov of the Shumen Archaeological Institute;
Dr. CE. King of the Heberden Coin Room, Ashmolean Museum; Dr. E.
Oberländer-Tarnoveanu of the National Historical Museum, Bucharest; Dr. A. Poulter,
University of Nottingham; Dr. M. Roxan, Institute of Archaeology, London; Ms. E.
Rucker; Dr. R. Tomlin, Wolfson College Oxford; Dr. S. Torbatov of the Museum of
History, Dobrich; Helen Whitehouse of the Department of Antiquities, Ashmolean
Museum; Prof J.J. Wilkes, Institute of Archaeology, London; Dr. D 3. Woolliscroft of
the Department of Archaeology, University of Manchester.
17
From the above I must single out Dr Dimitar Draganov for special thanks for
inviting me to speak at the symposium in Jambol, for putting his department's collection
of sealings at my disposal, for negotiating the loan of a large private collection to enable
me to study it, and for producing much-needed maps of Bulganan provenances.
Replies concerning an absence of sealings in their museum collections were

received from the following: P.H. Allioux of the Cabinet numismatique de Ia
Bibliothèque Nationale et universitaire de Strasbourg; Hélène Chew of the Musée des
Antiquités Nationales, St-Germain-en-Laye; Dr F. J. Hassel of the ROmisch-
Germanisches Zentralmuseum, Mainz; Jean-Pascal Jospin of the Musée Dauphinois,
Grenoble; Jacques Lasfargues of the Musée de Ia Civiisation Gab-Romaine, Lyon;
JoëlIe Pournot of the Cabinet des Monnaies et Médailles de Marseille; Bernadette
Schnitzler of the Musée Archéologique de Strasbourg; Annie Tarin of the Musée
d' Archeologie Méditerranéenne, Marseille.
Lothar Schwinden of the Rheinisches Landesmuseum Trier informed me that

there are some unpublished sealings in the museum's collection but that Heinz Cuppers is
hoping to publish them.
I would particularly like to thank my supervisor Mark Hassall for his welcome
help and advice on matters both large and small during my period of research.
I would also like to take this opportunity to thank Richard Reece, my secondary
supervisor, for reading drafts of, and offering advice on, the statistical chapters and for
general encouragement.
Finally, I would like to thank my mother for her patience and understanding,
without which this work could not have been completed.
18
INTRODUCTION
Several authors have expressed the need for an up-to-date detailed study and
corpus of Roman lead sealings (Turcan, 1987, 10; Vons, 1980, 44-5; Dembski, 1975,
49). The most recent attempt at developing a typology and dividing sealings into
categories had been at the turn of the century (Rostovtzeff, 1900, 1-21) That important
work still forms the main pillar of lead sealings studies and rightly so. The rise of the
metal detector in recent years, however, has produced large numbers of sealings which
need to be integrated with the old material in the hope of enriching our understanding of
their role in Roman life. I have therefore attempted to produce such a work. This has
required a knowledge of glyptics, numismatics, trade mechanisms, military organisation,
customs dues and tolls, etc.. Doubtless, specialists in these fields will be able to find
many inaccuracies in my discussion of their subjects but it remains a fact that somebody
had to make a start on this daunting task.
The work is centred on a catalogue of Roman lead sealings drawn together from
many sources. It is the first corpus of Roman lead sealings from across the empire ever
to have been formed. The majority of pieces have already been published although I have
been able to correct many readings and interpretations in the light of finds from other
sites. I am also pleased to have been able to include many important examples, mostly
from Bulgaria and Hungary, which have never been published. This catalogue is by no
means complete since there must be many more examples hidden in museum stores,
especially in areas which I was unable to cover. Due to the restrictions of time I
concentrated on writing to museums in those areas which had already produced
published sealings Even if I had been able to cover all museum collections there are
many examples in private collections and, of course, new examples are constantly being
discovered.
Once the catalogue was complete, I was able to use it as the basis for a
discussion of many aspects of sealings which have never been addressed previously,
19
including the use of statistical analysis. This overview will hopelblly draw together the
valuable information which appears in the continual stream of isolated articles covering
museum collections and metal detector finds.
20
LIST OF FINDSPOTS ON MAPS
(not including sites which have only produced examples in appendices)
No. on map Site Pros ince Modern Country
Aldborough Britaiinia Great Britain

2. Ahimir Moesia Inferior Bulgaria
3. Bainbridge Britannia Great Britain
4. Baldock Britannia Great Britain
5. Beirut S na Lebanon
6. Belosem Thracia Bulgaria
7. Benwell Britannia Great Britain
8. Binchester Brilannia Great Britain
9. Brettenham Britannia Great Britain
10. Brough under Stainmore Britannia Great Britain
11. Burgh Castle Britannia Great Britain
12. Buxton Bntannia Great Britain
13. Cabyle Thracia Bulgaria
14. Caerleon l3ritannia Great Britain
15. Carlisle Britannia Great Britain
16. Carthage Africa Proconsularis Tunisia
17. Catterick Britannia Great Britain
18. Chesterhoim Britannia Great Britain
19. Chesters Britannia Great Britain
20. Cirencester Britannia Great Britain
21. Cotchester Bntannia Great Britain
22. Combe Dov n Britannia Great Britain
23. Constantine Numidia Algeria
24. Corbridge Britannia Great Britain

25. Dimitrovgrad Thracia Bulgaria
26. Ehi Gerrnania Superior France
27. Enns-Lorch Noricum Austria
28. Felixstowe Britaiinia Great Britain
29. Frójus GaIlia Narboncnsis France
10. Gardun Dalmatia Croatia
23

No. on map Site Pros ince Modern Country
31. Gigen Moesia Inferior Bulgana

32 Glava Moesia Inferior Bulgaria
33. Ickham Britannia Great Britain
34. llindcnt,i Thracia Bulgaria
35 Istanbul Thracia Turkey
36. Izmir Asia Turkey
37. Izvoarele Moesia Inferior Roinania
38. Kalugcrovo Thracia Bulgaria
39 Katzelsdorf Pannonia Superior Austria
40. Kenchela Numidia Algeria
41. Kirkby Thore Britannia Great Britain
42. Kirminglon J3ritannia Great Britain
43. Kocherinovo Thracia Bulgaria
44. Koinare Thracia Bulgaria
45. Kotel Moesia Inferior Bulgaria
46. Lancaster Britannia Great Britain
47. Lanchester Britannia Great Britain
48. Leicester Britannia Great Britain
49. Leithaprodersdorf Pannonia Superior Austria
50. London Bntannia Great Britain
51. Lyon Gallia Lugdunensis France
52. Mainz Germania Superior Germany
53. Medo o Thracia Bulgaria
54. Mellieha Bay Melita Malta
55. Mohcs Pannonia Superior Hungary
56. Newsicad Britannia Great Britain
57. Nijniegen Germania Inferior Netherlands
58. Old Winteringham Britannia Great Britain
59. Ostia lialia Ital)
60. Pakenham Britannia Great Britain
61. Peakley Hill Britannia Great Britain
62. Pctronell Pannonia Superior Austria
63. Piercebridge Britannia Great Britain
64. Pittinglon Farm Britannia Great Britain
24

No. on map Site Proince Modern Country
65. Ravenglass Britannia Great Britain
66. Ra enna Italia Italy

67. Ravenscar Britannia Great Britain
68. Rhci nzabcrn Germania Superior Germany
69. Richborough Britannia Great Britain
70. Rome Italia Italy

71. Rudchester Britannia Great Britain
72. St.Albans Britannia Great Britain

73. Sewingshields Britannia Great Britain
74. Silchestcr Britannia Great Britain
75. Silistra Moesia Inferior Bulgaria

76. Skikda Numidia Algeria
77. Soli n Dalmalia Croatia
78. South Shields Britannia Great Britain
79. Stanwix Britannia Great Britain
80. Strebersdorf Pannonia Superior Austria

81. Svishtov Moesia Inferior Bulgaria
82. SzOny Pannonia Superior Hungary
83. Thorpe in the Glebe Britannia Great Britain
84. Trier Gallia Belgica Germany

85. Velsen Germania Inferior Netherlands
86. Vetren Thracia Bulgaria
87. Vienne Gallia Narbonensis France
88. Wagna Noricum Austria
89. Wailsend Brilannia Great Britain

90. York Britannia Great Britain
25
Chapter 1
TIlE USE OF LEAD SEALINCS BEFORE AND AFTER TIlE

ROMAN PERIOD
Pre-Roman use of lead sealings
If lead sealings of the Roman period have suffered from limited research in the
past then those of earlier periods have been virtually ignored. It is very difficult to
construct a chronology of these sealings although the earliest would appear to have been
recognised in Bulgaria. One can suspect, although with no evidence whatsoever, that this
practice had actually spread from Anatolia and probably further east.
The site of Cabyle, near modern Jambol, Bulgaria, is said to have produced a lead
sealing dated to the fifth century B.C. (Gerasimova-Tomova, 1982, 142). It is thought to
have originated in Apollonia on the Black Sea and shows an inverted anchor with a crab
and part of the letter 'A'. This design is said to be identical to that on coins of the city
dating to 440-400 B.C. However, in the same article, Gerasimova-Tomova claims that
Garrucci (1862, 408ff., pl.xv) published the oldest known lead sealing, from the late sixth
century B.C., showing Darius I. I have looked at this reference and the piece of lead is a
lessera of some sort which Garrucci never suggested was a sealing.
The same site has also produced two lead sealings believed to date to the fourth
century B.C. (Gerasimova-Tomova, 1982, 143-144). The sealings show a man on
horseback with a wreath below the horse. Apparently, this combination is only found on
the coinage of the Thracian king Seuthes III, the association being confirmed by the
similarity in style and proportions of design on both scalings and coins. These scalings
may show that goods were being sent from Seuthopolis to Cabyle in the fourth century
BC.
26
Several of the scalings included in my Catalogue may in fact be pre-Roman.
Indeed, many could be Hellenistic but it is virtually impossible to separate them from the
Roman examples.
There is a distinctive group of scalings from Sicily, bearing designs which appear
to belong to the Hellenistic and Byzantine periods.' These consist of two roundels linked
by a folded strip (for a section of a Roman example of this type see Fig.1. no.4).
As for parallels, there are many examples of Hellenistic clay scalings (Selinus,
Sicily: Salinas, 1883, 288ff., pls.V1I-XV; Neo Paphos, Cyprus: Nicolaou, 1971, 51-53;
Uruk, Babylonia: Rostovtzeff, 1932, 1ff, pls.1-XI; Richter, 1968, 165, nos.658-661;
Seleucia: McDowell, 1935; Elephantine: Rubensohn, 1907, 7ff., pIll; Priene: Wiegand
& Schrader, 1904, 465, no.23 5) but apparently only two collections (see below) which
enter the Roman period. All of these clay scalings are believed to have once been
attached to rolled papyrus documents held in archives (Spier, 1992, 167).
The best known collection which includes examples from the Roman period is the
group of several thousand scalings found in the Nomophylakion at Cyrene in Libya
(Maddoli, 1965, 39-145). Although the preserving fire here occurred cAD 117-118
during the Jewish Revolt, many of the impressions are said to date from the late
Hellenistic period of the first century BC. However, perhaps we should bear in mind that
many of the gems could still have been in use in the early second century AD and that
therefore the scalings themselves may have been impressed much closer to the time of
the fire. Thus we may also be seeing the continued use of this system well into the
Roman period, perhaps not surprising when one considers Cyrenaica's retention of
Hellenic culture.
The other collection which partially dates to the Roman period consists of
thousands of scalings from an unknown site (probably near Diyarbakir in Turkey,
perhaps ancient Douche) and seems to date from the first century B C. to the early
27
second century A D. (Spier, bc. i1.). The collection has been reconstructed theoretically
from various museum collections and sale pieces 2
Post-Roman use of lead sealings
From the eighth to tenth centuries A D we find Islamic lead scalings inscribed in
Kufic. These are often square, although sometimes circular, and have holes in the
reverses through which a wire was passed.
Possibly the greatest use of lead scalings, however, was made during the
Byzantine period. These were in use right up to the end of the empire. They are found in
much larger numbers than those of the Roman period and served several purposes. One
of these was to secure official letters. We are told by Pseudo-Kodinos that the emperor
would use lead scalings when corresponding with I)espolai, Patriarchs and other
archontes of high rank (Verneaux, 1966, 175-6):
irpoç & öcctotaç, iratptapxç icat touç otitouç apxovtac tcov evtt.totcxov
[c$paytctJ öta toXt43ötvqg I3ouXXig
He would use scalings of gold or silver when dealing with people of higher rank.
The majority of scalings belonging to officials, both civilian and military, and clergy,
were found in Istanbul since this was the administrative centre to where reports etc.
would be sent (Zacos & Veglery, 1972, vii). These scalings were always double-sided
and were struck from cold blanks. The blanks had been pre-formed in moulds which
were shaped to give a raised semi-cylindrical housing across the diameter for the
thread-hole. The hole was kept clear using wire which was removed once the lead had
solidified. The scalings were impressed by placing the blank between the
retrograde-engraved jaws of a pair of iron pincers, the boulloterion, after threading the
cord attached to the document through the hole in the sealing. The head of the
boulloterion was then struck with a hammer and in this way the sealing was stamped with
28
a two-sided design while the thread was held in the crushed channel (Zacos & Veglery,
op. cii., xi).
Another type of Byzantine lead sealing was that belonging to kornrnerkiarioi

(Zacos & Veglery, op. cii., 131 &135). These officials were in charge of the imperial
warehouses and the collection of associated taxes and customs duty. In the provinces
they supervised the supply and distribution of goods, while in the capital they dispatched
them to the sales depots. They were directly attached to the emperor and were therefore
allowed to use his portrait on their sealings. These sealings date from the second half of
the sixth century AD. to the first half of the ninth century A.D. Some belonging to the
sixth to early eighth centuries are unusual for Byzantine sealings in that they have a blank
reverse, often showing the impression of a coarse material (Morrison, 1987, 5 & 7,
fig.3). They must therefore have been stamped with a single matrix once the
merchandise, thought to be silk, had been wrapped up.
Despite its high cost, Byzantine silk was highly sought after for its quality. In this
way the lead sealings, which were really for customs purposes, came to be seen as a
guarantee of quality. Goods of this sort reached the West through the trade fairs held at
Pavia where they were sold by traders from the Byzantine-dominated Venice, Amalfi and
Salerno. It appears to have been in the ninth to eleventh centuries that lead scalings came
to be regarded as trade marks rather than customs labels among the cloth merchants of
Western Europe (Endrei & Egan, 1982, 47-75). It happened at this time due to the
ubiquitous nature of Byzantine goods in the area. By the twelfth or thirteenth centuries
the scalings were even being imitated.
The use of lead sealings then radiated out from the two main textile exporting
centres, Flanders and northern italy. From Flanders it spread to neighbouring centres of
woollen cloth production in north France, the Netherlands and western Germany. By the
early fourteenth century the practice had reached the south of France while the first
mention of it in England appears to be in 1380, although it had almost certainly been
29
current for some time by then (Sharpe, 1907, 145-6). It was not until the fifteenth
century that the sealing in lead of woollen cloth began in southern Germany and
Czechoslovakia and only in the sixteenth century that it was adopted in Poland, Hungary
and Scandinavia.
Lead sealing for the cotton industry started in north Italy and then spread
northwards in the fourteenth century. The linen industry took up the idea in the fifteenth
century.
In England, cloth was examined by alnagers who would attach one sealing to
each cloth to show that it had been checked and passed and that the subsidy revenue had
been collected.
During the sixteenth century a number of examinations and sealings by various

authorities were allowed at different stages of production such as after weaving, after
frilling and after dyeing. The manufacturer would apply his sealing stating the length and
then the alnage sealing would be added. This meant that, in theory, there could be over
half a dozen scalings on each cloth. However, this seems to have been extremely rare
(Endrei & Egan, 1982, 56).
There are three usual types of British cloth seals, single-, two- or four-disc
(Endrei & Egan, op. cii., 59). Single-disc examples are one piece scalings with a hole
running across the diameter. They were probably connected with the dyeing process
since designs on some of them bear similarities to the crest of the arms of the London
Dyers' Company. Two-disc sealings are the most common of this period. They consist of
two plates joined together by a thin strip One of the plates bears a projection which was
pushed through the hole in the other plate, prior to it being struck between two dies.
As far as alnage scalings are concerned, the change from two- to four-disc
sealings occurred c.l610 or 1611, although the two-disc variety continued in use for
30
other purposes until the nineteenth century. Four-disc scalings consist of four plates in a
line connected by thin strips. In use this line of plates was folded at the middle strip so
that the projection in the plate at one end was passed through the hole in the plate at the
other (Egan, 1994, viii). In its finished state, after striking, the sealing would have the
appearance of two scalings i.e. a figure of eight configuration This type was the standard
alnage seal from the reign of James I. Many of the post-Restoration scalings bear the
monarch's portrait and the similarity to representations on coins increased with time,
including the adoption of a seated Britannia. In 1982 Geoff Egan knew of only two
extant matrices used for cloth sealing: one for bay cloth from Bocking in Essex dated to
the sixteenth! seventeenth century and another from Suffolk said to date from the middle
of the fourteenth century (Endrei & Egan, op. cii, 67 and notes 80 & 209).
Alnage ceased in 1724 but in 1765 'searchers' were given the right to seal
examined cloths. In 1821 it was recommended that the laws governing the practice were
abandoned since they were completely ineffective but it was 1889 before they were
repealed. However, the sealing of cloth had effectively finished long before this (Endrei
& Egan, 01). cii., 58).
Starting with the first quarter of the eighteenth century, sealings were used more
and more as company labels bearing control marks and consignment numbers. One of the
major users of this type of sealing (both two- and four-disc) was the United East India
Company (Egan, 1990, 87-89).
The lead scalings of most other countries are less well-researched but those of
Portugal may hold some interest for us. Dc Mello mentions two types, mercantile and
official (de MeIlo, 1979, 211-23). The mercantile bear only initials but the official show
the Royal coat-of-arms and the badge of the particular fiscal department involved He
refers to several examples bearing the name of the ("asa dos ('lisco which was under the
control of the Lisbon custom house and took its name from the five per cent tax charged
on merchandise entering by land or sea.
31
A similar usage to cloth sealing was that of bale sealing. These lead seals were
attached to packed bales, not individual cloths, by the packing companies themselves.
This origin is shown on the actual sealings which usually state the name of the company,
its location and the fact that they are packers (Endrei & Egan, 1982, 60 and note 154).
Some examples of these were found in Minnesota and investigation into the
London-based company names dated them to the early nineteenth century (Birk, 1975,
73-84).
The ninteteenth and, presumably, the early twentieth century saw the use of lead
for the sealing of sacks. These sealings appear to have been pre-moulded as two plates
joined together by lead at the top and bottom (i e. if we take one plate as the face of a
clock, from c. 11.00 - 1.00 and again from c.5.0O - 7.00). There is another, much thinner,
strip of lead joining the two plates at c.9.00. Many of these sealings are found by metal
detector users on farmland but most bear initials or indecipherable company names.
However, the present author has two examples in his collection which seem to shed
some light on their use. The first of these, bought in an antiques shop in Rochester, Kent
and said to have been found in nearby Frindsbury, has the legend
'ANGLO-CONTINENTAL GUA[no] WORKS LTD' while the other, presented to me
by the finder, Mr.L.Smith and found near Dartford, has a mostly illegible impression
from which can be extracted the words '[Pe]RUVIAN' and 'GUANO'. Therefore it can
be stated that one certain use of these sealings was to seal sacks of guano fertiliser. This
also explains why they are found discarded on farmland. However, the author has also
seen (on a coin dealer's stall in Oxford) an example bearing the name of 'SUTTONS
SEEDS' which may suggest bulk supply of seed to farmers but also points out that even
these lead sealings were not associated with a single product. The author also has in his
possession several sealings bearing Cyrillic legends. The purpose of these is unknown but
it reminds us that sealings of this period were not just a British phenomenon.
Lead sealings are still in use today, in widely differing roles. They are used by
customs officers in many countries, usually to prevent additional items being added to
32
previously checked baggage/containers. Turkish officials at the Syrian border seal up
crates of archaeological finds travelling by land from Syria so that their colleagues at the
European end of the country can ensure that no Turkish artefacts are being added and
smuggled out.
Another modern use is that seen in Greece, where replica ancient vases have a
loop of wire passed through the handle and then sealed with a pre-moulded sealing of a
type very similar in form to the nineteenth/twentieth century examples mentioned above,
but smaller. This practice does not seem to be restricted to one company or area of
Greece.
It may seem odd to go so far beyond the chronological limits of our main area of
study. However, I feel that this is valid since any information as to the application of lead
sealings could be of some help in understanding the varied uses to which they were put in
the Roman period The possible similarities will become evident in later chapters.
33
Notes
1. See Type no.4 and note 6 in chapter 2, entitled 'Typology of Lead Scalings'.
2. Many of these clay scalings are identical. They have appeared in several publications,
especially Maaskant-Klcibrink, 1971, 23-63.
34
Chapter 2
TYPOLOGY OF LEAD SEALINGS
It would be very convenient if all lead sealings could be divided up into set
categories based on their overall shape. Rostovtzeff produced the first of these
typologies (1900, 8-9) although, strangely considering his usual influence, it would
appear that virtually no other authors have adopted this. Certainly, Culic devised his
own for the publication of his largest collection (1975, 237). On the basis of my research,
I would like to propose another typology which is outlined below and illustrated in fig.!.
These illustrations are idealised and in most cases do not represent any particular sealing.
The position of the string holes is also generalised and should not be taken as correct for
all examples of that type.
The shape of a sealing is closely connected with the method of use since it is
purely fbnctional and not part of the design of the impression. Since use is dealt with
elsewhere under separate category headings, however, we shall attempt to look at the
main shapes found in order to see whether they can provide any dating evidence. This
could be used in the future to assess the likely age of similarly shaped sealings.
Unfortunately, many of the sources used do not provide the necessary information but
the following details have been assembled from Catalogue entries taken either from
sources which do, or from personal observation. If it can sometimes be difficult to allot a
sealing to a particular type even with first-hand knowledge, then it can often be virtually
impossible when only a written description is available Therefore it has only been
possible to use certain examples from the Catalogue for this particular survey.
35
Two-sided with different sized obverse and reverse resulting in flange
around flan.
Uses: Imperial, Provincial; Military; Miscellaneous.
Examples: Imperial - [0009], [0271]
Provincial: [03121
Military - virtually all scalings found in Britain (at least) of legions
[0390-0423], alae [0444-0453], cohorts [0454-0523], beneficiarii
[0524-0526], equites singulares [0527-0528] and pedites
singulares [0529-05 3 1].
Miscellaneous - [0678-0679]; [0694]; [0727]; [0731]; [0735 -
0736].
Date Range: Late first century - Fourth century.
These scalings were made using two separate (or possibly hinged) matrices. The
matrix for the obverse must have taken the form of a container with sides. The molten
lead was poured into this and then the matrix for the reverse was pushed into the
container, thus forcing out some of the molten lead around the edges. This resulted in a
large reverse with irregular edges and a smaller obverse with very regular, shaped edges.
As I suggested above, these two matrices could have formed something resembling a
boulloterion (a tool formed from two dies hinged together like tongs and designed to be
hammered), rather than being separate. However, with military examples it is usually the
container-matrix which bears the title of the unit and the basic stamp matrix which bears
the name of the man in charge. This could suggest that the two were not joined because
the more skilftilly made container-matrix was designed to be used by various officers,
either in the long term due to promotion or in the short term due to different shifts.
Another reason for assuming that the two halves were not joined is that some people
appear to have chosen to use their ring intaglios in place of the separate stamp.
Obviously, their engraved stone was just as much a personal signature as any stamp
bearing their name or initials. One possible reason for the fact that some scalings bear the
impressions of stamps and others of gems could be that the officer would use his gem
36
whereas his men would use stamps bearing the officer's name or initials. However, and 1
prefer this idea, it could be that only centurions and decurions were allowed to seal
packages but that many owned intaglios of an unsuitable material for impressing in
molten lead and therefore had to have made special stamp matrices in order to conduct
this sort of work. It has been said that a stone such as red jasper could stand such heat
(King, 1872, 395).'
It should be noted that the sealings which do not follow this 'rule' of having the
unit name in the container-matrix are those which use a general purpose reverse which
does not refer to individuals e g. exp(edi'it) on examples from !egio IlAugusta [0391 (7
examples); 7392 (4 examples); 70400 - 70401) and legio VI [0403] and ex(pedivit) on
examples from the benejiciarii consulares [0524 - 0526]. This probably also applies to
the legio I! Augusta examples with FIT on the reverse, although I have not been able to
check this In these cases the manufacturers of the dies obviously never envisaged the
need for close identification of the officers responsible so an easily interchangeable
reverse was not required.
One of the Imperial examples of this type 2 [0271] is dated to the joint reign of
Septimius Severus and Caracalla (AD 198-209). The Provincial example is dated to the
fourth century, by its apparent reference to the province of Maxima Caesariensis, as well
as by the Chi-Rho. There is one legionary example [04171 from a dated context. It was
found in a Roman drainage ditch with mid-second century material. I have limited the list
of sealings of cohorts from dated contexts to just those which definitely conform to the
shape in which we are interested Two [0503 and 0504] were found with mid-second to
early third century material, another [05061 was from a third/fourth century level while
another [0507] was from a fourth century context. One of the examples belonging to the
equites singulares [0527] is dated by context to c ADI6O-c.200. Two of the
Miscellaneous examples [0678-0679] were found in a pit dating to the late first or early
second century whereas [0727] is believed to be from a context dated c.ADI6O-c 200
and [0736] may be mid-second to early third century (Clay, 1980, 318).
37
The range of categories in which this style of sealing was used is heavily biased
towards the official side, particularly the Military. This is clearly in keeping with the
amount of effort and expenditure needed to make the specialised two-part mould
necessary for the production of this type. It is difficult to see anything but a
government-backed body being able to marshal the resources of die-engravers and
bronze- or iron-workers on such a project. Apparently, none of these moulds has ever
been found (the assumption that they have never been found, rather than just not
recognised, is even safer if the two halves were originally hinged together). They were
probably melted down for scrap or, due to their official nature, deliberately destroyed in
the same way as coin dies or Byzantine boulloteria (Oikonomides, 1985, 3).
However, the irregular outlines to many of the container-matrices and the untidy
inscriptions on both these and the top dies suggest that we may not be looking at metal
equipment at all but perhaps at ceramics. The problem here is that these would surely
have been found, even if deliberately broken, on account of the durability of pottery
(unless of course they were crushed beyond recogni6on for security reasons).
2. Two-sided with various shapes of flan.

Uses: Imperial; Official; Provincial; Civic; Miscellaneous.
Examples: Imperial - [0091]; [0208]; [0211]; [0272]; [0273]; [0275]; [0278]
Official -[0290- 0291].
Provincial - [0314-0315].
Civic - [0369].
Miscellaneous - Numerous records (see Catalogue) but few in
terms of percentage of total Miscellaneous category.
Date Range: Late second century - Fourth century
This is the most common type of double-sided sealing, if only because it covers
so many possible methods of manufacture. Scalings of this type may have been made in
38
several different ways: (i)Pre-moulded blanks impressed with a boulloterion struck with a
hammer in the Byzantine style; (ii)pre-moulded blanks impressed using a fixed lower die
and a hand-held upper die, just like coins; (iii)molten lead poured into something like a
boulloterion which was then just squeezed in order to make the impression; (iv)molten
lead poured on to a fixed lower die and then gently impressed using the upper die. It is
quite possible that scalings of this type were made using any or all of the above methods.
However, it is virtually impossible to divide them up into more detailed types since it is
difficult to separate scalings made from pre-moulded blanks which have been formed
badly or which have been squashed out of shape by being impressed from those scalings
which were made and impressed simultaneously from molten lead and which therefore
also have irregular edges.
Moulds for blanks of Byzantine scalings are known (Morrisson, 1987, 3-4) but,
as yet, none have been found for those of our period. The blanks produced would
presumably not have needed any further heating for their final application. This is the
way in which Byzantine scalings were produced using boulloteria. Such tools may have
been used as part of this process in our period and there is a published example of such a
tool which bears an impression naming Constans as Caesar (Oikonomides, 1987c,
105-15). Some Byzantinists, however, believe that this particular tool is merely a mould
for forging coins. 4 The other way in which some of these sealings may have been used
would have involved the use of molten lead. There would have been some difficulty in
simultaneously positioning the cord or wire which had to pass through the scalings, but
the problem would not have been insuperable.
Assuming that the identification of the various emperors is correct, Record

[0091] can probably be dated to AD 193, [0208] is 'Late Roman", [02111 dates to
AD238-244, [0272] to AD 198-209 and [0273] to the fourth century. Record [0275] may
be early third century and [02781 could be mid-third century. The Official examples
[0290] and [0291] date to the fourth century, possibly even c.AD375. The Provincial
examples are also dated to the fourth century by the Chi-Rho which they both bear.
39
Here we see the Imperial category with a sizeable percentage of examples when
compared to the numerically superior (but smaller in terms of percentage) group within
the Miscellaneous category, Again, the resources needed for the production of a double
die (whether hinged or not) favour official use. However, the less complex system (i e.
just two basic dies rather than one being mould-shaped as in no 1) could explain some
apparently private usage within the Miscellaneous category.
3. Two-sided with squared-off appearance due to having been cut from a lead
bar.
Uses: Imperial; Miscellaneous.
Examples: Imperial - [0011-00141 squared, cut-off chunks and [0197]
squared, ?cut-off.
Miscellaneous - [1700].
Date Range: Last half of fourth century (with possible example from 3rd
century)
These sealings appear to have been chopped off from a flat bar of lead, had
thread holes made through them and then been impressed, either with or without
boulloteria. Hassall suggests that [00 11-0014] were struck cold (Brilannia x (1979),
350). This could be correct since the tool used for [0011-0014] may well have been a
boulloterion because on all four examples the obverse and reverse impressions are
positioned the same way up, exactly in relation to each other. This appears to be true for
[0197] as well. The less likely explanation for this positioning is that the sealer was being
very careful in locating his upper and lower dies. Hassall also suggests that the thread
holes were either drilled or punched through the metal. This seems rather unlikely when
the probable amount of these sealings used is taken into account Is it possible that the
bars of lead were specially made with a pre-formed hole running through their length?
Inspection of Hassall's drawings, with the position of the thread holes marked, does not
40
render this impossible although some of the channels are not quite parallel to the uncut
edges of the bar
Records [0011-0014] can be dated to AD360-363 and, assuming a correct

identification of the emperors, [0197] to AD367-375. Record [1700] is undoubtedly
official in some way but is rather difficult to date The style of engraving suggests that it
is considerably earlier, perhaps even early 3rd century. However, its place in this
category is only based upon the evidence of a photograph and so it may not belong here.
It is possible that this method was not used more often since it was realised that it
was more efficient to mould several individual blanks connected by thin sprues than to
mould thick bars which then had to be chopped up using some effort. Again, this amount
of planning and organisation would be in keeping with the bureaucracy of the time.
4. Two-sided consisting of two pfates, joined by a strip, clipped together.

Uses: Cohorts.
Examples: [0510] and two uninscribed examples mentioned under this entry.
Date Range: ThIrd century.
Roman scalings of this form are extremely rare. Presumably the tool which
crushed the two plates together also applied the inscription.
The inscription refers to Cohors V Gaiorum who are attested at South Shields
(the find spot of the sealing) in the third century
There are collections of similar examples, all from Sicily, which belong to this
type but which appear to be either Hellenistic or Byzantine. 6 It would not be
inconceivable for some of those examples to date to the Roman period but I have not
included them in the Catalogue. Otherwise, there are no other Roman examples of this
41
type of sealing known to me The style does appear later as a cloth seal, the earliest in
the Netherlands in the late thirteenth century and in England from the late fourteenth
century (Egan, 1994, 1-2). There are also examples from the seventeenth, eighteenth and
nineteenth century (Egan, 1994,95, 99) and perhaps even more recently.
5. One-sided with swelling on blank reverse (sometimes rising from

surrounding flan).
Uses: Imperial; Taxation; Provincial; Civic; Cohorts; Miscellaneous.
Examples: This is one of the most common shapes of sealing. There are
numerous examples in most categories. Here we shall enumerate
the examples with impressions of fabnc or wood on their reverses.
Fabric impressions
Imperial - [0047].
Civic - [0368]?; [0380]?.
Miscellaneous - [06661; [0864]; [1097]; [1411]; [1494]; [1647].
Wood impressions
Imperial - [0018]; [0022]; [0053-0054]; [0058]; [00671; [0153].
Taxation - [0301].
Cohorts - [0521].
Miscellaneous - [0753], [0835B]; [0837-0838]; [0842], [0869];
[0881]; [0960]; [1025]; [1051]; [1058-1059]; [1068]; [1076];
[1084]; [1100], [11331; [1164]; [1179]; [1248]; [1273]; [1321];
[1341]; [1369]; [1373], [1395]; [1396]; [1418-1419]; [1426];
[1428]; [1451]; [1478]; [1479]; [1508]; [1548]?
Date Range: Throughout Roman period.
42
The impressions on many of these scalings show the bezels of the rings or dies
used to make them. It is virtually certain that rings, even with gemstones, were
sometimes used. One sealing is said to show clearly the impression of alternating pearls
and hearts which formed the decoration on the ring bezel [12071 (Dissard, 1905, 47-8).
The swelling on the reverse of these scalings can take many forms, including
rounded, pointed, steep or shallow, since it is usually irregular. Quite often the slope of
the swelling does not start from the actual edge of the flan but from a few millimetres
inside. Sometimes the swelling can start from the very edge of the flan and be strongly
dome-shaped. Sealings of this type must surely have been formed using molten lead at
their actual moment of use, rather than being formed as blanks and then used later. We
can be virtually certain of this since the stamping of a single matrix on one side of a cold,
pre-shaped lump of lead would either leave little impression if the surface behind was
soft, or would probably damage the blank reverse if the surface behind was firm. As
neither of these has been observed we can probably assume that a small amount of
molten lead was dripped on to the 'string' to be sealed and was then impressed with a
matrix. This is reinforced by the depth of impression and rounded borders of some
scalings.
Another factor in assuming that the lead was applied in a molten state is that
some examples, albeit a minority, bear the imprint of either the weave of a textile or the
grain of wood on their reverse. The only examples bearing the impression of fabric which
I have personally examined are as follows: [0666] and [1647], with less certain examples
[0368] and [0380]. It is only on [1647], [03681 and [0380] that the fabric imprint
coincides with a large reverse swelling. It will be noted that the last two of these were
only possible examples but, fortunately, the remaining example, [1647], bears the clearest
of all fabric impressions The temperature at which the lead would become molten would
not have been sufficient to damage or burn either the 'string' or any fabric covering the
packages The reason that more scalings do not bear fabric impressions on their reverse
swellings may either be through chance due to their having been applied at a temperature
43
which was not conducive to the reverse imprint being formed, or simply because the
fabric was not present. However, we are still left to explain the enigmatic shape of the
swelling on the reverse of these sealings.
It would seem that whatever backing the molten lead was applied to was pliable
enough to give way and to allow the lead to form its own arbitrary shape. This is rather
difficult to understand since, as I have already mentioned, the lead would not have
passed through any fabric. It seems that in some cases we should consider a thin fabric
covering behind which is a soft commodity of some description which gives way as the
lead is poured on and stamped.
As for sealings bearing the accidental impression of wood grain, one unusual
example is [0022] which has a swelling at one end of a flat reverse. The flat part of the
reverse bears approximately six parallel lines, c. 1.5-2 mm apart. One of these lines
continues past the swelling. This may have been caused by the grain of wood but we are
left to wonder how the swelling was allowed to form. One possible explanation could be
that it flowed into a hole in the wood. 1 This is unlikely to have been a knot hole since the
line which passes close to the swelling does not deviate around it in the way that the
grain would. This phenomenon of the probable impression of wood grain on only the fiat
part of a reverse with a swelling can also be seen on [0018] and [0753]. The published
sources which give other examples said to bear the impression of wood grain on their
reverses do not include any information on the exact position of the impression in
relation to the swelling.
Jgnoring the presence of accidental reverse impressions, one way to make

sealings with this form of reverse is to drip the lead into sand or fine earth. However, this
gives little support for achieving a good impression with the matrix, and seems rather
strange if sealing packages was the aim. Perhaps we should consider drops of molten
lead being poured on to something that would be attacked and eaten away by the heat of
the lead, e.g. wax or resin.
44
The Imperial example with fabric impression may represent a young Caracalla
whereas those with wood grain refer to Marcus Aurelius [0050], Septimius Severus,
Caracalla and Geta [0053-0054] and [0058], and Caracalla alone [0049]
This method of manufacture seems to have been popular throughout our period,
presumably on account of its simplicity and limited recourse to specialist tools
6. One-sided with flat reverse.

Uses: Imperial; Provincial; Civic; Cohorts; Miscellaneous.
Examples: Imperial - [0167]; [018 1-0182]; [0185]; [0190]; [0196]; [0230];
[0279].
Provincial - [0320 - 0321].
Civic - [0324]; [0338].
Cohorts- [0521]; [0523].
Miscellaneous - [0799]; [0823]; [09401; [1070]; [1080], [1085];
[1134]; [1177]; [1302]; [1440]; [1587]; [1589]; [1591 - 1592];
[1599]; [1603]; [1608]; [1615]; [1631]; [16931; [1707]; [1742];
[17681; [1782-1783]; [1794]; [1802]; [1809].
Fabric impressions
Imperial - [0279].
Provincial - [0321].
Miscellaneous - [1608].
Wood impressions
Civic - [0324].
Cohorts - [0521]; [0523].
Miscellaneous - [0940]; [1070]; [108oJ; [1085], [1134]; [1177];
[1302], [1440], [1802].
Date Range: Early first century - fifth century.
45
These scalings appear to have been formed and impressed simultaneously at their
moment of application, and not prepared as blanks to be used later. It is unclear exactly
how they differ from type no. 5. It may be that these examples were applied and
impressed using the same technique and that the only difference is that the molten lead
was dripped on to a harder surface which did not allow a swelling to form. In fact there
are examples of the same impression or similar ones referring to the same individual,
some of which have a swollen reverse and others of which have a flat reverse. This may
just indicate that these merchants were sealing different commodities or it could be that it
is evidence that scalings of types no.5 and no.6 only differ in form due to the vagaries of
manual application. As with type no.5 we find that some examples bear fabric imprints
and wood grain impressions on their reverses.
The earliest imperial example of this type of sealing is probably [0230] which
seems to bear a Julio-Claudian portrait. Three other examples, [0185], [0190] and
[0196], appear to show Caracalla. Record [0279] shows a mid-third century emperor,
perhaps Gordian Ill or Philip II while [0182] is thought to represent either Aurelian or
Probus. Finally, [0181] bears a portrait of either the third or fourth century. One of the
Provincial examples, [0321], also provides us with a date since it names a province which
only existed between AD314/5 and 325. Two Miscellaneous examples, [16031 and
[1608], bear monograms which would appear to date from the fifth century.
As for type no.5, this type of sealing was presumably popular for such a long time
on account of its simplicity of manufacture, even if the exact method is unknown to us at
present.
46
7. One-sided rectangular mould-formed bars.
Uses: Imperial; Civic; Miscellaneous.
Examples: Imperial -[0111]; [0192-0193].
Civic - [0367]
Miscellaneous - [1778]; [1785]
Date Range: Third - Fourth century.
The regular shape of these sealings and their flat reverses suggest that they were
formed in some sort of mould. They may have been used as blanks which were
cold-struck with the matrix or they may have been stamped while hot with a die which
may have been hinged to the mould or may have been held separately. Some examples
were formed in bulk inside a two-halved mould, for example [0367] which bears a seam
mark around the edges and the remains of a cut sprue at one end. This also appears to be
true for [0192]. Record [0111] does not appear to have any string holes but seems to
have been attached by a nail which was driven in from the obverse.
Only the three Imperial examples can be dated. They all bear the legend 'DN' in
one form or another and so can probably be dated to the third - fourth century.
Depending on the method used to make the blanks, this technique could be quite
simple, given a small amount of preparation. The mould could have easily been made
from pottery and so the only specialist equipment necessary was the matrix, which all
sealing methods need. Therefore we should not be surprised at any private use of this
technique.
47
8. One-sided with central nipple on reverse.
Uses Imperial; Official; Miscellaneous
Examples: Imperial - [0243]; [0260]; [0267].
Official - [0287A]
Miscellaneous - [1614], [1731]; [1745]?.
Date Range: Third century - Fourth century.
The presence of the central nipple would suggest that the reverses of all of these
examples were formed in a mould of some description. This is definitely true of [0243]
which has an otherwise smooth, shallow reverse with almost vertical sides, [1731] which
has a large dome-shaped swelling surmounted by the central nipple and probably also of
[1614] which has vertical edges.
The examples which can be dated seem to belong to the third - fourth century,
e.g. [0260] which shows a third century imperial family, [0267] which probably shows a
diademed fourth century emperor, [0287A] with its design of four soldiers with a
standard which may be connected with representations on fourth century coinage and
finally a possible entry in this type, [1745], which shows the fourth century (if not later)
image of the Good Shepherd.
At first sight it is surprising that there is so much Miscellaneous use of this

mould-manufactured sealing, since we have noticed that Government backing seems to
be important in the use of a mould. However, when we look more closely at our
Miscellaneous examples we can see that [1614] is an unidentified bust which could be
imperial and the Christian symbolism of [1745] could also indicate some Official use.
This leaves us with just [1731] which appears to bear a nomen preceded by an
abbreviated praenomen. This may be the sealing of a merchant as I have assumed
previously (Still, 1994, 391), being the exception that proves the rule, or it could possibly
be the name of an official, thus explaining the form, although no rank or position is
48
given. It should be pointed out that another sealing bearing the same impression [1 596A
does not appear to have the central nipple.
9. One-sided with pedestal-foot projection on reverse.

Uses: Legionary
Examples: [0439-04411.
Date Range: Early second century -?
These are the only examples of this form which are securely dated to the Roman
period. Gerasimova-Tomova believes that the projections were originally wedge-shaped
and were then pushed flat when they were used (Gerasimova-Tomova, 1992, 71).
However, she does not explain why they did not just bend to one side but instead splayed
out forming another flat plate. Her theory could work if the projection had been pushed
through a hole, perhaps in fabric, and had then been hammered flat, like a rivet. This
seems to be what Culicà had in mind (1971, 193-7). However, perhaps the lead was
poured into something while molten, which allowed the foot to form, and was then
stamped with the legionary matrix on the obverse. The 'rivet' through fabric
interpretation is preferable since tearing of the fabric could explain how we are left with
sealings with undamaged pedestal feet.
Culic dates these sealings to no earlier than the beginning of the second century
AD on account of the (non-) movement of legio XI Claudia. However, he does attempt
to narrow this down by suggesting that they might belong to the reign of Septimius
Severus.
There are three other lead sealings with projections on the reverse but these
projections seem to be much thicker and more solid than our examples here. These other
scalings, all from Sicily, are of uncertain date, although one inscription is in Latin.R
49
It is strange that we only have sealings of this form produced by just one legion.
It suggests that methods of sealing were organised on an ad hoc basis by the local
commander. However, we should not forget that this strange type of sealing may have
been used since it was necessary for whatever was being sealed
10. One-sided consisting of two plates

Uses: Miscellaneous.
Examples: [1511-1512].
Date Range: AD 15-30.
These are apparently the only examples of this form. They are extremely small
and Vons (1980, 44-53) believes that they were formed from two plates of lead between
which the string was sandwiched The plates were then crushed together and one side
impressed using 'tongs', by which he presumably means some sort of boulloterion. He
seems to suggest that this process may have taken place with or without the use of heat.
These examples are closely dated by their presence in a fort beyond the German
frontier which was only occupied for approximately fifteen years. They are some of the
earliest examples we have, although it is of interest that they differ so much from the
usual size and style.
Experimental manufacture of lead sealings by the author
Following on from the methods of manufacture in the Roman period, I shall detail
here my limited attempts to create sealings. Apparently, the only analysis ever carried out
of the lead used for sealings was published by Richmond (1936, 122-3) Two illegible
scalings from Brough under Stainmore were found to contain (i) 72 9% lead and 27.1%
tin and (ii) 69.97% lead and 29.83% tin. This was likened to plumbers' solder (67% lead
and 23% tin) which would be 'mushy or pasty' between 183°C and 253°C. There has
50
been one previous attempt to manufacture lead scalings but these were of an unusual
blank 'droplet' type into which string was poked and no technical details are provided
(Vons, 1980, 51).
I obtained a bar of leaded plumbers' solder (most is now lead-free for use with
potable water supplies) and some hemp string. The problems noted below may well be
due to a possibly different composition in modern plumbers' solder compared to that
available in the 1930s. Alternatively, they could be due to an inadequate heat source
(methylated spirits burner) which could only heat the solder to just inside its molten
range.
It was encouraging to find that the molten 'lead' did not burn through, or even
leave scorch marks on, the (undampened) hemp string. The same was true when the
'lead' was applied to a cork block, showing that the hemispherical/conical reverses on
some sealings (Type 5) were not caused by the burning-away of cork bungs. 9 I had some
limited success in recreating similar reverses by pouring molten lead on to earth/sand,
although I would not wish to suggest that this was the Roman method.
The main problem encountered was that the alloy I was using solidified too
quickly to allow any decent impressions to be made. This would suggest either that the
composition of the solder was wrong or that the temperature was insufficient to keep the
solder in its semi-molten state (183°C - 253°C) for long enough. This therefore casts
significant doubt over the relevance of the findings mentioned above.
51
Notes
1. This reference is from 1-lenig, 1974, 33.
2. There are other sealings bearing similar impressions but it is not known whether or not
they conform to this shape.
3. Other dated examples which are probably of this shape include Records [0454] (with
residual Antonine material in Severan foundations), [0500] (in construction trench of a
building provisionally dated c.AD85/90) and 0508 (in demolition layers over a drain in
front of a Severan barrack block).
4. Pers.comm. J-C. Cheynet (Centre d'histoire et de civilisation de Byzance, College de

France) 1992. The inclusion of S(acra) M(oneta) An(tiochae) in the exergue on the
reverse would certainly suggest this. However, the production of such a complicated tool
for nefarious purposes also raises questions over my earlier statements that the
manufacture of such tools would require the marshalling of official resources. My
defence would be that for the production of lead sealings the individual would not bother
to produce a boulloterion, whereas for future monetary gain the initial outlay might well
have been seen as justified.
5. Attested by RIB 1060.
6. The styles of the designs and inscriptions appear to fall either side of the period in
which we are interested, although some doubtful examples possibly belong here. The
majority are housed in the Department of Greek and Roman Antiquities of the British
Museum, but there are two examples in the Musée des Beaux-Arts, Lyon and two in the
Bibliothèque Nationale. Many of the British Museum's examples are published in Kaibel,
1890, 627-8, whereas the Lyon examples are in Turcan, 1987, 35 and the Paris examples
in Rostovtzeff and Prou, 1900, 161.
7. C. the wooden writing tablet (the text of which refers to legio 111 Cyreiiaica) from
Egypt which has a rounded hollow cut inside a square border in order to accept a
circular seal box (Devijver, 1989, 267-72)
8. They are in the collection of the Department of Greek and Roman Antiquities, British
Museum and are published in Kaibel, 1890, 627-8, 2415.2, 6-7.
9. This lack of damage would also explain how sealings could be applied in a
molten/semi-molten state to fabric which would then leave its imprint on their reverses
52
Fig.I

1 2 (i-u) 2 (iii -iv)

3 4

5 6 7
10
8 9
Chapter 3
IMPERIAL SEALINGS - TIlE EVIDENCE FOR ThEIR USE
Evidence from impression
Sealings referring to the Ratio Patrirnonii 10016; 0052; 01 19J - There are three
sealings known which refer to the ratio patrimonii. They are [0016], [0052] and [0119],
the first two of which are identical. They all appear to date to the reign of Septimius
Severus, the first two from the period AD209-1 1 due to the reference to three emperors
(ie. including Geta), the other probably from AD198-209 due to the inclusion of
Caracalla but absence of Geta. The ratio pa(rimonii was the (department of the) account
of income from crown property, although originally it had been the account of the
emperor's private income from his personal property. it first appeared in the first century
with the ratio palrinlonii for Alexandria being attested in the mid-second century (ILS
1491) and a procurator pairimonli in Narbonensis during the reign of Marcus Aurelius
(AE 1962, 183). It appears to have been reorganised by Severus with the result that in
the third century we hear more of local officials such as those in Belgica and the two
Germanies (ILS 1330), Bithynia and parts of Africa. Later in the third century however
the post of procurator pa/rimonhi, and presumably therefore the ratio itself, seems to
have disappeared since it is no longer referred to. The date of the ratio patrimonhi's
zenith is apparently consistent with the identification of the emperors on our three
sealings. There are several references to officials of the rallo patrimonif in Rome
revealing procurators, tabularii, a commentariis and a cuslos (C'IL VI 3962;
8498-8509).
It is likely that these sealings were affixed to goods emanating from imperial
estates (de Laet, 1949, 165). There is a parallel for this in the tiluli picli from Monte
Testaccio where some amphora sherds are marked fisci ralionis palrin?onil provinciae
Baeiicae with others referring to Tarraconensis (CIL XV 4097-4133 - Baetica;
54
4134-4137 - Tarraconensis; 4138-4142 - uncertain).' These have been interpreted as the
produce of imperial estates with the painted inscriptions denoting the exemption from
customs duties (de Laet, 1949, 292). It is possible that this exemption also applied to the
glass unguent flasks which have pairimoni (RIB 2419.130-4) and vectigalpatrimo (RIB
2419.154-5) hollow-stamped on their bases, although on other flasks the base usually
just bears the maker's name and would seem an awkward place for information possibly
needed in transit. The inscriptions are translated as '(product) of the (imperial) estates'
and 'revenue of the (imperial) estates' respectively, although the latter is placed around a
bust set inside a wreath. Returning to the scalings, their purpose may also have been to
signal exemption (de Laet, 1949, 165, 430, nI), but in addition I would suggest that
possibly it was also to secure the goods in order to reduce petty thieving or prevent the
inclusion of excise-payable goods by the transporters. Therefore the impression not only
proved exemption en route but also, at the final destination, showed that the goods had
not been tampered with.
Sealings referring to the Ratio Castrensis 10069 - 0072; ?0015; ?0032; ?0240J -
There are seven scalings marked R C' which are assumed to refer to the Ratio ('as/rensis,
with three other possible examples.They are [0069](two examples), [0070], [0071](two
examples) and [0072](two examples) with [0015], [0032] and [0240] as the possible
examples. The ratio castrensis has been described as the account of that part of the
military administration which acted for the emperor as head of the army (Grenier, 1934,
650-1). The same author then refers to it as the part of the administration which was
responsible for the emperor's bodyguard, his wardrobe and his residences in various
parts of the empire while fulfilling his role as commander-in-chief. In Rome we find many
references to officials of the ratio castrensis. They held various posts such as procurator,
tabularius and dispensalor and the majority appear to have been imperial freedmen ((7L
VI 8511-8536; 33735-33738).
The two scalings of Septimius Severus in his role of head of the army [0072]
which were found in Lyon could easily be explained as supplies for the Praetorian Guard
55
or for the emperor himself on campaign during the siege and subsequent occupation of
the city. However it is much harder to explain the presence in the same city of two such
sealings of Marcus Aurelius with Lucius Verus [0069] and three of Marcus with
Commodus [0070-007 1] since there is nothing to suggest that any of these three ever
visited Lyon at the head of an army. This tells us That we are wrong to assume that these
sealings travelled with, or to, the emperor on campaign and that we should perhaps see
them in a less romantic light, perhaps revealing the provision of supplies etc. possibly for
the Urban Cohort of Lyon 2 or even for the upkeep of permanent official residences in the
absence of the emperor.
Sealings referring to the Anabolicum 10042; 0048 - 0049; 0091 - 0094; 0106; 0271 -
02721 - There are ten sealings which refer to the anaholic,im. They are [0042],
[0048-0049], [0091-0094], [0106] and [0271-0272]. There are several theories as to the
meaning of anabolicum, most of which are listed by Turcan (1987, 21-3). His preferred
explanation is that it is a tax in kind collected in Egypt, where he cites a reference to a
procurakir ad anabolicum Alexandrinum (Vitucci, I)iz. Epigr., 943)•3 The J-IisIoiia
Augusta (HA, A,ir., 45, 1) lists anabolicae species as glass, paper, flax and tow. 4 Turcan
says that these raw materials and finished products were requisitioned annually in Egypt,
exported, and then redistributed from Rome for the benefit of the army. Apparently he
makes no reference to the definition of to avc43o?tKov (the Greek form seen on [0049]
without the definite article) which is given in Liddeli & Scott as 'deferment of payments',
the reference being PAmh. 2.131.
Our examples appear to cover the reigns of Marcus Aurelius with Commodus,
Pertinax, Septimius Severus with Caracalla and Caracalla alone.
Sealings referring to the fiscus Alexandrinus 10097A; 01 15A1 - There are two
scalings pertaining to this organisation [0097A; 011 5AJ Apart from the reference to the
fisczis, we can tell that these belong in the Imperial category since, in addition to the
inscription FISC ALEX, [0097A] bears the portrait of an emperor, although opinions
56
differ as to whether this is Hadrian (Rostovtzeff, 1897, 471, 4) or Antoninus Pius (CII.
XV 7974a). The fiscus of Alexandria was the treasury for the revenues derived from
Egypt which was in effect an imperial possession. It is difficult to tell exactly how this
differs from the ratio patrimonii of Alexandria which is attested in the mid-second
century (ILS 1491). It could, however, be a precursor to the anahol,cz,m. Both of these
sealings were found in Italy, [0097A] in Rome (in the Tiber) and [011 5A] probably in
Rome. The Tiber provenance suggests that the items thus sealed were bulky enough to
require unpackaging on the quayside. This renders unlikely a previous suggestion that
this particular sealing came from a bag of money sent by the 'receveur d'Alexandrie'
(Daremberg-Saglio, 'Tessera'). Perhaps the sealing was attached to sacks of Alexandrian
grain or any of the items listed above as anaholicae species.
Sealings possibly referring to the Annona 10096 - 0097J -There are three sealings
which may refer to the annona [0096-0097; 0116]. These attributions are all rather
doubtflul. Record [0096] bears an inscription which has been tentatively expanded by
Mowat (1895, 217-219, no.13) as [Traian. De}CIVS [et fit.] AVGG., ANN(ona)
although this does not hang together very well. Record [0097] is even less certain since
its inscription is CSA / AVG which has been expanded, again by Mowat (1895, 217,
no.11 & 219), as C(ontra) S(criptor) A(nnonae) AVG(usti). Record [0116] is equally
uncertain since it bears the inscription G star AFDNAVGTRP around S star A. Dressel
believed that S A stood for S(tationis) A(nnonae) (CIL XV, 7959). These three meagre
pieces of evidence can in no way be taken as evidence for sealings refemng to the
annona. It is at first sight surprising that we do not have some definite evidence of
annona sealings, considering the amount of sacks etc. which must have been involved.5
However, I would suggest that if they did ever exist they would probably only have been
used as protection from pilfering since the sacks would have been the main, if not the
only, cargo on the grain ships. It would therefore be ridiculous for every sack to be
marked as being free from customs duty when virtually the whole load consisted of
obvious imperial property It might be the case that documents were carried by the ship's
57
captain to vouch for the tax-free status of his cargo6 but these would probably have been
sealed with wax rather than lead.
Numbered scalings 10100 - 0101; 01201 - There are three scalings which apparently
bear numbers, [0100] which has Ci) (800) above busts of Marcus Aurelius and Lucius has
Verus, [0101] which has COB (802) above the bust of Antoninus Pius1 and [0120] which
the inscription THE I XXV on one side and an imperial bust (possibly Septimius
Severus) on the other. Obviously there could not have been a discrete number for each
sealing since they were being struck from matrices. Therefore the reason must be to
identifi (i) the user of the matrix, (ii) the department in which he was working or (iii) the
relationship of the consignment to others, which could help, for example, in ensuring the
use of perishable goods in the correct order. We have no way of knowing what the first
two scalings were being used for but the the(...) of the third example has been interpreted
as iheca (envelope/case) or thesaurus (treasury/strong-box) (Garmcci, 1847, 55; 1866,
77). Despite the attraction of case or strong-box we must again remember that, due to
their nature, sealings could not have been used to mark individual packets or boxes with
separate numbers.
Dated scalings 10046; 0095J - Two scalings, [0046] and [0095], bear a date in the field
to the right of a bust of Trajan. There is also the possibility that a dated sealing showing
Serapis [1019], listed under miscellaneous in the catalogue, may also have some
connection with whatever authority was using the two depicting Trajan. These dates use
the Alexandrian regnal year system (Butcher, 1988, 115). Turcan states that our second
example must have sealed a box belonging to the imperial post and coming from
Alexandria (1987, 17) For our first example however, he says "comme Ic précédent, ii
scellait un envoi du service imperial venu d'Alexandrie" (1987, 18). It seems likely that
these scalings did indeed belong to some branch of the imperial service, but we cannot
conclude from the fact that they are dated that they necessarily sealed anything sent by
cursu publictis (whether he means boxes of actual mail or not e). I feel that it is possible
58
to look too deeply into the reasons for the dating of these sealings. We should bear in
mind the high level of bureaucracy in Egypt which is mirrored in the fact that Alexandria
was one of the few places in the empire to place a date on its coins. 9 Therefore I believe
that there is no special reason for the dates on the sealings which means that they cannot
help us to assess the use of those scalings.
Evidence from form and reverse
More information on the way in which the sealings were formed and attached can
be found in chapter 2. Here we shall try to use the evidence of different forms in order to
assess use.
If we look at the types of sealing then we find that the following are
double-sided:
[0001], [0010-00141, [0023], [0064-0066], [0091-0093], [0099], [0106],

[0114-0115], [0119-120], [0125], [0187], [0197], [02081, [02111, [0236], [0239],
[0271-0273], [0275], [0278].
This may suggest that they were designed to be read on both sides and may have
been loosely-fitted to the item or document to facilitate this. It is tempting to think of the
Byzantine use of scalings where the majority were attached to thread which sealed letters
etc. en roule whereas others hung from documents and acted as a form of signature or
guarantee of the contents (Oikonomides, 1986, 5-13, 1987, 101). However, if we return
to the idea of bulk goods then there are several ways in which this type of sealing could
have been used: (i) tied around knot/loose ends of rope securing the neck of a sack in
order to prevent the sack being secretly opened (Still, 1993, 408); (ii) tied around
knot/loose ends of rope bound around a bale or rope netting in which the bale is wrapped
(ibid.); (iii) hanging from thread which has been passed through a catch or lock
59
mechanism on a box (Richmond, 1936, 106 although he is actually referring to military
scalings).
Rather than list all of the numerous single-sided scalings we shall extract those
examples which provide some evidence as to the nature of their attachment.
The reverse of one of the sealings mentioned above in connection with the fisciis
A!exandrinzis is depicted in Daremberg-Saglio ('tessera') as being blank with a swelling
rising from the surrounding flan. Appearing at right-angles from this swelling are the four
ends of two cords which presumably join in a knot inside the sealing. This is a sensible
use of a sealing despite the fact that virtually all other scalings only have one cord
passing through (although it is possible that the normal, apparently continuous, cord is
actually two ends knotted together inside the sealing). The arrangement of the cords on
this fiscus sealing suggest that it sealed neither a money bag nor a sack of grain but
something more square, like a box or bale.
The following scalings bear the accidental impression of wood grain on their
reverses:
[0018], [0022], [0049-0050], [0053-0054], [0058], [0067-0068], [0085],

[0094-0095], [01001, [01051, [0153].
This suggests that these scalings were applied to wooden boxes, crates or even
barrels. There is also a possibility that they sealed wax tablets, although there is no direct
evidence for this.
The following show evidence of accidental cloth impression.
[0046-0047], [0069], [0083], [0098], [0101-0102], [0110], [0279]
60
It seems likely that these examples would have been affixed either to sacks or to
bales covered in cloth
Whereas the above examples were all held in place by thread passing through
them, the next three were apparently nailed in place:
[0063J[01 131(0235].
This suggests that they were fixed to wood, presumably a crate. It would not be
impossible for this technique to seal the lid of a crate by virtue of the nail passing
through, although it seems rather strange.
Evidence from association with findspot
Severan scalings at South Shields 10021 - 0022J - The group of sixteen lead sealings
found at South Shields are generally said to be connected with the use of that fort as a
supply base for Septimius Severus' campaign in Scotland (Frere, 1987, 159-161).
Despite the danger of attempting to link archaeological evidence to specific historical
events, this seems likely although we do not know the nature of the supplies thus sealed
is it a question of food, clothing or arms? The archaeological evidence for South Shields
being used as a supply base is that twenty or so additional granaries were built which left
little room for anything else. We also know that Corbridge was the contemporary depot
which handled weapons and armour (Dore & Gillam, 1979, 63) so it seems likely that the
South Shields sealings were attached to supplies of food, especially corn. Dore & Gillam
(op.cil., 64) believed that the corn was being brought up by sea from further south in
England, while Birley adds to this the Rhineland (1988, 173). Johnson actually states that
analysis of a sample of the corn has revealed that it is more likely to have come from the
continent and he again chooses the Rhineland, although he fails to tell us his source
(1989, 85, 105-6) This corn had presumably been collected either as compulsory
61
purchase (frurnenium emptum) or as one of the 'irregular practices' Which eventually led
to the annona mi/hans, taxation in kind for the provisioning of the army (Rickman,
1971, 278; Rostovtzeff, 1957, 1 484, 517; 11 712, note 15),'° or even as the rent or tax
gathered from the tenants of imperial estates (Thompson, 1987, 560; 566-7)
Thus it appears that these Imperial scalings had little connection with the person
of the emperor himself. The only exception to this would be if the produce collected by
compulsory purchase or as the annona militani.s only received imperial sealings when it
was required for specific campaigns involving the emperor, in which case the proximity
of the emperor would indeed have some bearing on the presence of the sealings. Despite
this possibility, the most important point remains the fact that these Imperial scalings
were not for the emperor's personal use but for that of his army, and then seemingly only
when on campaign. Broadly speaking, the emperor, through various agencies," was
always responsible for supplying the army and so these sealings should not come as a
surprise (Garnsey & SaIler, 1987, 89). There may be a fourth century development of
this expeditionary supply use in chapter 4 ('Official Sealings - The Evidence for their
Use') where the Praetorian Prefect of the Gauls appears to be supplying an army on
campaign.
Trajanic seahngs at Svishtov 10162 - 01641 - There is a similar case at Novae

(Svishtov) in Moesia Inferior, where three scalings were found, all of which apparently
refer to the emperor Trajan. These are also used as evidence of supplies for a military
campaign, this time in Dacia (Mrozewicz, 1981, 82). Again, we are apparently seeing
Imperial scalings being used for military campaigns conducted in person by the emperor.
We do not know if Trajan himself ever visited Novae during these canaigns' 2 or on his
tour of the Danube frontier in AD98/99 but it seems more likely that, like the South
Shields examples, these sealings belonged to military supplies rather than the emperor's
personal baggage. Lepper and Frere (1988, 92-3) list three main invasion routes which
are popular among strategists and only one of these lies close to Novae. However, I do
not mention this in order to use the scalings to justify this route (following the valley of
62
the River Ott), only to suggest that if the invading armies used one of the other routes,
the sealings could belong instead to the period after the conquest and relate to the
consolidation of the new province of Dada. In this case they could presumably still be
considered as military supplies organised by the emperor for a special task which kept the
units away from their bases Although these supplies could not have been the annoiia
milifaris as suggested for the Severan examples, they could easily represent items gained
through the earlier system of compulsory purchase.
Imperial sealings at Izvoarele 10134A - 0158; 0168; 0171-0176J - The presence of the
sealings here could have some connection with the suggestion that Sucidava (Izvoarele)
was one of the two unnamed frontier towns which were turned into customs
stations/markets after Valens made peace with Athanaric in AD369 (Diaconu, 1963,
548-50; Barnea, 1969, 23-4 and his note 7; Still, 1995, 354). The terms of the treaty
meant that the Visigoths had to confine their commercial relations with the Romans to
these two towns (Themistius 135 bcd). The comparatively large collection of imperial
sealings could have been attached to goods being sold at these markets which the
barbarians were allowed to attend. These goods could have been obtained by the imperial
authorities in the form of rent or tax in kind from imperial estates and then sent to
Sucidava (and the other unknown site) to take advantage of this virtual monopoly. If we
reject the notion of imperial goods for sale, but retain the idea of the Visigothic market,
then the sealings could simply denote supplies to the army unit which would undoubtedly
have been present to oversee the proceedings. The site is sometimes described as a fort
so the military presence may well have been permanent although this still does not
explain the relatively large number of imperial sealings.
Continuing with the idea of military supplies, we know from Ammianus

Marcellinus (27.5.2) that a couple of years earlier, in spnng AD367, Valens had drawn
up his army at a base near a fort called Daphne, prior to crossing the Danube by a bridge
of boats. This fort has been linked with the site at Izvoarele (Diaconu, 1971, 3 11-3 18)
63
and so perhaps our sealings can yet again be connected with a particular campaign led by
the emperor himself.
Imperial sealings throughout Thrace 10230 - 0270J - It should be noted that the
majority of sites producing imperial sealings in Moesia Inferior were forts whereas in
Thrace approximately half, if not more, were possibly road stations of one type or
another. 13 This opens up several possibilities, though not all are mutually exclusive. The
situation is confused by the current lack of knowledge concerning the development of
road stations in general. Even our standard idea of the scalings being attached to military
supplies is no longer as clear-cut as usual:
(i) supplies for armies passing through on well-planned campaigns.'4

(ii) supplies for frontier troops (Iirniianei) and mobile field-army units (comilalenses,).
(iii) supplies for military units stationed at the road stations.'6
Other options are:
(iv) supplies for the emperor's personal entourage when passing through on planned
campaigns - see (i) above.17
(v) supplies for official (but not overfly military) staff of the road stations.'8
(vi) supplies for the animals of the cursuspublicus.'9
Several of these ideas are clearly untenable for certain sealings which are thought
to show Nero and other early emperors.
Another option suggested to me for one site, Kalugerovo (ancient Arzus), is that
it fulfilled the role of an emporium (pers. comm. Dr. Gerasimova-Tomova) which could
explain the large number of scalings of different categories 20 However, unless we
envisage the state selling off a surplus of rent/tax from imperial estates, as suggested at
Sucidava, the imperial scalings would still have to fit into one of the above options
64
Imperial sealings at Ickham 10010 - 0014J - These sealings, one of Constantine II as
Caesar (AD317-37) [00101 and four of Julian (AD360-3) were found on the site of a
large Roman masonry building (2granary) near Canterbury. The site is on the nearest
high ground to a late Roman water-mill, I lOOm away, used for metal-working as well as
corn-milling, which apparently supplied the Saxon Shore forts of Richborough and
Reculver (Young, 1981, 32-40; Still, 1995, 347-356). If we can accept the military
nature of the industrial site and a connection between the two sites then we again have
imperial sealings in a military context.
Frere obviously sees these as provisions for campaigns since he links Constantine
Ii's scalings with the expedition of Constans in 342/3 and those of Julian with the
campaigns of 367, saying that 'the packages concerned had been sealed up some years
previously in each case' (RIB II, 1, 88)21 This is not impossible but does seem rather
laboured so perhaps we should see them as regular supplies even though there are so
few. Another apparent problem is that Frere therefore connects these supplies with a
field army (comilalenses), whereas Young links the nearby industrial site with the Saxon
Shore forts and therefore the limiianei (ibid.). However, we should remember that the
post of comes Briianniae and the associated mobile field army may only have been set up
towards the end of the fourth century (Mann, 1989, 11), so any preparation of supplies
for a mobile army arriving in the south-east would, by default, have to be carried out by
the men of the Saxon Shore. Alternatively, if the officer in charge of the Saxon Shore
had already been promoted from dzix to comes at this date it would remove the
distinction, since Mann thinks that the post could have been of fill comitival rank, at
least by AD367 (Mann, 1976, 6). On the other hand it could be as simple as the
newly-arrived comitalenses commandeering the masonry building as a storehouse with
river access to the English Channel.
The limilanel would not have needed extra supplies for campaigning since they
were permanently stationed on the Saxon Shore but if we see the comes Liloris Saxonici
commanding a mobile army (Johnson, 1979, 146) at the dates represented by the
65
scalings, then we could possibly imagine some minor expeditions which qualified for
extra supplies. This is only a hypothesis and should be treated with caution. Another
option is that we should see these scalings as belonging to everyday supplies for the
limitanei, although this does not fit in with our evidence for the earlier periods.
Evidence from written sources
We know that any goods related to the imperial fisc were exempt from duty. This
is explained clearly in Dig. XXXIX, 4, 9, 8:
Fiscus ab omnium vectigalium pracstationibus immunis est.
The author, lulius Paulus, wrote during the reign of Severus and Caracalla,
Caracalla, Elagabalus and Severus Alexander but this was obviously still the state of
affairs under Justinian when the Digest was compiled. It was presumably the same
throughout the empire since there is little point in charging oneself.
This immunity, however, was apparently lost once the goods were being traded
by private individuals, as can be seen in the next sentence of Dig. XXXIX, 4, 9, 8:
Mercatores autern, qui de fundis fiscalibus mercari consuerunt, nullam iinmunitateni soluendi
publici vectigalis usurpare possunt.
The emperor's exemption also extended to his son and wife, as can be seen from
Dig. XLIX, 14,6, 1:
Quodcurnque privilegii tisco competit, hoc idem ci Caesans ratio ci Augustae habere sold.
66
Examples dated by context
Despite the relative quantity of imperial sealings, there are apparently none for
which a date has been suggested on the grounds of the archaeological context in which
they were found
Conclusions
All imperial sealings would have guaranteed the physical integrity of the contents
and would have acted as proof of exemption from custom duties at any collection points
which they may have passed.
The many different types of imperial sealing preclude any single description of
their exact use but it would appear that, in general, the scalings found in the provinces,
and not naming any government agencies, were attached to supplies for the army and are
certainly not to be used as proof of the emperor's presence. It may well be that he was
present at the head of his army but this is puirely coincidental. The main problem is
whether or not the supplies thus sealed were only organised for special campaigns or
whether they were for everyday use. The Trajanic and Severan scalings suggest that they
were for expeditionary use only whereas the Ickham scalings, and possibly those from
Izvoarele, would suggest that regular use is possible. Another problem is the almost
parallel usage of the putative sealings of the Praetorian Prefect of the Gauls. I have
attempted to answer this in the relevant section of chapter 4.
The supplies in question could have consisted of food, clothing or weapons,

although food is probably the main, if not the only, candidate. Davies states that the basic
peace-time diet consisted of 'corn, bacon, cheese, and probably vegetables to eat and
sour wine to drink; the soldier would also have access to salt and olive-oil' (1971, 125 =
1989, 189) Presumably the corn, wine, salt and olive-oil would have been brought in
67
from elsewhere but we cannot be so sure about the bacon, cheese and vegetables which
could have been produced in the lerrilorium or prata around the camp). Wine and
olive-oil can be virtually ruled out as recipients of lead sealings since none have been
found in association with amphorae on shipwrecks. If we say that one of the functions of
sealings was to prevent petty pilfering then we could probably also rule out bacon and
possibly cheese since any theft of a part would be instantly noticeable (although this
would depend heavily on the size and shape of portions). This leaves us with corn, salt
and vegetables. Salt may have been transported in pitch-sealed ceramic containers
(Milne, 1985, 107) whereas corn and vegetables would have been packed in sacks. Any
of these could have been sealed with lead and it is only the position of corn as a staple
and the quantity in which it was required to feed an army that leads us to prefer this.
There are of course many other items of food which were consumed in military camps
but we would not expect these luxuries to have been imported under the imperial seal.
68
Notes
1 The Baetican examples cover the period from AD2 14 to 230 whereas the one dateable
Tarraconensis example (CIL XV 4137) is from AD235 (dates provided by consuls named
in accompanying thu/i pic(i on the sherds). There is also an example from Chester found
with 3rd century material on the site of a possible mansio. The notes for that example tell
us that 'this formula is attested for the third century when the south Spanish export trade
was taken over by the state This is usually linked with the presumed confiscation of the
estates of adherents of Clodius Albinus in 196. Professor Almeida, however, gives
ADI 17-260 as the extreme date range for the sherd (implying some shipping by imperial
agents from the beginning of Hadrian's reign) and, within that period, AD220-26 as the
most likely date of the vessel' (Brilannia xii (1981), 381-2, note 44).
2. Perhaps the Urban Cohort enjoyed a special position in the army hierarchy.
3. This otherwise unpublished ftmneral inscription from c.2nd C. is apparently in the

collection of the Comune di Roma: D. M. Optati Aug. 1. qui pr(ocuravil) ad anabolicum
Alex(andriae) item in provincia Cilici(a). Turcan presumably takes the word item to
mean that Optatus was a procurator in Cilicia and not a procurator ad anabolicum in
Ci/icia. If he was the latter then it would spoil Turcan's preferred idea that the
anabolic,,m relates only to Egypt.
4. Vectigal ex Aegjpto ii'r,Thi Romae Aurellanus vitr('e)i, chartae, huh, stupp('a,)e atque
anabohicas species aeternas constiluil.
5. We must remember that the species annonariac also included large amounts of oil and
smaller quantities of wine, fat and fruit (Warmington, 1954, 60).
6. This is an extension of an idea which I have previously expressed elsewhere (Still,

1993, 408).
7. It was Turcan (1987, 1 8-19) who suggested that these Greek letters represented
numbers. However, the fact that these are our only examples and the numbers are so
similar raises the possibility in my mind that the omega is the initial letter of an unknown
word, with beta as the second letter of this word.
8. If he does mean actual letters does he believe that the sealing acted as a frank or
postmark on the box to denote the date of dispatch?
9. There is also a sealing [1019] in the Miscellaneous category (although it may actually
be Imperial or Official) which appears to be dated in the Alexandrian system.
10. In the same way that taxes in kind were raised from the Egyptians for Severus' visit
in AD 199 and for Caracalla's campaign in the east in AD214-5, although even these may
have been paid for (Rickman, 1971, 280).
69
II. In Strabo's time it was the responsibility of the imperial procurators (iii 4, 20). In
later periods the task fell to the Praetorian Prefect (Jones, 1964, 371).
12. Lepper and Frere (1988, 84) suggest that Cichorius' casts 86-88/scene xxxv from
Trajan's Column shows him disembarking at either Oescus or Novae, although they
prefer Oescus.
13. We should bear in mind that the scalings from Moesia Inferior are found in forts and
those from Thrace are found in road stations probably for the simple reason that these
are the main centres of officialdom in those provinces.
14. The 'Main Route', (on which all of these possible Thracian road stations are placed)
identified in the Antonine Itinerary by van Berchem (1937, 117-201), is said to be the
planned route for Caracalla's progress to the east in AD214-5 naming places where
supplies for the journey were to be collected and stored. Similar preparations must have
taken place on every such occasion (Chevallier, 1989, 36-7; Still, 1994, 392).
15. Rickman suggests that some mansiones were used as rear supply bases for the
frontiers and he seems to include the cornllafrnses in this system (1971, 287-9). He
describes these sites as 'great central supply dumps into which officials like the
primipilares could gather supplies before dividing them between individual forts'. Jones
(1964, 627) points out that, since they had no fixed stations, the comila(ense,s were
usually given warrants to receive rations from the provincial governor but that the Coder
Theodosianus (VIII.iv. 17.385) says that under Gratian the comilatenses in Illyricum
received their supplies from storehouses which had been stocked by prinupilares (I e the
supplies were brought from another province by the method usually adopted for
limilanei).
16. There is much conflicting evidence as to the date of any possible military presence in
road stations (Chapman, 1978, 173-80). Some authors point to cases where the army
was involved from an early date (Chapman, op.cil.,175; Chevallier, 1989, 184) whereas
others see the introduction of military units as being a move to protect the annona
mi/hans (Pflaum, 1940, 189-240). However, further confusion is caused by the fact that
some authors think that the annona was set up by Septimius Severus (van Berchem,
op.cil., 146ff) while others believe that it was only put into fill operation by Diocletian
(Rostovtzeff, 1957, I 484, 517; II 712, n.15; Rickman, op. cii., 278 -83), although it
should be noted that these opinions are not mutually exclusive. It may be prudent to say
that some, but by no means all, road stations may have had a military presence from an
early date.
17. Unlikely if we take our earlier conclusions into account but perhaps we should
consider each case separately.
18. HA Severus 14 says that Septimius Severus transferred the running of the cursus
publicus from the private domain to the imperial treasury, so perhaps from this time
70
onwards the staff would have been paid (in kind?) by the government. In addition to
these civilian staff we should perhaps also include the paramilitary police services such as
the frumenlarii who, from Diocletian onwards, were replaced by the agenles in rebus
19. Following on from note 16, the government would presumably have also paid for the
fodder for the horses etc.
20. Arzus has 26 imperial, 13 civic and 34 miscellaneoua sealings, but this large
collection may simply reflect the amount of metal detecting conducted on the site. For
another problem with the emporium idea see Still, 1995, 355
21. Neither Constantine II nor Julian ever visited Britain so we can immediately reject
the ideas of personal baggage and campaigns led by the emperor in person. In fact, we
are not aware of any military expeditions to Britain in those reigns, which has led to
Frere's suggestion.
71
Chapter 4
OFFICIAL SEALINGS - TilE EVIDENCE FOR ThEIR USE
None of the inscriptions on any of the Official sealings in the catalogue make any
reference to the use of those sealings. However, two examples, [0286] and [0287], refer
to named officials along with their titles. A further four examples, [0288-0291] (with a
possible fifth [0292]), name a single official, the identification of whom I believe I have
discovered.
The sealings of provincial officials 10286 - 0287J - Dressel (('IL XV 7970) expanded
the inscription on [0286] as [.V]OLVSSIVS AEMILIAN[us] [pr]OC(urator)
AVG(usti). He also suggested M. VMBRI PRIM(i) PROC(onsulis) AF(ricae) as an
expansion for [0287], with a possible candidate (GIL XV 7969).' Although the
abbreviation for proconsul is usually procos, the limited space on the sealing could have
brought about this unusual abbreviation. If we insist on proc(urator) then we have to
assume that this man is previously unknown, by no means impossible since the list of
known holders of this post is rather short. We do not know where Volussius Aemilianus
was posted2 but Primus would appear on the evidence of the sealing to have been in
Africa. Record [0286] was discovered in the Tiber and [02871 may also have been
discovered in Rome. Therefore it would appear that these sealings show the officials
sending items into the capital. The discovery of the example in the Tiber (i.e. thrown
away at the dockside) suggests that this, at least, was attached to bulky merchandise of
some sort which had to be unpacked there as opposed to official documents such as
letters or lists of accounts which would only need to have been opened at their final
destination. It is difficult to say whether gold or silver representing collected taxes would
be unpacked on the wharf for ease of transport or whether they would be taken to the
treasury intact. It is possible that the merchandise was grain since the annona was
72
presumably considered to be another form of tax or rent (Garnsey & Sailer, 1987, 49-50)
and therefore within the domain of the procurator. On the other hand, John Mann,
admittedly talking about collection for use within a province, says that there was no
corn-tax in the Principate and that any compulsory purchase was conducted by the staff
of the governor, not the procurator (1985, 21-22). This last point may explain why, if
our other sealing represents the transport of grain from Africa to Rome, it was being sent
by a proconsul.
The seaiings of Maximinus 10288 - 0292J - The identity of this Maximinus has been
quite an enigma for some time. The C hi-Rho on the reverse of all four/five of the sealings
immediately tells us that we are not dealing with either of the emperors of this name,
both of whom were pagan. The bust on the obverse, around which the name is
circumscribed, was thought by GObl to resemble Magnentius, although I believe that the
fact that it is bare-headed has helped this suggestion. Therefore this Maximinus was said
to have been a previously unknown high-ranking military officer of Magnentius (1969,
58-9). However, I think it is much more likely that this Maximinus is the anti-hero of
Books 28-30 of Ammianus Marcellinus, the Praetorian Prefect of the Gauls from
AD371-376 (I'LRE vol.1, Maximinus 7)•3 The bust could easily be Valentinian I but is
probably better seen as Maximinus himself depicted in the style of the time.
At the risk of searching for an historical explanation, the campaign of Vaientinian

I in AD 375 which set out from Trier (seat of the Praetorian Prefect of the Gauls) and
then made its headquarters in Carnuntum for three months (the findspot of [0288-0289])
for the preparation of equipment and supplies, would appear to provide an ideal reason
for sealings of the Praetorian Prefect of the Gauls being found in Illyricum (Amm.
Marc.xxx. 5). The other two/three sealings which were found at unknown locations in
Hungary can be interpreted as relating to Valentinian's subsequent moves to Aquincum
(with an excursion into the territory of the Quadi at NógradverOce), Savaria and
Brigetio.
73
It is not clear whether this campaign is the same event referred to by Ammianus
when he says earlier:
quod ad tulelam Illyrici Gallicam militis validurn accesserat robur (xxix.6. 16)
If not it could present another opportunity for these sealings to have arrived in
the area.
Further examples of these scalings, which are not in my catalogue, are said to
have been found in Enns (ancient Lauriacum) and Sisak (ancient Siscia) (Deringer, 1965,
220ff; Althldi, 1931/2, 6ff). Neither of these findspots presents any real problem since
military detachments could easily have travelled to these places, bringing with them
supplies packaged in Trier.
Jones (1964, 627) says that a fourth century emperor engaged in large-scale
operations would be accompanied by his praetorian prefect, or the prefect of the area in
which he was campaigning, who would personally arrange the supplies for the army,4
although this also seems to be the state of affairs from day to day (op.ciL, 371). It is
presumably due to the fact that Valentinian had led his army from Trier that Maximinus
was given the task of organising supplies. The fact that praetorian prefects were in
charge offabricae and were also responsible for the annonae at this date is of little help
in narrowing down the items sealed. He had to supply the army with everything and,
anyway, there is no way of distinguishing whether our scalings were attached to these
shields etc. and rations, or some other commodity such as clothing. However, assuming
that I have chosen the correct man, the basic suggestion of equipment/supplies for a
campaign still stands since there are apparently few other reasons for a Praefecius
Praelorio Ga/lianim's scalings to be found in such relative quantity in someone else's
prefecture. Previously Maximinus had held the position of vicariiis urbis (AD3 70-1) and
praefeclus annonae (AD368/70). I have been unable to connect these offices with the
findspots of the scalings, although Pannonia was in the Italian prefecture. It is
74
unfortunate that the sealings themselves do not mention the office which he held,
although, since we apparently have no other sealings for any holders of these offices, the
sealings are unique regardless of whatever position he was in.
The usage of these sealings is virtually the same as that suggested for some
imperial scalings in chapter 3. Some of those sealings probably represent everyday
provisions for soldiers whereas there is the possibility that others may be connected only
with specially-prepared military expeditions. They date from throughout the empire and
so we probably have evidence from within the lifespans of several different administrative
techniques. However, one group of imperial sealings is dated AD360-363 (or later) and
another may be connected with events in either AD369 or possibly 367. Our 'praetorian
prefect' scalings apparently date to AD375 whereas we are told by Ammianus
Marcellinus (xiv. 10.3-4) that even back in AD354 the troops were blaming the
Praetorian Prefect, Vulcacius Rufinus, for the delay in the arrival of their supplies.
Therefore we cannot say that the difference is simply one of date and a change in
administration. One possibility is that the responsibility for military supplies in the fourth
century, and perhaps even prior to Diocletian (Jones, 1964, 371), was in the hands of the
Praetorian Prefect but that he ostensibly exerted his authority on behalf of the emperor,
using imperial scalings. Maximinus, however, may have produced scalings of his own for
this purpose. 5 I hesitate to mention it, but perhaps this confidence is a foretaste of the
'insufferable arrogance/overbearing conduct' for which Ammianus tells us Maximinus
was eventually executed by Gratian (xxviii. 1.57).
Anepigraphic sealings
Three anepigraphic scalings [0283-02851 in the Official category all bear a similar
impression. This shows a man on horseback, holding a spear, riding right on a pearl
groundline. This general design is said to bear a strong resemblance to coin reverses of
Constantius II, as noted for [0283] by Mattingly (l3ushe-Fox, 1949, 138, no.164).
However, [0283] from Richborough is slightly different to the others. Therefore we
75
have [0284] (together with an identical uncatalogued example - Mills, 1995, 74,
p1 M215) from Britain and the also identical [0285] from Trier. This distribution is not
surprising in view of the fact that Trier was the seat of the Gallic prefecture of which
Britain was one of the constituent dioceses. The design provides no real clue as to the
origin of the sealings, although it is possible that the matrix had been cut, and was being
used, in the mint city of Trier itself Perhaps even more likely is that several similar
matrices were cut in Trier and then distributed to the various agencies for whom they
were made, including those in Britain.
Another three anepigraphic sealings show soldiers holding standards. Again,

these are presumably linked to coin reverses of the 4th C. One example from Trier
[0285A] shows two men holding a standard between them, while [0287A] from Moesia
Inferior and [0292A] from Pannonia bear similar designs of four men holding spears and
standards.
There is little information to be gained from the physical characteristics of the

sealings in this category. Records [0288-0292] are all two-sided so there is the possibility
that they hung loose from threads, although this does not mean that they were not able to
secure whatever they were attached to. The rest are all one-sided although we have
information on the shape of the reverse of only two examples, [0284], which displays a
shallow swelling, typical of many one-sided sealings and [0287A] which has a shallow
swelling but with a central nipple. There are no known impressions of fabric or wood
grain on any of the examples.
76
If it seems unlikely that all sacks of grain belonging to the annoiia were sealed
with the sealing of the I'roczira/or Aiigusti or even the governor (see above), there is an
interesting text from Egypt dated to 2BC which sheds some light on the care of the grain
during its journey down the Nile to Alexandria (JJP iv, 1950, 106-15; given in Lewis &
Reinhold, 1966, 141-2) and which may give us some clues'
From the Oxyrhinchite nome
Ammonius son of Ammonius, pilot of a gosernment boat with the emblem... attached to him as
escort being Lucius Oclatius, soldier of Legion XXII, second cohort, century of Ma'cimus Stoltius; and
Herinias son of Petalus, pilot of a second boat with the emblem Fgypi, attached to him as escort being
Lucius Castricius, soldier of Legion XXII, fourth cohort, century of Titus Pompcius. This is a sample of
the cargo we have loaded from the har'est of the 28th year of Augustus - Ammonius, 433¼ artabs of
wheat, loaded to the rail, Hermias, likewise 433¼ artabs of wheat - a total of 866V2 artabs of wheat
consigned by Leonidas and Apollonius, .cilologoi of the Lower Toparchy, Eastern Division, plus the
supplement of one half artab of wheat per hundred We loaded from the 2nd of Hathyr to the 4th of the
same month, and we have sealed [this jarj with both our seals, Ammonius' with a figure of Ammon,
Hermias' with a figure of Harpocrates. Year 29 of Augustus, 1-lathyr 4.
This text was written on ajar containing a sealed sample of the grain loaded on to
both boats. The seals in this case were presumably impressed in some sort of mortar
which had been spread over a bung in the neck of the jar. 6 On arrival at Alexandria the
jar would be opened and the 'control sample' would be compared with the grain on the
boats to ensure that the cargo had not been spoiled during the journey. It is not
impossible that there was a similar arrangement for the journey to Rome (Lewis &
Reinhold, ibid.), with the procurator giving his guarantee of a sealed sample using lead
scalings in some way. Although Egypt is often a special case and we cannot be sure that
these rigorous precautions were observed anywhere outside of that province, the words
of Seneca (Brev. vit. xix, 1) suggest that great care was indeed taken of this lifeblood for
Rome:
Simile tu putas esse, utrum cures Ut incomiptum Ct a fraude advehenlium Ct a ncglcgentia

frumentum transfundalur in horrea, ne conccpto umore vitietur Ct concalcscat, ut ad mensuram
pondusques respondeat.
77
Conclusions
Apart from the special use of the suggested sealings of the Praefecftis Praelorio
Gallianim, there is very little information to be gleaned from these few scalings This
paucity precludes any attempt to see a pattern in their use or, with the exception of the
Pannonian scalings, their place of discovery. Although the sealings cannot tell us, it is
likely that, due to their official nature, they usually indicated exemption from customs
duty, in addition to any guarantee of physical integrity of the contents.
78
Notes
I. As Dressel suggested, our man is probably Umbrius Primus (PIR' V 596) who is listed
as being proconsul of an unknown province in the time of Papinianus (i e. in the Severan
period). He died in office and is believed to be different from M Nummius Umbrius
Primus Senecio Aibinus (PIR' 11189), consul in AD206. We have a detailed c,,rsus
honorum for the latter and there does not appear to be any connection with the man of
our sealing. We also know of M. Umbrius Primus (PIRE I, IJmbrius Primus 4) who was
consul suffect in AD289 We have no further information on this last man.
2. The only information relating to Volussius Aemilianus in P!R' is taken from this
sealing!
3. Maximinus is depicted by Ammianus as an evil, manipulating schemer Before

achieving high office he is described as 'tamquam subterraneus serpens per humiliora
reptando' (xxviii.1.7) and then on being allowed to preside over a court case as
praefeclus annonae he behaved 'Ut saepe faciunt amphitheatreales ferae, diffiactis
tandem solutae posticis'. He was himself a Pannonian and the invasion of the Quadi and
Sarmatians which was being repulsed had actually been caused by the underhand tactics
of his son, Marcellianus, who had been created d,.'x of the Pannonian province of Valeria
at his father's suggestion (Amm. Marc. xxix.6.3f). Maximinus' brother-in-law,
Valentinus, had been exiled to Britain and was executed for subversion there (Amm.
Marc. xxviii.3.3ff). Animianus tells us that he will, in due course, give the details of the
death of Maximinus who was executed during the reign of Gratian (xxviii. 1.57).
However, as John Matthews points out (1989, 211), this is one of the two unfulfilled
promises in Ammianus' work.
4. The responsibility in earlier periods appears to have ostensibly been that of the
emperor himself (see chapter 3).
5. Roger Tomlin (perc. comm.) says '...for such a senior official to usurp a quasi-imperial
prerogative, the striking of 'coins', seems to me perilously close to maieslas.'. This
argument appears to centre on the presence of the portrait of Maximinus rather than the
actual sealing since purely epigraphic examples naming other officials are known, as seen
above. The problem would surely lie with the fact that Maximinus was advertising
himself as the person who was providing for the army. However, perhaps Tomlin is
broadly correct. If the practice was soon stopped then it could explain why these sealings
are the only examples known for any Praetonan Prefect.
6. This method was used for sealing amphorae. The space above the cork bung was filled
in with an opercidum of mortar or pozzolana which was sometimes stamped (Peacock &
Williams, 1986, 11-12, fig 4; 50).
79
Chapter 5
TAXATION SEALINGS - TIlE EVIDENCE FOR ThEIR USE
Sealings bearing the inscription STAREL INP 10293; 0295-03011 - These sealings,
mostly from Lyon but with one example from a similar waterfront context in London,
have been interpreted as belonging to the Sfalio Arelatencis, the customs station at Aries
(Dissard, 1905, 14; Grenier, 1934, 654). No attempt has ever been made at expanding
the abbreviation INP on these sealings.' There are many possible expansions but I have
narrowed down my suggestions to the following:
ST(ationis) AREL(atensis) INP(ressio) - the stamp of the Aries slatio.

ST(atione) AREL(atensis) INP(ressatus) - having been stamped by the Aries slatio.
ST(atione) AREL(atensis) INP(ositus) - put on by the Aries slalio.
ST(atio) AREL(atensis) IN P(ortu) - the Aries stalio in the docks (perhaps as opposed
to any others on the land routes into the city?).
The first two are probably closest to the truth. The last suggestion is interesting
since it would not require the change from imprimo to iliprimo, or impono to inpono,
(not that this is a serious obstacle in itself) and could also make more sense of the one
example [03011 on which the inscription may read STARELIP, which would otherwise
be explained as an error on the part of the die-cutter (or an error in Dissard's reading).
However, although the 'N' is strictly unnecessary, perhaps the abbreviation of 'IN' to 'I'
would be unusual.
Grenier (loc.ciL) saw these sealings as denoting that the quadragesima

Gaiiarum had been paid on the goods at Aries and that they were then able to circulate
within Gaul with exemption from any firther payment, especially at the central slalio at
80
Lyon. De Laet followed this and suggested that one of the duties of the internal customs
stations would have been to check that the payment had been made (1949, 168-70).
The comparatively recent discovery of an example from London [0293], while

providing information on the minimum distance travelled and possible routes taken, does
not shed any light on the functioning of the qriadragesima Ga!Iiarum.2
Sealing possibly referring to a tax farmer of Cyrenaica 102941 - This sealing bears
the inscription MI[ .](OTH together with a bust of Zeus Ammon. Turcan suggests that it
could refer to a tax collected by a tax farmer (ni,9otç) although this is far from certain
(1987, 31). The bust has led him to suggest that the area of origin was Cyrenaica where
the oracle of that god was situated. The bust of Zeus Ammon, however, also appears on
coins of Cassandreia in Macedonia, although the fact that this was a colony whose coin
legends were in Latin may possibly make Cyrenaica the more likely choice.
Sealing referring to the s/a/ia ftrraria at Ostia 103021 - Garrucci (1862, 418)
suggested STAT(ionis) FERR(ariae) FOR(iculorum) OS[t](iensium) while Rostovtzeff
(1900, 10) preferred STAT(ionis) FERR(ariarum) FOR(i)? OST(iensis). Little is known
of the work of the s/a/jo ferraria although Meiggs believes that the 'passage of iron may
- have been under direct imperial control' (1960, 302), citing T. Petronius Priscus who
was procurator Aug('usti) ferrariartim el annonae Os/is (GIL XIV, Suppl. 4459). The
fact that this sealing was found in Ostia is not as strange as it seems since the goods/raw
materials, after being sealed on arrival at Ostia to denote payment of duty, could easily
have been used in the town rather than being transported further afield.
Sealing possibly referring to qualtuorpublica Africae 103051 - Rostovtzeff (1900, 11)

suggested (quattuor) P(ublicorum) A(fricae) AVG(usti) N(ostri) B [1. He believed that
the qua//Nor publica Africae were four areas of porlorium collection in Africa (1904,
402). Others have preferred to see them as four different forms of taxation (Thibault,
81
1888, 44; Guérin, 1900, pascim) Exactly what these were is uncertain, although de Laet
suggested porlorium, vicesima liherlalis, quinta et vicesima i'enalizim mancipiorum and
vicesima hered,taium (1949, 247-54).
There are two further sealings which Rostovtzeff thought referred to the qualluor
ptiblica Africae (1900, bc. ciIi) but which I have entered as Imperial on account of their
uncertain inscriptions ([01 13]; [0118]).
None of the sealings connected with the payment of taxes are double-sided and
none bear the impression of fabric on their reverses Two examples, [0294] and [0301],
do however have the impression of wood grain on their otherwise blank reverses.
The absence of double-sided sealings may show that these scalings were used to
physically seal the goods to which they were attached, thus ensuring that no untaxed
items could be added later to the packages.
Evidence from assodation with findspot
The scalings of the ArIes slalio found in Lyon [0295 - 0301], like the other
scalings from this site, were probably discovered in the Saône, presumably having been
removed from packages as they were unloaded from boats and opened on the quayside.
The scalings would appear to have been attached to the goods at Aries, near the mouth
of the Rhône. This was probably at the time that the goods were transferred from
sea-going ships to river craft (King, 1990, 115) although the application of scalings
would suggest that unloading on to the quayside (as opposed to ship to ship transfer)
82
was necessary. This is because the scalings were formed from molten lead as they were
applied, not pre-moulded, and this would be a rather difficult task on board a ship.
There does not appear to be any written evidence directly connected with the use
of lead scalings for tax collection.
There are no examples dated by the archaeological context in which they were
found.
Conclusions
The evidence provided by these scalings is very slight. They certainly tell us little
concerning the organisation and operation of these agencies. The apparent paucity of
scalings in this category (which after all is considered by many to be one of the main uses
of scalings) may partially be explained by the numbers of Civic scalings which possibly
fulfilled a similar function. It has also been suggested that Imperial/Official scalings were
also used for this purpose (CUppers, 1974, 172; Dembski, 1975, 53), although I do not
support this idea filly. As noted above, it is interesting that there are no double-sided
examples since this may suggest that securing of packages was the purpose, to ensure
that no extra goods were added following the payment of dues.
83
Notes
1. Dissard (1905, 14) merely gave 'in p.....? ' whereas Frere (RIB 2411.39) offered
'inp(...) or in p(...)'.
2. Frere (loc.cit.) points out that 'several types of import, including olives, dates and
certain kinds of wine came from sources round the Mediterranean beyond Gaul'.
84
Chapter 6
PROVINCIAL SEALINGS - THE EVIDENCE FOR THEIR USE
Sealings of this category provide little information in their inscriptions other than
the name of the province.
Some of the inscriptions actually use the word 'province', either as PROV or just
the initial P. These are [0306-03 10], [0312-0315], [0320-0321] and [0323].
The other inscriptions just give the name of the province, although it is debatable
as to whether some of these refer to an actual province or not. The Latin inscription on
[0316] may refer to the province of Pamphylia after its separation from Lycia in the
fourth century but sealings [0317-0319] offer us the ethnic of the name in Greek. This
reference to the people of the area suggests that we are not dealing with a Roman
province but with the Kotvov. This need not necessarily precede the Roman period since
otva continued under the empire, and indeed, where they had not already existed, may
even have been created.
Another sealing [03 11] bears a legend which names the province and adds a
qualiing adjective, Brlla(nnia) Sanc(ta).
Sealing inscribed PROV / PONTI 103201 - This sealing helps to date itself by virtue of
its inscription. The only point at which there was a province called simply 'Pontus' (i.e.
without mention of Bithynia or some identi1'ing adjective associated with the name) was
the period between the AD23Os and the beginning of the fourth century (Mitchell, 1993,
158-9). Thus, assuming that this is not just an abbreviated reference to one of the other
85
provinces referred to above, we can date the probable period of use of this sealing to
within c.80 years.
Sealing inscribed PROV I AEGVPT / JOVIAE 103211 - Like the previous example,
this sealing also helps to date itself by virtue of its inscription. The period of use of this
sealing can be narrowed down even further since the province of Aegyptus lovia existed
only from AD3 14/5 until 325.
The majority of these scalings are one-sided with a swelling on the reverse.
Four of the seatings in this category are double-sided, [0312], [03 14-0315] and
[0323]. The presence of an impression on each side suggests that these sealings were
intended to be viewed from both sides and therefore may have been suspended loosely
from a thread. However, this does not preclude the use of this type of sealing as a
security device (e.g. the thread could have secured a lock or something similar). All four
of these scalings bear similar designs, probably refemng to P(rovincia) M(axima)
C(aesariensis).
Only one example, [0321], bears the impression of fabric on the reverse, while none
show signs of wood grain.
Sealings of Britannia Inferior and Bntannia Superior 10306 - 0310; 03131 - These
two groups of sealings are inscribed PB! and PBS with the image of a bull and a stag
respectively (see chapter 13 on iconography for more information). It is interesting to
86
note that, with one exception, they are always found inside their issuing province. Only
[0310], a PBI sealing, was found in a different province, Britannia Superior. Although it
may seem obvious, these !indspots can at least tell us that the identification of the
province was not placed on the sealing because the item sealed was destined to leave the
province but just to represent the authority by which the item had been sealed.
Sealings of Maxima Caesariensis found in Trier 10314 - 03151 - These scalings,

marked PMC with a central Chi-Rho, are thought to refer to the 4th century province
which was presumably centred on London. A similar sealing [031 2J was found in
Silchester and the provenance of another [03231 is unknown.' The examples from Trier,
if interpreted correctly, are a further indication of the links between Britain and Germany
in the 4th century which I have also noted in chapter 4 ('Official scalings - the evidence
for their use'). These sealings may have been used for sending supplies (or documents?)
to the capital of the Gallic prefecture.
Sealing of Britannia Superior from Combe Down 103091 - This sealing was found in
the area of a building thought to be under procuratorial control. 2 Jones and Mattingly
(1990, 217) suggest that the building may have been connected with stone quarrying and
point out that 'the use of lead seals attached to blocks and columns is paralleled at
Rome'. 3 The scalings from blocks of marble found at Rome, however, tend to be
somewhat larger than our normal type of lead sealing, as can be seen from Appendix I in
the Catalogue.
Sealings of Pontus and Aegyptus lovia from Thrace 10320; 03211 - These scalings,
mentioned above under 'Evidence from impression', provide us with dated scalings
which can be related to individual sites.
The Pontus sealing (probably dating from between AD230 and the early fourth
century) found at Cabyle may have been found in the city or perhaps, considering the
possible official nature of the sealing, in the associated auxiliary fort. For suggested
87
reasons for the presence of scalings from Asia Minor in the Balkans see the section
'Evidence from association with findspot' in chapter 7 ('Civic sealings - the Evidence for
their Use').
The presence in Dimitrovgrad of the sealing from Egypt (dating from

AD3 14/5-325) would appear to suggest official involvement in such long-distance
transportation.4 This is rather interesting, whether the site turns out to be the mansio of
Carassura or the emporium set up by Septimius Severus at Pizus
There are apparently no written sources which have any bearing on the use of
provincial scalings. However, perhaps we should look at the evidence for travel warrants
which gave permission to use the curstis publictis. These were apparently in the gift of
the provincial governor and so, if sealed, may have been provided with provincial lead
scalings. Chapman has extrapolated from Trajan's reply to Pliny (Ep. X. 46) that these
were 'official blanks counterstamped by the emperor in Rome and sent out in batches to
provincial governors for their use' (1978, 59). This is probably correct although it is
debatable whether the source can actually support it. Chapman later says that the tablets
were probably stamped or branded with the emperor's name (op.cit. 60) but perhaps I
should suggest the possibility that imperial scalings were used . Unfortunately there are
no detailed descriptions of these postal warrants and so we are unable to progress
beyond the basic suggestion of some type of lead sealing being used, to denote validation
either by the emperor or by the provincial governor.
88
None of the provincial sealings are from securely dated archaeological contexts,
but see 'Evidence from impressions' above for two dated examples.
Additional example
It should be noted that a sealing entered in the Miscellaneous category has been
recognised at a late stage as being Provincial [1691]. It has therefore been impossible to
include it in its rightful place in the calculations in chapter 12 ('Categories of sealings
found in the provinces') or to renumber the illustration. The sealing in question is
inscribed 11AA / HHEI / POC which is an abbreviated form of fl&xxtoç Hitctpoç. This
refers to the fourth century province known in Latin as Epirus Vetus which was in the
diocese of Moesia (later in Macedonia). The sealing was found on an unknown site in
Pannonia.
Conclusions
With regard to the sealings which have PROV or P in the inscription, the very
fact that the word 'province' has been referred to suggests that these sealings were
produced officially and therefore were presumably issued, for whatever purpose, on
behalf of the governor and his officia.
89
Notes
I. Since we have no indication of where this sealing was discoved, it may refer to
P(rovincia) M(auretania) C(aesariensis) in Africa, rather than Maxima Caesariensis. In
fact, the abbreviation PMC usually refers to the African province (ILS 781)
2. RIB 179 from the site refers to a freedman who was an assistant of the procurator.
3. Jones & Mattingly actually refer to 'a large number of lead seals' having been found
on the site, although they do not give a reference for this and I have been unable to
locate any further information. If there are any other sealings from the site then they are
probably anepigraphic since EE IV (1881), 209, no.707 does not mention them in
connection with the Britannia Superior example.
4. Unless, of course, it sealed expensive exotic goods from further afield which would
have covered the otherwise prohibitive transport costs.
90
Chapter 7
CIVIC SEALINGS - TIlE EVIDENCE FOR ThEIR USE
There is very little information as to actual use in any of the impressions on Civic
sealings. All of the examples which I have collected are epigraphic since the presence of
the name of a city is the criterion for identifying examples in this category. It is almost
certainly true that some of the anepigraphic sealings in the Miscellaneous category,
especially those showing figures from mythology, are actually Civic and bear the image
of the tutelary deity of the city. For those sealings that we have here, however, the most
important piece of information to be included by the engraver was considered to be the
origin (or issuing authority?) of the sealing. Therefore, the only possible way in which we
can attempt to use these sealings is to examine the stated place of origin, and then try to
suggest a reason for that city to be using sealings. However this can be highly subjective.
The form of reference to the city may also be of some interest, since some sealings use
the city name in the nominative whereas others use the genitive plural of the ethnic.
Various abbreviations of these two most often used forms of reference render the case
and type of noun found on many inscriptions uncertain.
The following sealings definitely bear the city name in the nominative:
[03281 (Ephesus); [0330] (Rusicade); [0348-0349] (Hypaepa); [0357] (Metropolis);
[0364] (Rusicade); [0376] (Cyme); [0387] (Hypaepa); [0388] (Cyme); [0389]
(Rusicade).
The following are best seen as the city name in the nominative, although an
abbreviated form of the ethnic is possible:
[0325] (Smyrna); [03311 (Smyrna); [0339-0347] (Smyrna); [0354] (Coloe); [03 58-0360]
(Smyrna); [0380-0385](Smyrna).
91
The following sealings definitely bear the genitive plural of the city's ethnic:
[0333-0336] (Ephesus); [0337] (Laodicea ad Lycum), [0338] (Magnesia ad
Maeandrum); [0350] (Hypaepa); [0351] (Magnesia ad Maeandrum); [0353] (Ephesus);
[0355-0356] (Coloe); [0368] (?Tyre); [0369-0372] (Ephesus); [0374-0375] (Ephesus);
[0379] (Philippopolis).
The following examples could be either name or ethnic, the case of either of
which is unknown. This uncertainty is for various reasons including abbreviation:
[0329] (Tavium), [0352] (?Tyre); [0361] (?Adramyttium); [0362] (?Antandrus); [0365]
(?Savaria); [0366] (Pictavium); [03731 (unknown city in Galatia); [0377-0378]
(Pergamum); [0386] (Apollonia Mordiaeum).
Some inscriptions furnish their city with a title. The examples of this are [0324]
which apparently refers to R(es) P(ublica) G(Ievensium) [.] A(...), [03261 which may
bear the inscription C(ivitas) COR(i)EL(tauvorum), [0332] which has TYPIWN
MH[tpj(oiroXtc) and [0367] with COL(onia) BER(ytus/ytensis). Of course, any of these
expanded titles could equally be in the genitive.
There are three related sealings, [0329], [0373] and [0386], which bear the name
of the province, Galatia, in addition to the name of the cities.
Another example, [0365], appears to bear the legend FORTVNA [S]AB[a](ria),

although the expansion is rather doubtful and the sealing may not even be Civic.
The majority of the sealings in this category are single-sided with the usual
swelling, of varying shape, on the reverse.
92
Three are double-sided, [0326], [0365] and [0369], which means that they
probably hung loose on a cord, although they could still have performed the flrnction of
actually securing the item to which they were attached.
Two scalings from this category, [0368] and [0380], show evidence of the
impression of fabric on their reverses, while another two bear the imprint of wood grain,
[0324] and [0332]
It is of great interest that one example, [0332], still bore fragments of its binding
thread, described as purple silk thread, in its thread-hole when seen by Paul Dissard.
These have since disappeared but Dissard's notes are referred to by Turcan (1987, 29,
no.28).
Sealing of Smyrna found at Ickham 103251 - This sealing has already been the subject
of an article which examines closely its links with its findspot (Still, 1995, 347-56). I only
mention it here to highlight the distance which these sealings can travel.
Sealing of Tavium found in Trier and seating of Apollonia in Galatia found in

Thrace, 103291 and 103861 respectively - The sealing from Tavium in Galatia is marked
TAO[.]I / fAAA and was found near the Mosel in Trier. Like the Smyrna example
above [0325], it demonstrates the distance over which some scalings are known to have
travelled.' In his publication of this sealing, Binsfeld suggested that it had secured a
package of scarlet dye or scarlet-dyed wool, the local speciality (1988, 14 & 16).
However, in an article currently in preparation, I have pointed out that there is a
previously unpublished sealing from another Galatian city which is stylistically very
closely linked [0386]. 1 believe that this sealing, marked AflOA I TAAA, must have
come from Apollonia Mordiaeum. This city was inside Galatia until the reign of
93
Diocletian and is the only likely candidate. The article examines several possible reasons
for why the matrices, almost certainly engraved by the same person or workshop, should
have referred to cities so far apart (c.400km as the crow flies or c.550km by road). 2 The
favoured answer is that the issuing and use of the original matrices were probably
controlled by some form of central administration, presumably the provincial governor
and his officia. 3 The idea that the scalings were used to mark local produce with labels of
origin can probably be set aside now, since it is difficult to see why any organisation at a
provincial level would have been interested in arranging this, purely for the glory of the
cities.
Sealings from cities in Asia Minor found in Moesia Inferior and Thrace: Apollonia
Mordiaeum f0386J; Coloc 10354 - 03561; Cyme 103761; Ephesus 10333 - 0336; 0353;
0369 - 0372; 0374 - 0375J; Hypaepa 10348 - 0350; 03871; Laodicea (?ad Lycum)
f0337J; Magnesia ad Maeandnim 10338; 03511; Metropolis 103571; Pergamum
10377 -0378J; Smyrna 10339 - 0347; 0358 - 0360; 0380 - 0385J - The proximity of
Asia Minor to the Balkans is certainly a major reason for the location of these scalings
but can they tell us any more? The main theme of past research conducted on the
movement of goods between Anatolia and the Balkans has been that Anatolia gave more
than it received, with the provisioning of the Danubian garrisons being the main reason
for this (Gren, 1941, passim). Mitchell, however, has added to this that the supply of the
Balkan army must have been organised by the state (1993, 251). This is because the task
was too important to leave at the mercy of private enterprise and also since the cost of
land transport for this quantity of grain would have been prohibitive to private merchants
(Mitchell, 1983, 139; 1993, bc. cii). This, however, does not help us since military
supplies would surely have been exempt from import duties throughout the period of the
empire4 and yet it seems likely that these Civic scalings were attached to denote the
payment of dues. Perhaps they were fixed to items which were taking advantage of the
supply-trains, in the same way that La Graufesenque samian ware was possibly
transported from the kiln-sites as part of the system of pack-animals which carried silver
from the mines of La Rabasse.5
94
Boon (1991, 318) sees the scalings from Smyrna (and presumably those from
Ephesus) as being connected with Julian's restoration of revenues to certain cities (Amm.
Marc. xxv. 4.15). This would certainly fit in with the general date (i.e. mid- to late fourth
century) assumed for these examples. The reason for the similarity of the dies could lie in
the fact, implied in Ammianus' reference to Julian, that the organisation of new civic
dues was subject to the permission of the emperor (de Laet, 1949, 46 1-2 citing Dig.
XXXIX, 4, 10 and Cod. Just. IV, 62, 1-3). For this late period, it was the comes
sacrarum largiliomim who took care of the details following the approval of the
emperor (de Laet, op. cii., 462 citing GIL III, 7152).
Boon's other suggestion (bc. cii.) is that the Smyrna scalings reveal the presence
of a branch of the imperial customs-system for Asia (Quadragesima poriuum Asiae)
which was based at Ephesus. This could also explain the similarity of the design of the
matrices being used. If Smyrna, and other cities in Asia, were acting as subsidiary
collection-centres then it is highly likely that the matrices for this would also have been
supplied by the central administration at Ephesus. 6 The drawback, however, is that we do
not have any examples which are from ostensibly earlier periods.
Unfortunately, the first century AD customs law discovered in Ephesus dates

from a much earlier period than our scalings which are apparently fourth century
(Engelmann & Knibbe, 1989, passim). It is a similar story for the late first or early
second century inscription from Smyrna which referred to the collection of dues for the
crossing of the Hermus (IGRRP IV, 1427).
95
Unfortunately, there are no examples which can be dated by the archaeological

context in which they were found.
Conclusions
While the general tone of this chapter has been to view these sealings as being
normally connected with the collection of customs dues, it is by no means impossible that
some were used as labels to advertise the place of origin, perhaps guaranteeing quality.
This is, however, very difficult to prove, whereas disproving it for some sealings is
considerably easier, as shown above in the section on Galatian scalings
96
Notes
1 The goods themselves may, of course, have travelled much further, only receiving the
sealing as they passed through Tavium.
2. The presence of a similar, damaged, sealing [0373] from an apparently different but
unidentified Galatian city only serves to confirm the province-wide basis of these
sealings.
3. In later times this would apparently have been organised by the comes sacrarum
largilionum (de Laet, 1949, 462 citing CJL III, 7152). Despite the official involvement,
the revenue collected on goods entering these cities was actually destined for the cities'
coffers.
4. Dig. XXXIX, 4, 9, 7 tells us this (see chapter 8 entitled 'Military Scalings - the
Evidence for their Use'.
5. Middleton (1980, 190) says that 'Given the scale of the workings and the mineral
veins exploited, an official and not improbably military interest is certain ... It would have
been a simple matter to tack on to this existing transport facility the mule-transport
system from La Graufesenque.'. He does not make it clear why a private mule train
would require the presence of an official one, unless of course the private traders simply
took advantage of the well-maintained road.
6. This sounds rather similar to the state of affairs suggested for earlier in the empire in
Galatia - cf. note 3 above.
97
Chapter 8
MILITARY SEALINGS - TIlE EVIDENCE FOR ThEIR USE
This chapter will cover all types of military scalings but will of course make
reference to the individual categories, where applicable.
Sealings bearing evidence of rank of supervising officer ILegionary: 0406; 0409;

0410 - 0411; 0414 - 0415. AIae: 0444 - 0450; 0453. Cohorts: 0464; 0473; 0477 -
0479; 0483 - 0486; 0490; 0495; 0509; 05151 - Many military scalings bear letters which
are presumably the initials of the officer's tria nomina, but not all refer to his actual rank.
All of the legionary and auxiliary cohort scalings mentioned above bear the symbol '7' or
a variant of this Most of the scalings of alae have the initial D after the initials of the
man's name, although see below for the two exceptions to this.
Tomlin has provided important notes on the meaning of the centunal symbol '7'
on the legionary and cohort scalings (Brifannia xxii (1991), 298, note 22). I-Ic says that
since the examples with initials alone must refer to centurions and not centuries, this
therefore shows that the symbol should be read as 'centurion' not 'century'. Tomlin is
probably correct but he does seem to be placing a lot of faith in the intelligence and
comprehension of the die-cutters. There seems little difference to me between the
implicit understanding involved in '(the seal) of Aelius Cominus, centurion' and '(the seal
of the century)' of Aelius Cominus, to use Tomlin's examples. I believe that his reference
to documentary evidence which reveals that '7' placed before the name means 'century',
but when placed after means 'centurion' provides better proof of his theory, a1thougJi
even this accepted use is still open to the blundering of the die-cutters (RUR, pacslm;
Tab. Vindol. 3.A8 & 10; B2 & 13). Tomlin also points out that only our [0464] and
[0414 - 0415] indicate that the centurion's name is in the genitive.
98
There has been little discussion as to the use of the letter D for decurion,
although we can probably assume that it too was intended to be in the genitive. In one
case the letter appears before the initials [0446] but we should obviously not extrapolate
from the centurion theory since the decurion was in command of a Itirma not a decuria.'
Another example, probably from the A/a Sabiniana not Sehosiana as it has previously
been published, has the abbreviation DEC following the name VAL [0444]. This has
been taken as decurio (or perhaps more correctly decurionis) but could quite possibly be
the cognomen Dec(-imus or imi) without any indication of rank. On the other hand,
when an individual decurion is being referred to on stone or in papyri, the initial D seems
never to be used, the most abbreviated form apparently being DEC, usually placed after
the name.
It may seem obvious but the presence of the letter D on these sealings appears to
indicate that the matrix was used by the decurion himself, since a general reference to the
Itirma would surely have involved the use of the much more common formula '(urma of
so-and-so' which appears in inscriptions on instrrnnenla domesilca and in papyri (but
only once in Britain on stone - RIB 1445). This is variously abbreviated as I on
ins!riimenta and T or, less regularly, TU on papyri. The absence of I and the presence of
D on these sealings of alae may therefore be seen to support Tomlin's theory of
reference to centurions rather than centuries on the sealings of auxiliary cohorts.
It should be noted that the initials on the reverses of many scalings belonging to
cohortes equilalae end in D [0476; 0483 - 0484 (6 ex.); 0487 - 0488 (6 examples); 0491
-0496 (33 ex.); 0498 - 0499 (10 cx.); 0504; 0513 - 0514, 05171. 2 It would seem likely
that on many of these occasions the D stands for decurio and that the work was being
carried out by members of one of the lurmae belonging to the unit. We should however
be aware of the fact that some may just give a name without rank since some inscriptions
on the reverses of scalings belonging to cohories pedila(ae also end in D. Thus we
cannot use the presence of D as a final letter to tdll us whether a little-known unit was
equutala or not, eg. [0506].
99
Some sealings bear the impressions of intaglios on their reverses instead of initials
[Legionary: 0395; 0416; 0418; 0438. Alae: 0451. Cohorts: 0459 - 0461; 0503, 0508 -
0509. Ped,ies Singulares Consiilaris 0530]. Unfortunately this can tell us little since any
soldier may have owned such a ring, not just an officer. By analogy with the matrices
naming centurions and decurions, however, we may be able to say that these intaglios
belong to men of the same rank. The reasons for the use of such intaglios are examined
in chapter 2 ('Typology of Lead Sealings') under Type no.1.
It comes as no surprise that centurions and decurions are named on these sealings
since they were in charge of one of the main administrative divisions within the army.
Working parties were usually under their control3 and the century or lurma (identified by
the name of the centurion or decurion, i.e. the century/Iurma of so-and-so) was often
used in inscriptions to indicate ownership of items, with or without the added name of a
soldier.4
Sealings bearing the inscription METALLA 104651 - The only military sealings which
may give us any more information than unit or personal name/rank are a small group of
five examples belonging to cohors II Nerviorrim [0465]. The reverse is marked with a
heavily ligatured inscription which may read meial(la) or melal(h,m). This has been
interpreted as referring to lead mines (or their product) which were near the unit's 3rd
century base of Whitley Castle (Richmond, 1936, l09). It is rather strange that this
inscription appears in the place where most other scalings apparently bear only names
and ranks but it appears to have been read correctly. It may even be that some of the
initials on other scalings provide us with more information than has previously been
thought, e g. four scalings of cohors VI Raeiorum which have the inscription CAD on
their reverses [0472], nine examples belonging to cohors VII Thracum marked MV /
CAD [0494; 0498; 05171 with another possible example CAD / MV [07721 (but see note
2). Due to the presence of zinc in the general area in which these scalings probably
originated, there is a possibility that this could refer to one of its ores, cadinia, although
there is no proof that it was ever mined in Roman Britain 6 It should be noted, however,
100
that if this is correct then it is even more specific than the above reference to metal or
mines.
Sealings bearing the inscription EXP 10391; ?0392; ?0400 - ?0401; 0403; 0425;
0427 - 0428; 0443; ?0524 - 05261 - Thirteen of the scalings belonging to !egio II
Augusta [0391 (7 examples); ?392 (4 examples); ?0400 - ?0401) and one of Iegio VI
[0403] have the inscription EXP (or a blundered version) on their reverse which is
usually translated as exp(edivii) - (Legion ??) dispatched (this). 7 However, EXP is also
the sole legend on 9 examples from Lyon [0425; 0427 (7 examples); 0428] and I from
Syria [0443]. Turcan has interpreted this on [0425] as exp(ediiio). He asserts that his
example would have been attached to packages destined for the army while on campaign.
He even suggests that parcels associated with these scalings found at Lyon contained
food intended for the army of Septimius Severus on its way to Britain in 208. It is just
possible that EXP on these single-sided sealings may stand for something other than
exp (edi vii) since there is no subject for the verb, however, one would expect similar
scalings to be found in Britain where instead we find imperial scalings relating to the
same campaign [0021 - 0022]. If Turcan is correct, could the EXP scalings from Lyon
represent supplies packaged by the military units themselves using surplus food already in
the granaries of their bases as opposed to the imperial scalings in Britain being brand new
supplies straight from the fields?
Can Turcan's translation be used for the British EXP examples? They were all
found at Brough under Stainmore which seems to have been a centre for repackaging
and redirecting goods It is usually thought that the goods are being sent from further
north and are travelling south (Richmond, 1936, 108) and if we stick to this then
expedi vii must be favoured. These particular examples, however, may have been
attached to packages sent north to a vexillation on campaign in the area of the wall (or
even further north if these too date to Severus' expedition) in which case either meaning
would be possible. They may even indicate that they were being sent by a vexillation on
101
campaign. It cannot be explained exactly why they were unpacked here at Brough but
then this is a question which applies to the whole collection from that site.
All three sealings belonging to the beneficiarii consularis [0524 - 0526] have EX
inscribed on their reverse. Frere and Tomlin have both taken this to be analagous with
EXP which seems sensible (RIB 24 11.246 & 267; Brifan,,ia xxii (1991), 302, no.33).
Sealing bearing the inscription AR I LEG I XIV 104261 - Dissard (1905, no.45)
thought that AR might be A R(alionihus) - ' accountant', but this must surely be wrong
since the a ralionibus was a civil official connected with the treasury. I have not been
able to find any reference to this title being used in a military context, the post usually
being filled by either the signfer or a librarius. If Dissard had instead meant 'from the
accounts' then, on limited evidence from papyri,' 0 this should probably be ex
rallonibus.It is much more likely to be the initials of a personal name, as suggested in
('IL XIII 10029.47.
More information on the way in which the scalings were formed and attached can
be found in chapter 2. Here we shall try to use the evidence of different forms in order to
assess use.
The majority of military sealings are double-sided, perhaps on account of the

amount of information to be applied. The single-sided sealings are as follows:
Legionaty - Britannia [0397; 0404]; Dalmatia [0424], Gallia Lugdunensis [0425 - 0426,
0427(7 examples); 0429 - 0437]; Moesia Inferior [0439 - 0442] and Syria
[0443].
Cohorts - Britannia [0457 - 0458; 0474 - 0475; 0487 - 0488 (6 examples); 0500, 0516];
102
Gallia Lugdunensis [0519; 0520 (2 examples), 0521 (4 examples); 0522 (2
examples); 05231.
This shows that all of the Legionary sealings from Moesia Inferior, the single
example from Syria and all but one of the sealings from Gallia Lugdunensis are
single-sided. This is in stark contrast to those from Britain where the single-sided
sealings are a tiny minority. Even the scalings of cohorts reinforce this picture, when
viewed in relation to their totals. This could suggest a more rigid and bureaucratic
system in operation in Britain than elsewhere which in turn could relate to a special use
which had to be closely monitored.
The following scalings bear the accidental impression of wood grain on their
reverses:
[0520 (2 examples); 0521 (4 examples); 0523].
The following sealings bear other marks:
[0433] - impression of string on reverse; [0436] - impression of coarse fabric on

reverse; [04571 - nail hole.
All of these marks reveal to some extent the nature of the material to which they
were originally attached. The other single-sided sealings may have been affixed to similar
items but, for reasons of alloy temperature, have not picked up the impression.
The possible uses of double-sided scalings have already been examined under this
heading in chapter 3 ('Imperial Scalings - the Evidence for their Use'). However, military
scalings can tell us one thing about double-sided scalings and that concerns the surviving
white string that can be seen on several examples of auxiliary scalings from Brough
under Stainmore. In particular, one of the sealings listed in the catalogue as [0499] is
103
split along the thread hole and clearly shows the white string running through. This string
apparently displays a small knot which may be the reason for the positioning of the
sealing or may just be an earlier repair which has fortuitously been included within the
lead. If its inclusion was deliberate and typical of others then it would totally refute the
idea of double-sided sealings as dangling labels. However, the value of the suggested
materials sealed by the sealings from Brough would probably have rendered the label
theory redundant anyway, even without the added evidence of this knotted string.
The single-sided scalings of legio Xl Claudia found at Silistra (0439 - 0442),

marked k(astrum) leg(ionis) Xl, were found in the canahae of that legion's fortress
(Culicâ, 1971, 193-7). Their reverses seem to have splayed out when the scalings were
struck with the die or even when they were actually formed. They can be likened to
rivets, in that they may have been pushed through a hole (in the fabric of a sack or in
hides) and then hammered with the die on a hard surface to give the impression on the
obverse and a non-return foot to the lower haffi The presence of the scalings close to the
issuing legion's own fortress is probably best seen as the result of goods being sent back
to base by a detachment and the scalings then being thrown away or disseminated in
some other way to the population of the canabae.
Brough under Stain more 10390 - 0403; 0444; 0455 - 0499; 0524 - 05251 - The largest
known collection of military lead scalings is that from Brough under Stainniore, with 131
examples from cohorts [0455 - 0499], 25 legionary examples [0390 - 0403], 2 examples
from beneficiarii consularis [0524 - 0525] and one example from an ala [0444]. The
disposition of the various units represented makes the 3rd century the most likely
occasion for the accumulation of these scalings in the rubbish dump near the fort
(Richmond, 1936, 114)11 As mentioned briefly above, Richmond saw the site as a centre
where items received from various places (implicitly further north) were unpacked (thus
104
giving us the sealings) and then presumably forwarded in bulk to the 'central
administration' (1936, 108).12
Hassall has suggested that the unit listed at Brough in the Nolilia Digni/aizirn, the
numerus Directorum, was connected with this work (1976, 111). He points out that the
noun direcior is derived from the verb dingo (actually dengo but not usually
distinguished) meaning 'to aim' or 'shoot' and says 'it seems to me that it is just
conceivable that the meaning intended is that of "direct", in the sense of "send" or
"dispatch".'. This is an interesting idea, although the objects of the verb are usually
things like ships, horses or the course of a journey.
The main problem that I can see is one of date. The sealings are said to be 3rd
century (and there is no evidence of any later examples) whereas the Nolilia list is from
the late 4th century. Furthermore the Notilia unit is presumably one of those which Frere
covers when he says 'the rest have ... fancy names typical of troops raised during, and
perhaps late in, the 4th century...' (1987, 219). However, the 3rd century unit could have
been renamed, or replaced by a new unit conducting similar work but in a period (i.e. the
4th century) when lead scalings were no longer used for this particular purpose.
Janet prefers to translate directi as 'sharp-shooters', echoing Frere by saying that

'such a title would be typical of the fourth century, when the numerus was presumably
formed' (1994, 70).
It should be pointed out that dingo (in its correct form) can also be interpreted
as 'to set in a straight line' and can thus mean 'to build regularly', suggesting to me the
possibility of a unit of engineers.
On the subject of names, while not wishing to suggest that the nature of the
activity on the site influenced its name (or vice-versa), it may have crossed the mind of
the soldiers stationed there that the Latin name for Brough under Stainmore, Verteris,
105
shares its first syllable with the verb verb, one meaning of which is 'to turn towards' or
'to direct towards'.
Although touched upon above, we should examine in more detail the question of
what was being sealed and then sent to Brough. The suggested reference of melal(Ia) to
lead mines raises the possible alternative of its associated product, silver. The higher
value of this material could also explain the meticulous attention to sealing, often with an
indication of the person responsible. Boon describes the Aiston lead field as being
'silver-rich', citing the fact that it supplied the Carlisle mint under Stephen and David of
Scotland (1991, 319). Tylecote's general tone is that British lead was not sufficiently
argentiferous to make its recovery viable, but close examination of the figures for
Cumberland and DurhamfNorthumberlandlWestmorland in his Table I suggests that in
1847 and 1923 cupellation was still economical (Tylecote, 1964, 26). Since the higher
veins of lead ore often bear the most silver, it is quite possible that in Roman times the
Aiston field was of considerable importance.' 3 In addition, if the only reason for the
activity here was to further the production of lead, it would seem ridiculous that these
lead scalings were just thrown away in such quantities. '
The word mekilla can also refer to quarries. Although there is no evidence of any
major Roman stone quarries in the immediate area this is typical of the whole country
and should not be used to rule this out as a possibility. There would, however, seem to
be little point in transporting the local sandstone any great distance and, additionally, the
sealings from Brough are not of the type normally associated with blocks of stone.'5
Therefore, for these examples, we can probably return to the idea of lead ore,
argentiferous or otherwise, being sent to Brough. Richmond doubted that all of the
sealings at Brough were connected with mining on account of the number of units and
the different styles of impression used on the scalings although he could not offer any
alternatives (1936, 110-111 & 115). It is difficult to assess whether or not this is correct
since we do not know where the unit represented by the largest group of sealings., cohors
106
VII Thracurn, was based, or where the presumed vexillation of legio II Aiigusia was
stationed. The quantity of sealings belonging to cohors Li! Thracum led Eric Birley to
suggest that this was the unit actually based at Brough (1958, 49)16 I believe that this is
well worth considering since it appears from the Iegio XI ('lazidia examples cited above
that detachments would send back sealed consignments to their bases. I also think that
these Thracians could have been sending in lead ore, since it is also found in this area.
If it is true that the majority of the units represented in the sealings at Brough
were sending in consignments of lead ore then what happened to it here? Richmond, as I
have said above, considered that the material from various units was probably
'concentrated' at Brough for delivery to the 'central administration' (1936, 108). This
seems sensible but one wonders why the material needed to be unpacked. If it were
already in man-portable sacks or boxes surely it would have been preferable to have kept
it in these rather than emptying them into a large wagon? There are several possible
explanations such as: the lead ore was being delivered here in order to be smelled into
lead and/or desilverised; a procuratorial official was present who had to receive and
weigh the ore. It is of course possible that the product had already been separated into
lead and silver and that we are seeing the delivery of one or both of these metals. This
supports the idea of a procuratorial official being stationed at Brough (Richmond, 1936,
109). It would have been his job to take receipt of the ingots and then despatch them
under his own department's seal.' 7 However, although none of these occurrences need be
ruled out, the simplest reason for the sealings being removed here is that it was a 'simple'
collection site and the boxes, or perhaps the donkey-pariniers, were the property of their
original unit and needed to be returned in order that they could be used again.
There is some evidence for other activity on the site. It has been said that
brooches were made at Brough (Birley, 1958, 42) but there are other signs of
bronze-working on the site. It could be that some of the other items had been collected
as scrap for the brooches but if so why were they discarded? For example, two bronze
'flawed castings' (a terret and a 'lock bolt' which may possibly be the handle of a folding
107
hare and hound knife) and one bronze trial piece (of an enamelled seal-box lid) were
discovered in the dump here (Collingwood, 1931, 84, nos.25-7). Collingwood described
the seal-box lid by saying 'the work is rough and the casting bad, and the object looks
like a trial piece which has never been worked up into a complete seal-box in addition,
J.D. Cowen, who referred to the site as an 'undoubted centre of bronze-working',
published a Celtic sword pommel from Brough which he suggested had been taken there
as scrap to be melted down (1937, 67-7 1)
There is another item from Brough which at first I believed may have provided
more evidence for the manufacture of bronze pieces, especially seal-boxes, on the site.
This was a so-called imperial lead sealing (British Museum mv. no. 74.12-28.79) which
is actually the lid of a seal-box. This piece of lead, which is not entered in my catalogue,
is in the shape of a piriform seal-box, inside which it is said to have been formed, and
bears a bust identified as Caracalla by Richmond (1936, 120). Martin Henig describes the
impression as 'male head right (?bearded) - ?caduceus behind' and says that it probably
represents Septimius Severus (1974, 84 & no.8 19). I believe that the caduceus is certain
but it is difficult to decide whether or not the 'beard' is fortuitous. The main reason why
this cannot be a lead sealing formed inside a pear-shaped seal-box is that it possesses a
decorated terminal knob at the end opposite the hinge. The hollow space for the terminal
knob is sometimes included on the inside of the box but there is no means by which the
lead therein could have been decorated with a roundel. In addition the main 'it pression'
is itself pinform which is not a shape usually associated with intaglios or dies. Quite apart
from this the object is nothing like the normal type of lead sealing since it is flat with no
thread hole. Therefore this object at first appeared to be a lead trial piece for the lid of a
seal-box. The presence of this piece, and the other seal-box lid mentioned above, may
have suggested that the manufacture of seal-boxes, among other items, was taking place
on this site. Unfortunately this theory is no longer tenable since there are two other
lead/pewter seal-box lids with rather similar busts with caducei (identified as Mercury)
which were found in Lyon, along with another complete, hinged, lead seal-box showing a
figure of Victory moving r. (Dissard, 1905, 299-300, nos.1946-1948) This latter
108
example in particular infers that these lead seal-boxes were actually being used and were
not just trial pieces. Therefore the Brough seal-box lid, although now paralleled and
hopefully interpreted correctly, cannot be taken as anything other than general rubbish on
the site.
Despite this, we could perhaps suggest that, together with lead ore, raw materials
for other metal-working on the site were among the items sealed. Copper is not
apparently found in this area but in note 5 we saw that fluorspar was found at Whitley
Castle. Perhaps this was being sent to Brough for the manufacture of enamel to be used
in decorative bronze-work?
Having said this, however, we should not forget that bronze-working in some
form would have taken place on many Roman military sites and so this evidence cannot
be used to claim any special function for the site.
In fact, we must be careful even in drawing any inferences from the presence of
the sealings on this site. After all, they have only come to light by accident (erosion by
the river) and there could quite possibly be similar collections still lying undiscovered
near other forts.
Sealings of cohors VGa!!orum at South Shields (0508- 0514J - These 7 scalings were
all found in the fort at South Shields. The unit is definitely attested here in AD213
(Brlla,inia xvi (1985), 325-6) and again in AD222/3 (RiB 1060). Dore and Gillam
interpreted the scalings as belonging to a period prior to the latter date (the earlier
inscription had not been found at the time of their writing), partly on the assumption that
sealings are not found in the fort where the issuing unit was based (1979, 67-8).' They
believed that the unit was stationed at Cramond during the Severan campaigns and sent
the scalings from there. Theoretically this could still hold true even in the light of the
earlier inscription. However, their assumption concerning the findspots of sealings is
logical but unfortunately incorrect.° This could mean that the scalings are later than
109
AD222/3. On the other hand, I believe that the scalings may still date from the Severan
period but that they serve to resurrect the idea that coho,s V Gal/orum was split between
South Shields and Cramond (ibid.). 2 It would be quite possible to envisage the bulk of
the unit at Cramond (or elsewhere) sending items back to the small detachment left in
charge of the granaries. It may therefore be possible to use the scalings as partial
evidence for the same theory that Dore and Gillam were using them in order to reject.
In fact, the seventh sealing [0508], found in 1990, was located in the demolition
deposits of a building dating from the time of the conversion to a supply-base (Brila,mia
xxii (1991), 232; xxiii (1992), 322, no.42). This sealing is rather confusingly said to
suggest that the unit was 'already associated with the fort in this earliest phase of the
supply-base' although the building is described as having had a short life (ibid., 1991).
Sealings of military units found at Lyon 10425 - 0437 0518 - 0523J - There are 19
legionary scalings and ii scalings of auxiliary cohorts, all apparently found in the Saône,
where they had presumably been removed from packages as they were unloaded from
boats. This in itself raises the question of why would the army immediately unwrap
goods on the quayside? There is no obvious reason, regardless of whether the packages
were to be stored in a waterside warehouse or immediately taken to their destination,
unless of course there was no other way of telling what the contents were! If this were
true it would be surprising, considering the use of lead labels, military and otherwise, in
Austria, Germany and Britain23 and the writing in ink (liftili picti) on amphorae. Another
reason could be that the packages were so large that they had to be opened in order that
the contents could be man-handled once on the quayside. This is quite possible since the
actual unloading of large items would have been conducted using a crane of some
description. The possibility that much of the cargo had already been transferred from
large ships to smaller river-going craft at ArIes (King, 1990, 115) does not hinder this
idea and we should not expect to see the goods being repackaged at that point.
110
The legions represented here are the XIV Gemina [0426], XX Valeria Vicirix
[0434 - 0435], XXII Primigenira [0429 - 0431; 70432; ?0437] and the XXX Ulpia Victrix
[0436]. There are also 9 other sealings which are categorised as legionary but which have
been dealt with above under the heading 'Sealings bearing the inscription EXP'.
Unfortunately we cannot date these sealings by their presence here since the
associated imperial sealings apparently cover the period from the 1St to the 5th century.
In addition, the disposition of the legions helps little since the small number of sealings
means that we could often be looking at individual packages sent hundreds of miles on
special occasions, not localised everyday use. For example, the stationing of Iegio XIV
Gemina in Pannonia from AD9213 need not necessarily mean that the sealing dates to a
period earlier than that. It is interesting to note, however, that detachments of legio XXII
Primigenia and legio XAX Ulpia Victrix replaced the urban cohort of Lyon after
February 197 (Turcan, 1987, 25). This may be a better reason for the presence of their
sealings than Grenier's opinion that they showed Lyon to be the 'base intermédiare des
deux armées du Rhin' on the route to Italy (1934, 653).
The sealings of (presumably auxiliary) cohorts [0518 - 0523) are all rather vague
in their representation of the unit to which they refer. They are certainly no match for the
detailed information concerning names and ranks that we are given in the British
examples. This may possibly be used to suggest that the Lyon sealings were therefore
sealing packages of less importance or intrinsic value. The double-sided sealing [0518]
bearing the inscription COuP on the obverse and CME on the reverse is one of the only
scalings from Lyon to belong to our Type no.1, the common military, imperial and
official style of double-sided sealing
Virtually all of these military sealings from Lyon are single-sided which
immediately renders them different to those in Britain. Even the two scalings of legio XX
Valeria Victrix differ in this respect from the six of that legion found in Britain.
Ill
Sealings of military units found at Carlisle 10404 - 0409A; 05271 - Ian Caruana has
recognised the importance of his group of scalings as one of the few to have been
excavated from stratified contexts (forthcoming, 'lead objects'). This stratigraphy has
enabled him to make some valuable comments on the find spots and the association of
sealings belonging to different units. His most important point is that 'they have a spatial
distribution which stands in contrast to most other classes of object. All the scalings
come from within the forts and none were found in dumped material outside the forts or
associated with buildings beyond the ramparts. In this their spatial patterning most
closely resembles that of some small iron objects or personalia where it is arguable that
casual loss is involved rather than any process of recovery for re-use or systematic
rubbish disposal. If this is so, it implies that the scalings are likely to be close to their
point of loss and unlikely to be residual.'. It is rather difficult to understand exactly how
lead scalings can be 'lost' in the same way as iron objects, since they were not in
constant use as tools or jewellery. It may be a question of the scalings being removed
from packages and then thrown away on the ground rather than collected as rubbish or
being cut from packages and falling through gaps in floorboards before they could be
collected as rubbish. The fact, however, that iione were found in extra-mural rubbish
dumps would tend to suggest that either the men opening the packages were very
lazy/careless or that Caruana is wrong to say that they were not being collected for
re-use (i.e. melting down). In the past it has always been assumed that lead scalings were
usually thrown away, the numerous examples found at docksides being the most
convincing evidence. Perhaps, however, some were occasionally kept as scrap for the
production of new scalings.
There is only one apparent literary reference to a military lead sealing, 'The Acts
of Maximilian' (ACM, no.17) referring to events in Numidia in AD295. I have already
commented elsewhere on this and have stated my disbelief that any of the scalings known
112
to me were used as bullae/identity tags around the necks of recruits (Still, 1993, 406-7).
However this is the place for a filler examination of the evidence.
The matter is somewhat clouded by Roy Davies saying that the signaculum was
'a piece of lead with a seal and with the man's name, which he carried fastened on a cord
round his neck .., rather like a modern identity disk' (1989, 14). This sounds rather like
our auxiliary sealings from Brough, but there are problems. First of all, the presence of a
seal on the lead is never mentioned in the original text, although it could be argued that
the very word signaculum suggests that a seal was indeed used. The relevant part of the
text says:
Dion ad officium dlxii: Signetur. cumque reluctaret, respondit: Non accipio signaculum saeculi;
ci si signaveris, rumpo illud, quia nihul valet, ego Christianus sum, non licet mihi plumbum cob portare
post signum salutare Domini mel lesu Christi (jul Dei ivi...
More importantly, the man's name is never said to be on the lead. Davies has
inferred this from his interpretation of a phrase later in the passage:
Dion dixit: S(crne nomen elus.
Davies paraphrases this as 'Dio ordered that his name should be removed from
the signaculsim' whereas it need only refer to the crossing out of the name from the list
of recruits.
Since Vegetius apparently tells us that recruits of his time (late 4th century) were
branded (1.8; 2.5), Davies seems to use this text as his basis for saying that lead scalings
were used instead in the Principate and that this text 'is best interpreted as an isolated
recorded example of the practices of former times surviving into the Later Empire'
(op.cil., note 63). This would make the use of these sealings run concurrently with the
majority of our military lead sealings, but it does seem to be a rather far-fetched
explanation. If this practice really did take place then we do not seem to have any
recognized examples of these identity discs.
113
Perhaps one of the best reasons for not accepting that our sealings were used for
this purpose is the fact that military scalings are known from so few provinces.
Less direct, but perhaps more useful, written evidence can be found in the Digesi
(XXXIX, 4, 9, 7) where we are told that materials destined for the army were exempt
from duty:
Res exercitui paratas pracstationi vectigalium subici non placuit.
This obviously refers to official supplies and so would not have covered items
being transported by soldiers on a personal basis.
The absence of evidence
Afier examining the evidence of the lead scalings which we do possess, perhaps
we should consider the absence of military scalings in many provinces. We have seen
earlier that imperial scalings may perform some of the services which one would consider
to be military and this may go part of the way to explain the paucity of military scalings
in some areas. However, the fact still remains that some provinces, most notably Britain,
have many military examples whereas other provinces, e.g the Germanies, have virtually
none. The accident of excavation, and metal-detecting, can only be blamed to a certain
extent. I feel that we must be seeing a real absence of these scalings, not just a perceived
one. This suggests that the units in areas with a dearth of military sealings were either
working on different tasks which did not require such security or, even more likely, were
using different materials/methods for their sealing. It may be that the heavy use of lead
scalings among the army in Britain was due to the relative abundance of the metal in that
province. Perhaps the army elsewhere was using unfired clay for its scalings. The lead
labels which we have mentioned above (see note 23) are not suitable for sealing packages
114
in the sense of providing security. They seem to be used more for displaying ownership
or the nature of the contents.
In Britain itself we should note the units which are not represented by sealings.
These include the legions which were here in the early period of the occupation (legio II
Adinirix transferred to Danube c.AD87, Iegio IXllispana ?left c.121; Iegio XIV Gemina
finally left in 70) and the units of cunei, equites, mimeri and vexillationes which are
attested for the late fourth century in the Notitia Dignitalum, but which seem to appear
from the second quarter of the third century (Jarret, 1994, 68-73). This could be taken to
suggest a terminus post quem for military sealings in Britain of c.AD121 and, less
certainly, a terminus a/lie qiiem of c.AD225. However, this is clearly overstepping the
mark since although the three remaining legions have all produced scalings, by no means
all of the auxiliary units in this time-span have. This should warn us that the absence of
sealings belonging to cunei, equites, mimer! and vexillaliones may be due to the fact that
these units were of a type which, like some of the a!ae and cohorts, simply did not
conduct activities which required lead sealings. It could be argued that this may apply to
the three earlier legions, but I feel that the minimal evidence of dated contexts for
military sealings does tend to favour a date later than AD 121, and certainly later than the
70s/80s.
Another form of missing evidence is that provided by sealings which are

categorised as miscellaneous (e.g. anepigraphic impressions from intaglios) but which
were found in military contexts. These will be examined more closely in chapter 9
('Miscellaneous Scalings - the Evidence for their Use').
In the past it has been considered that some lead sealings would have been used
to seal military diplomas (Martin Henig suggested this in 1992, either to me personally or
earlier the same week during one of his lectures for the British Museum 'Seals of the
World' seminar). Margaret Roxan points out that no diplomas have been found with lead
115
scalings attached whereas there are known examples with the remains of wax seals still
adhering (perc. comm.)
Examples dated by context (see catalogue for sources)
Legionary [0405; 0407; 0408] AD 160 - 200.

[0417] with 2nd C. material.
Cohorts [0454] with residual Antonine material in the Severan extension

to the fort at Brough by Bainbridge.
[0500] found in the wall construction trench of a building
provisionally dated c.AD85/90
[0503; 0504] from rubbish pits, or spreads of disturbed material
of mid-second to early third century
[0506] found in 3rd/4th C. context
[0507] from a 4th C. context in Milecastle 35
[0508] found in demolition layers over a drain in front of a
Severan barrack block
[0515] found in a probably early 3rd C. context
Equites Singulares [0527] AD 160- 200
It can be seen from this limited survey that [0500] is rather earlier than the other
examples and [0507] (with the possibility of [0506]) is somewhat later than the others.
We should note that the context date given for [0500] is only provisional and that the
sealing is single-sided (rare for a military sealing although perhaps indicative of an early
date) and, perhaps more importantly, is not definitely military, since the only inscription
on it is [ ..]LIN which has been assumed to represent a cohort of Lingones.
116
The sealing from a 4th century context [0507J is quite similar in method of
manufacture and style of inscription to the majority of military sealings which are
believed to date from the late 2nd to early 3rd century It is not impossible, however, that
the techniques survived into the 4th century although Tomlin said 'the date of this sealing
is ... unknown' suggesting that he viewed the association of the sealing with its context
as being debatable (Britannia xii (1981), 394, no. 104, note 119).
Conclusions
We can see from the information provided that, in general, military sealings were
used to protect the contents and, quite often, to name the person responsible. The
question of whether or not customs immunity was implicit in these sealings does not arise
for the majority since these were only used in Britain. The examples found in the area of
the docks at Lyon, however, may indeed have possessed this significance.
117
Notes
1. The use of '1' in front of a name in the genitive to refer to the Itirma of so-and-so
does not appear on any lead scalings but is attested in several inscriptions, including RIB
2410.12; 2415.63 &65; 2425.4-5; 2426.3, 16& 20.
2 However, this would mean that the examples from the cohors VII lhracum bearing
TVD [0476; 0491 - 04931 which is usually taken to mean tu(Iu)d(i!) - struck (this), may
refer to a decurion, T(...) V(.) Our case is helped perhaps by the reverse of the sealing
of ala II Ashirum bearing the inscription AETVD. The only other suggested uses of this
word, or parts of it, are [0506] (cohors I Aelia Classica - believed to be pedilala) and
[0531] (pedites singidares consularis - definitely pedilata!). The possible interpretation
of D as decurion would also help to explain away the only scalings thought to place a
centunal sign in front of the initials [0483 - 0484]. This symbol is perhaps better seen as
an S. It is difficult to say whether we should include the cohors VI! Thracum reverse MV
/ CAD [0494 & 0498] in this suggestion since CAD alone is also found on the reverses
of scalings of cohors VI Raelonim,although there is the possibility that this was equllala
(Jarret, 1994, 65)). It is even possible that the inscrption could actually read CVITR
with the T and the R ligatured [cf.0478]. It is known that cohors VI Thracum was
equitata and so the D may possibly refer to the decurion. The apparently single-sided
sealing [0772] which bears the inscription CAD / MV would, however, render it unlikely
that the D stood for decurion.
3. See, for example, the numerous entries for legionary centuries involved in the building
of Hadrian's Wall in RIB I Epigraphic Indexes, 1983, 9.
4. Examples of inscriptions with the added name of a soldier are RIB 2425.2-5;
2426.1-2; 2427.2, 16-18 & 20. Those without the name of a soldier are RIB 2410.1-7 &
12; 2415.58-59 & 65-66; 2427.3, 13 & 19; 2428.1 & 4.
5. These lead mines at Alston are only known to have been started in the medieval period
but Collingwood reminded Richmond that lead ore had been found in the fort at Whitley
Castle (Richmond, 1936, 109, footnote marked tf). However, Collingwood also
mentioned the presence of fluorspar Boon suggests that the unit ran a penal colony
rather than doing the work themselves (1991, 319) Fifty miles further north, and
possibly 100 years earlier, a document from Vindolanda probably listing the work duties
of the men of cohors VIII Balavorum tells us that some were working ad phimbtim
which, on comparison with other entries, is taken by the editors to mean that 'men were
detailed to acquire (or perhaps work with) lead' (Tab. !J'indol. Al, 77-79).
6. The presence of zinc in this area is shown by Jones & Mattingly (1990, 179, map 6.1)
but not listed in the relevant table (op.ciL, 181, table 6.1). However, Bishop and
Coulston say that for Roman Britain the nearest source of zinc may have been Aachen in
Germania Inferior (1993, 35).
118
7. The use of expedlo to mean 'to dispatch', however, appears to be poetical. The more
interesting prose meanings of the word which may apply here are 'to bring forward,
procure, make ready, prepare' although these usually refer to items which have been
folded or put away. Other meanings are 'to provide' and 'to hold in readiness', the latter
usually applying to weapons. Flassall has suggested (pers. comm.) that EXP may stand
for expedite - quickly - representing a 'first class' express delivery service.
8. Apart from [0428] which has the EXP on both sides.
9. See the discussion in chapter 3 under the subheading 'Evidence from association with
findspot'.
10. RMR 70 (a i 28; ii 25; b 1 9, 22) from Egypt gives ex priore ration('e) as meaning
'from the previous accounting'.
11. Birley points out that the dump contains 'a great quantity of objects in various other
metals' (1958, 41). He suggests that the packages may even have been opened
elsewhere. The rubbish dump, as it is usually referred to, is on the north side of the fort
and the scalings and other objects were found in the Swindale Beck which had changed
course and was eroding the slope. It seems possible that it was formed by refuse being
thrown from the northern rampart and allowed to roll down the slope. Dr. David
Woolliscroft, who has conducted surveys on the site, agrees with me on this. I had also
considered the possibility that what is usually identified as the Roman rubbish dump may
be the spoil heap resulting from the building of the Norman castle inside the fort, against
the northern rampart. In this way the scalings would originally have been disposed of
inside the fort, although still possibly as rubbish in pits. Some of the material found with
the sealings is said to be medieval (Smith, 1866, 144 & 146). Woolliscroft, however,
believes that during the Medieval period earth from inside the Roman fort was dumped
on to the Roman ramparts in order to strengthen them, and was not dumped down the
slope. His amendment to my idea is that there is physical evidence that the Roman
ditches were redug as defences for the Medieval castle and that therefore it is the rubbish
deposits contained within these which were dumped down the slope. Incidentally,
Woolliscroft informs me that he has now completed an authorised metal detector survey
of the entire north front of the fort and some way down stream (using one of the most
powerful detectors available) without locating any further scalings.
12. The 3rd century date usually assigned to these operations, gives us the possibility that
the goods' next destination was York, the capital of Britannia Inferior, rather than
London. However, it is possible that they could have been sent to the higher ranking
governor of Britannia Superior in London (and thence onwards?). If the product sealed
was silver then it may have gone to the treasury in London before being sent to Rome.
13. In 1847 Cumberland had 9oz per ton (270 parts per million) whereas Durham!
NorthumberlandAVestmorland had I 2oz per ton (360 ppm) In 1923 Cumberland had
6oz per ton (180 ppm) as did Durham etc. Jones and Mattingly point out that
119
desilverisation is economical at 120 ppm and even better above 170 ppm (1990, 190).
They note that a lead ingot from the Roman workings at Charterhouse, prior to
cupellation, had 560 ppm and perhaps we could imagine a similar figure for the higher
levels at Aiston.
14 The ordinary soldiers may just have been doing their job but one would hope that this
would occur to someone of higher rank. However, the quantity of the material sealed
may have rendered the wastage minimal in comparison. It would also have been
necessary to remove the tin which had been alloyed with the lead to make the sealings
(Richmond, 1936, 122-23).
15. Cf Appendix 1.1 - 36. The most important difference is that the type normally
associated with blocks of stone are formed by pouring molten lead into large holes
gouged out of the stone, i.e. they are large and one-sided unlike our small double-sided
scalings. The fact that the scalings for stone are from diverse sources, not just one
province, and belong to the imperial and miscellaneous categories, renders it unlikely that
the Brough scalings are a British alternative.
16. Birley also said that there was no doubt that cohors V.17 Thractim was equitala like
the other Thracian cohorts. This has only recently been proven by a diploma of AD 178
published by Margaret Roxan (RIt'II) 184). Janet has pointed out that the size of an
equitate unit would make it 'difficult to believe that it could have been accommodated in
the known fort at Brough' (1994, 67). Collingwood, however, had suggested in 1927
that the fortlet at Maiden Castle was garrisoned by a ft,rma from Brough (1927, 175-77).
Woolliscroft agrees with this (pers. comm.), apparently dismissing Birley's colourful idea
of a militia garrison (1958, 49). Woolliscroft adds that there are at least five towers in the
vicinity which would have needed to be manned, including Johnson's Plain which he
excavated (Brilannia xxii (1991), 235-7), although he points out that these may be rather
early.
17.There is only one possible procuratorial lead sealing known in the whole empire
[0286] but perhaps imperial scalings were usually used.
18 It is possible that other seal-box lids have been found, since Birley, précising Ecroyd
Smith (1866, 146), mentions 'amulets - chiefly of bronze, including many varieties of the
the heart shape ... and a large proportion of phallae (sic - presumably phalli is intended),
one of pewter, lately found, bears an imperial or other bust' (1958, 42) The 'sic' in this
quote is Birley's own and I prefer to believe that Ecroyd Smith meant phalerae, not in
the strict sense, but meaning pendants for horse harnesses. He then presumably refers to
our lead trial piece with the bust which could be misinterpreted as an item of 3rd century
equine equipment (cf Bishop & Coulston, 1993, 156, fig. 112) Whether or not his
heart-shape amulets can also be taken as seal-boxes/lids is open to question and they may
be late Roman strap-ends. The possibility that a cohors equitata was based here would
certainly make the presence of various pendants understandable.
120
19. Only 6 of the sealings [0509 - 0514] had been found up to the date at which they
were writing
20. The example of Iegio XI (iaud,a, already given above, appears to illustrate this, as
do [0453] and [0515].
21. This also seems to be the view of Tomlin (Brllan,iia xvi (1985), 325, note 29).
22. This may be reinforced by Breeze and Dobson's comment that the refitting of the fort
at Cramond may indicate that it was to play a part in the sea-borne supply line from
South Shields (1987, 136). It is, however, rather difficult to imagine what was being sent
back to South Shields.
23. There are examples of lead tags in Austria from Magdalensberg (Egger, 1967, 195ft)
and Kalsdorf (ROmer-Martijnse, 1990), German examples from Trier (Schwinden, 1985,
121ff) and British examples from several sites (RIB II 24 10.1-23)
24. 1 must thank Dr.David Woolliscroft for forwarding to me a draft copy of Caruana's
chapter on lead objects.
121
Chapter 9
MISCELLANEOUS SEALINGS - TIlE EVIDENCE

FOR ThEIR USE
The category of Miscellaneous sealings covers all those belonging to private

individuals together with those which cannot easily be assigned to any other category. It
therefore includes scalings bearing (na nomina, other unidentifiable inscriptions and
anepigraphic sealings with representations of deities and animals, many of which appear
to be impressions from intaglios.
Many Miscellaneous scalings bear the names, in various states of abbreviation, of

private individuals. These are assumed to be merchants (see section below on 'Evidence
from association with findspot'). The more recognisable of these names can be seen in
the indices entitled 'Nomina of Individuals and Families' and 'Cognomina of Individuals'
although there are many more which take the form of the initials of (na non,ina.' On
account of the general lack of punctuation found in these inscriptions, many of them can
be interpreted in several different ways. This is not he'ped by the fact that the initial
letters of some Iria ,ion,ina appear to form words in themselves One of the best
examples of this is that given in GIL XIII pars 3, fasc.2, p.722. This uses my [1202]
which has SEC on the obverse and SOL on the reverse. These are interpreted variously
as
S. E(...) C(...) (et) S. O( ..) L(...)

SEC(...) (et) SOL(...)
S. E(...) C( ..) SOL(...)
SEC(...) S. O(...) L( ..) (sc servi)
122
The last expansion, with servsis in the genitive understood and in apposition to
the slave's name Sec(...), reminds us of Grenier's belief that 'bisque Ia marque comporte
quatre initiales ii faut reconnaItre, sans doute, dans Ia dernière, celle du nom de l'agent,
homme fibre, aifranchi ou esciave, a Ia suite des initiales du patron de Ia maison de
commerce.' (1934, 657_8).2 Parker, however, when talking about similar sets of initials
in amphorae stamps, suggests that the fourth letter was 'an added initial or control mark'
(1987, 641).
Interpuncts are not always as helpfiul as they should be. For example, one
suspects that some of those in [0763] were added for decoration. It also appears that
some impressions have less than would be expected: [10211 has one but needs a second;
[1048] also has one but needs a second; [1064] has the second but needs a first, as do
[1072] and [1170].
Problems such as these can lead to great uncertainty in the case of some sealings.
although it is probably best to say that there was no set formula and that people included
anything that they wanted to in their matrix, limited only by space.
The figural representations depicted on the sealings tell us virtually nothing about
the use of the sealings and are dealt with separately in chapter 13 ('The Iconography of
lead scalings').
The inscriptions found on these scalings however, both with and without any
additional figural representation, can provide us with some interesting information.
Many scalings apparently refer to family businesses in that they give a nomen in
the genitive plural, i.e. telling us that this is the sealing of that family/company. These
scalings are as follows:
123
Aelio(rum) [0850], Cossior(um) [0930]; lgnior(um) [0987]; III IuI(iorum) He(...)
Sab(...) Senat(...) [0988]; Lucilior(um) [1055 - 1056]; Marior(um) [1074],
Modestio(rum) [1105]; Oratiorum S(exti) et M(arci) [1128 - 1130]; Satriorum [0832;
11971; Tertinior(um) [0751; 0835-0836; 1230 -1233].
All of these are believed to have been found at Lyon, apart from [0751] which is
from London.
Other sealings bear witness to partnerships by giving the abbreviated Iria nomina
of the two men involved, one on the obverse, the other on the reverse:
S. M(...) P(...) (et) L. F(...) S(...) [0798]; Calvini (et) Frontini [0873]; C. A(...) M(...) (et)
Q. 1(...) V(...) [0875]; C. A( ) T(...) (et) S. E(...) R(...) [0883]; C. E(...) D(...) (Ct) P.
A(...) T(...) [0891 - ?08921; C. S(...) R(...) (et) L. 1(...) V(...) [0945]; L. C(...) P(...) (et)
C.V(...) [ . 1 [1012]; L. D(...) A(...) (et) L D(...) M(...) [1013]; L. G(...) A(...) (et) L.
G(...) S(...) [1017]; L. V(...) V(...) (Ct) M. V(...) V(...) [1062 - 1063]; M. C(...) M(...)
(et) M. E(...) R(...) [1083]; S. C(...) S(...) (et) S. E(...) S(...) [1200]; S. E(...) C(...) (et)
S. O(...) L(...) [1202]; Q. C(...) T(...) (et) M. V(.. ) V(...) [1510].
Other sealings give the two sets of abbreviated iria nomina on the obverse:
M. C(...) P(...) ET T. N(...) M(...) [1085]; M. I(...) H(...) I C. M(...) V(...) [1097]; M
1(...) S(...) I C. M(...) V(...) [1100], P / MLM / P = P. L(...) P(...) (et) M. L(...) M(...)
[1149]; T. L(...) L(.) IA. T(...) R(...) [1235].
Other sealings may refer to these partners by abbreviations of their nomina or

cognomina (although the first example could fall into our previous group):
DOM(itii?) (et) PAN(...) [0956],NOB(ilis ?) (et) VIC(toris'?) [1126]; SER( ..) (et)
EPIC(...) [1203].
124
Many sealings bear the same abbreviated Iria noniina on both obverse and
reverse. It is impossible to tell whether these simply refer to the same man or a father and
eldest son. The former, however, is more likely These examples with repeated legends
are as follows:
MAM [0735]; LAM [0788]; ACF [0848]; CAM [0874]; CLP [0912]; CMD [0914];
CSS [0946]; CTM [0948]; GNT [0971]; LCM [10111; LED [..JD [1015]?; LII [1021];
LIP L[.]P [1023]?; LMR [1032]; LNI L[.] [1033]?; LRG [1035]; LSC [1040]; LSS
[1049]; MAL [1069]; MAP [1071); MIL M[..] [1098]?; PDD [1138]; QCM [1167];
SCV [1201]; SI-IM [.]HM [1204]?; SMA [1209]; SMM [1212]; TCT [1222]; MVR
[1562];
It is possible that other companies are also represented in sealings. One example
from Lyon is inscribed C[..]AS S[.]CIOR [0882]. CIL XIII, 10029.201 expands this as
C. [.1 [.1 (et) [ . 1 A(...) S(...) S[o]CIOR(um) thus interpreting it as the sealing of a
partnership (socii). A similar partnership may be referred to in [1774] which has AR /
SOC. although the names of the partners would not seem to be hidden in the
abbreviation AR. Turcan suggests, perhaps more plausibly, that this names a different
kind of grouping, the AR(gentariorum) SOC(ietas/ietatis), the association of silversmiths
(1987, 39).
While discussing partnerships we should mention the six sealings marked DIFF
[0955). Dissard believed that these referred to the corporation of dffiisores olearii ex
Baelica (1905, xii-xiii). This may be possible since the corporation's warehouse was not
far from the spot where the scalings were discovered. 3 The expansion of D I(.. ) F(...)
F(...) (Sc. servi - genitive) given in C!L XIII, 10029.115 is rendered unlikely by the fact
that Decimus is a relatively uncommon praenomen with, for example, only one or two
doubttiil entries in our main index to the Catalogue
125
One sealing which names an individual is of great interest since it also names his
town of residence [15071. The inscription says:
COCTDASIFORIVLIF
and is expanded as C(aius) Oct(avius?) Dasi(us?) For(i) luIi(i) F(ecit?).4
The great majority of the sealings in this category are single-sided with the usual
swelling, of varying shape, on the reverse.
The examples of single-sided sealings which bear wood or fabric impressions on

their reverses are listed in chapter 2.
Several sealings still have traces of 'string' (often described as hemp) in the holes
which run through them. They are as follows:
[0748]; [0751]; [0988]; [11201; [11541; [11651; [1198]; [1232]; [1444]; [1463]; [1467];
[15061; [1549]; [17751 and [1809].
In addition, [1652] still had traces of 'string' when purchased in 1950 but these
have since disappeared.
Several double-sided examples, all mentioned briefly under Type no.1 in chapter
2, conform to a style of manufacture usually encountered in impertal, official and military
sealings, i.e. one side is formed in a container-matrix and the other impression is created
by pressing a separate matrix on top of this.
The examples in question are [0678 - 0679] (and possibly [0677]), [0694],
[0727], [0731]; [0735 - 0736] These sealings apparently belong to private individuals
126
but perhaps we should consider the possibility that some may be concealing references to
government-backed organisations.
Sealings from Lyon and London - Virtually all of the sealings found in Lyon and
London were discovered in waterfront deposits (Dissard, 1905, v-viii; Turcan, 1987,
13-14; RIB 2411, passim). It is generally believed that they had been thrown into the
rivers when the goods to which they had been attached were unpacked for storage in the
warehouses (Grenier, 1934, 646; Turcan, bc. cii.). If this is correct, and there is little
reason to doubt it,5 then it suggests that the merchants named on the sealings were at the
place of dispatch. Turcan, however, draws attention to several references to individuals
who are possibly connected with people mentioned on the sealings (op. cii., 39, 41 2).6
While preserving my scepticism (mentioned in Note 1) of looking too closely at these
bland names, we should at least consider the possibility that the men referred to on the
sealings were based in Lyon and that the sealings, while not acting as address labels,
were announcing that the merchant's agent at the place of dispatch had sealed the goods
in his name.
As we have seen above, it has been suggested that some scalings may include the
name of the agent, either free man, freedman or slave (Grenier, op.cil., 657-8). Despite
the fact that the latter would presumably fit in better with the use of a single initial, it
may be that a freedman would also use a single initial in order to save space by not
repeating the praenomen and nomen of his patron If it is correct that one initial refers
to an agent, regardless of his status, this could indeed suggest that the company owner,
named on the sealings, was not based at the place of dispatch It could, however, just be
a security arrangement so that goods could be traced back to the individual responsible
for sealing particular packages. Even if these inscriptions have been interpreted correctly
there are comparatively fewer of them than of other sealings of merchants.
127
Basically, we have no way of telling from the scalings where our merchants were
based. The sealings would have fulfilled a dual role, in that they would have prevented
tampering by the ship's crew and would also have acted as labels to enable merchandise
to be recognised. This would have been necessary when cdliecting goods from ships
which probably carried mixed cargoes belonging to several individuals/companies For
both of these rOles the only necessary information needed on the sealing was the
individual's/company's name or some recognisable symbol.
One way in which we could connect the merchant who dispatches his own goods
with people in the receiving port who bear the same nomeii is if the merchant sent his
relatives, perhaps his son(s), or a freedman to oversee the reception of the goods at their
destination. S There is little reason to doubt that the use of a son for this purpose would
be almost as likely for individuals named on scalings as for overt family partnerships.
Sealings found in River Tees at Piercebridge [0758 - 07611 - These scalings were
found by divers in the area of the Roman bridge, along with a large amount of pottery
and votive objects (Britannia xx (1989), 277 & 337). Tomlin surprisingly links the
scalings with the votive deposits which were apparently dropped from the bridge. I am
rather sceptical about this and would prefer to view the sealrngs (and presumably an
unspecified proportion of the other objects) as having been thrown away as rubbish,
perhaps after having been unloaded from boats although not necessarily.
At least one of the scalings, [0759], is almost certainly military although [0760] is
definitely not, apparently being that of a Greek merchant.9
Anepigraphic Miscellaneous seahngs from Brough under Stainmore 10705 - 0725j -

It would be interesting to know whether these scalings, virtually all apparently taken
from gemstones, represent goods sealed by merchants and then delivered to Brough or
whether they indicate items sealed by soldiers and sent to Brough like the majority of the
128
sealings found there (see chapter 8 entitled 'Military scalings - the evidence for their
use').'°
Some of the sealings of auxiliary cohorts found at Brough bear the impressions of
intaglios on their reverses. These all belong to cohors II Nerviorum and consist of five
examples showing Jupiter seated [0459], one example with a bird facing a vase
containing a victory palm (?) and three examples [0461] with combination figures. It is
these last three which are almost identical in design to eight of our one-sided
anepigraphic scalings [0718 - 0725]. I have examined the one-sided scalings but have not
been able to compare them for possible die-linking with the overtly military scalings. The
poor state of the one-sided examples has made it virtually impossible to attempt this even
within their own group.
A look at the epigraphic Miscellaneous scalings from Brough [0682 - 0704]

reveals an unusual collection, some probably misunderstood military examples [0691;
0704], and the rest quite unlike any merchants' scalings." This can only serve to
strengthen the feeling that the anepigraphic scalings were in fact serving a military
purpose, or at least being used by military personnel.
Evidence from association with item sealed
Of all known lead scalings, of any category, there is only one which was found
still attached to the item on which it had originally been fixed. This is a sealing from
Carnuntum (modern PetronelL/Bad Deutsch Altenburg) which bears an impression
showing a capricorn swimming to the right [1680] 2 The interesting feature is that the
otherwise blank reverse displays the marks of having been applied in a molten state on to
a piece of lorica hamala, ring mail. There are even two complete rings and some
fragments of more still embedded in the reverse. This military connection fits in well with
the capricorn since this was the symbol of, among others, legio XIV Gemina Mania
129
Vicirix, which formed the garrison of Carnuntum In his article on the sealing, Dembski
denies that it could have been produced during an idle moment by a soldier since he
believes that an official matrix was used (1979, 157) The shape and size of the
impression, however, suggest to me that it was made using an intaglio ring which has
even left the mark of its bezel. Dembski therefore believes that the sealing may represent
a general identification mark arranged by the legionary command to label armour
belonging to it or, following HannsjOrg UbI's suggestion, that it was a production mark
of a workshop belonging to Iegio XIV (op.cil., 157-8). In the absence of any examples
with similar associations it would be unwise to accept this idea at the moment. Since I
have no problem in seeing the matrix as having been a gemstone set in a ring, I would
return to Dembski's rejected ideas of a soldier playing around, marking his own armour
with his own signet for identification or wishing to show his membership of Iegio XIV
(op.cit., 157). 1 believe that the piece of lorica may even have been a scrap piece.
Dembski points out that none of the rings in the impression were broken but of course he
cannot say how large the original piece was (ibid.)
There are apparently no ancient written sources which have any direct bearing on
this category.
Examples dated by context (see catalogue for sources)
[0678 - 0679] are said to be from the upper filling of a pit of late first or early second
century, although this seems rather early to me, particularly in view of their form
mentioned earlier in this chapter under the heading 'Evidence from form and reverse'.
[0727] can apparently be dated by context to AD 160-200.
130
[0736] was found in a context said to pre-date the extramural cemetery at the West End
development in Leicester. No clear date has been provided for this but there is a
possibility that the context could be mid-second to early third century (Clay, 1980, 318).
[1511 - 1512] can apparently be dated by the fact that they are from a fort, Velsen 1,
which was only occupied between AD15 and 30.
[15821 is dated by associated artefacts found in a shipwreck off Malta. The wreck is
dated to the late second/early third century.
Surprisingly few of our Miscellaneous scalings can be dated closely by context

although it is interesting that [0727], [0736] and [1582] give such similar dates It is
quite possible that [1511 - 1512] do actually date to ADI5-30 since they are of a type
quite different to any other scalings (see chapter 2 entitled 'Typology of lead scalings'). I
am rather concerned about the early date proposed for [0678 - 0679] and would imagine
from their form that mid-second - early third century was more likely.
Conclusions
The category of Miscellaneous scalings is, by its varied nature, a hard one to
define. Many of the examples included within it are quite probably serving a military
purpose or being produced by soldiers for their own personal use. The findspots, form of
the sealing and their general difference to those examples accepted as merchants' scalings
all provide clues that this is the case. This, however, is as far as we can take this idea and
its implications for military supply by private traders, since their can be no firm proof.
The scalings which do appear to name merchants were obviously intended to be

recognisable. Whether this was for proof that the sealing was original and had not been
tampered with (thus ensuring the integrity of the goods sealed) or whether it was for ease
131
of recognition and recovery of one's goods from a ship's cargo, we cannot tell. Quite
possibly the scalings performed both functions. On the other hand, security may have
been the sole function with other labels, of a perishable nature, used to denote
ownership. It is unlikely that scalings were used as labels of ownership without the
security function, despite my past attempts to suggest reasons for this (Still, 1993, 405).
We can, however, be almost certain that these sealings have no connection with
the payments of customs dues, being just for the benefit of the private individuals named.
132
Notes
1. Grenier points out, quite correctly in my view, that even for the more complete names,
'leur banalité ne permet pas de les rapprocher des noms analogues qui peuvent être
fournis par l'epigraphie des diverses provinces' (1934, 658). However, this does not
seem to have stopped Turcan from attempting to do so (1987, passim)
2. Although [1190] suggests that the stave-name (if that is what it is) can appear first,
since there is no praenomen beginning with an 'R'.
3. The problem here is that no lead sealings have ever been found in association with
amphorae, even on shipwrecks. Stamped lead labels were sometimes used but these were
impressed strips of lead which were wound around the handles of the amphorae. The
best published collection of these is a group from a wreck off the Algerian coast
(Lequément, 1975, 667-680). I have also been informed by Jean-Pascal Jospin, curator
of the Musée Dauphinois, Grenoble, that there is an amphora from Vienne in the bottom
of which was found 'un plomb avec marque'. This is housed in the château de Moidiêre
at Bonnefamitle.
4. if gdt) is correct then this is one of only two places that the word may be found on
lead scalings, the other possible example being [06811 which has CBFEC, although I
have suggested that this may be the iria iiomina C. B(...) Fec(...).
5. Francis Grew has, however, drawn my attention to the fact that some of the
waterfront scalings from London may have been 'dumps of general refuse, imported
from anywhere in the City' (pers. comm.).
6. The best examples of Turcan's research are presented here, each followed by
references to relevant scalings in my catalogue. The name Q. Ignius is known at Lyon
(GIL XIII 1953; 2006il; 2070) and on scalings from there [0829- 0831; 1172; 1174 -
1176]. The iiomen is also present on [0983; 0985 - 0987]. The iiomen Satrius is found on
stone at Lyon (GIL XIII 1824; 1888; 1910; 2193; 2258) and also on scalings from there
[0821; 0941; 1197 - 1198]. The nomen Tertinius appears on stone several times in Lyon
(GiL XII! 498; 1854; 1897 - 1898; 2084; 2281) and again on scalings [0834; 0835;
0835A; 0835B; 0836; 1230 - 1233].
7. In his suggested expansions which I have included above, the editor of GIL XIII pars
3, fasc.2, Oscar Bohn, only refers to a slave being named, not a free or freedman (Cl!.
XII pars 3, fasc. 2, p 722).
8. 1 am suggesting this in the light of Parker's comment, based on the evidence of tiluli
picti on amphorae (especially Liou, 1980), that 'documents such as these show how
much of western imperial trade was run on a family basis; the family might be based in
Baetica, or southern Gaul, but a son or a trusted freedman would be posted to Ostia or
Rome to act at the far end of the main despatch route' (1987, 642)
133
9. It should be noted that I have suggested an improved reading of [0759] to Tomlin and
he agrees with this to a large extent.
10. Richmond believed that the auxiliary sealings with gemstone-impressed reverses were
those of the 'commandant' and that those 'unassociated with cohort-sealings also belong
to commandants of this or other units' (1936, 121-2) However, I have pointed out in
chapter 2 that any soldier could have possessed such a ring.
11. Richmond felt that the epigraphic Miscellaneous sealings had probably become
detached from their auxiliary scalings (op.cil., 122). This does not technically apply to
some which are double-sided but those examples are the ones which I believe are
misunderstood military sealings anyway.
12. I have referred to this sealing in print elsewhere (Still, 1993, 406).
134
Chapter 10
SEALINGS AS EVIDENCE FOR TIlE DISTRIBUTION OF GOODS
There are two ways in which we can assess how far sealings have travelled,' one
is absolute whereas the other is relative:
1. The absolute method uses sealings which actually bear the name of the town or
province, or its people, in the name of which they were applied.
2. The relative method involves the identification of two or more examples of the
same impression in different find spots. Although we are often unable to give a place of
origin, we can at least see the spread of destinations for the goods thus sealed.
In order to study Method no.1 we shall examine the evidence of the Imperial,
Taxation, Provincial and Civic categories of the Catalogue, since these are the only
examples to include place-names. 2 In the Catalogue these are ordered under their find
spot, but here we shall list them under the order of their origin so as to show more
clearly how widespread the distribution of some sealings was. The distances given are
those between place of origin and find spot. These are measured in a straight line so as to
avoid any preconceived ideas about ancient sealland routes. General details of some of
these routes are known from literary sources (i.e. Alexandria to Rome3) but there are
many more which cannot be securely identified. The example of Alexandria to Rome also
demonstrates that many of these routes could be much longer than the figure 1 have
given. Where one place, either origin or find spot, happens to be a province with no
further details, I have measured up to the nearest border, not to any putative provincial
capital. This has meant that some entries can have no measurement since the find spot
lies within the issuing province.
135

ORIGIN OF SEALING FIND SPOT OF SEALING DISTANCE RECORD NO.
IMPERIAL
Alexandria, Aegvptus Rome, Italia 2000km 10097A1

?Romc, Italia 2000km 101 15Al
TAXATION
Arciate, Gallia Narbonensis London, Britannia 1000km 102931

?Lyon, Gallia Lugdunensis 240km (0295-02961,
(0297J (2 ex.),
(0298-0301j
Ostia, Italia ?Ostia, Italia (03021
PROVINCIAL
Aegyptus lovia Dimitro grad, Thracia c. 1330km 103211

Britannia London, Brilannia 103111
Britannia inferior Aldborough, Britannia inferior (03061
Binches(cr, Britannia Inferior 103071
Felixstowe, Britannia Superior c. 140km 103101
York. Bntannia Inferior 103131
Britannia Superior Burgh Castle, Britannia Superior ------- 101081
Combe Dov n, Bntannia Superior ------- 101091
Epirus Vetus4 Pannonia c.lOOkm 116911
Maxima aesariensis Silchester, Maxima Caesariensis ------- 103121
Trier, Gallia Bclgica c.450km (03 14-03 15J
Pamph) ha ?Lyon, Gallia Lugdunensis c 2400km 10115A1
Izvoarcic, Moesia Inferior c 800km 10316-03191
Pontus Cabyle, Thracia c.275km (0320J
CIVIC

?Adramyttnim, Asia lmoarele, Moesia Inferior 550km (0161(

?Antandrns, Asia I'voareIe, Moesia Inferior 540km (0362j
136
I
Coloe, Asia Unknown site, Moesia Inferior c.SOOkm 10354-03561
Cyme, Asia KalugcrO%o, Thracia 350km 103761
Ephesits, Asia Trier, Gallia Belgica 2200km 103281
Iivoarcic, Moesia Inferior 720km 10313-03361
Unknown site, Moesia Inferior c.550km 103511
Cabyle, Thracia 540km 10369-03701
Dimi(rovgrad. Thracia 520km 10371-03721
Kalugero%o, Thracia 500km 10374-03751
I-Iypaepa, Asia I,voarcle, Moesia Inferior 680km 10348-4)3501
?Hypaepa, Asia Kalugerovo, Thracia 500km 103871
Laodicea ad L)Cum, Asia Iivoarele, Moesia Inferior 750km 103371
Magnesia ad Maeandrum, Asia Izvoarele, Moesia Inferior 725km 103381
?SishIov, Moesia Inferior 690km 103511
Metropolis, Asia Unknosn site, Moesia Inferior c.550km 103571
Pergamum, Asia Kalugerovo, Thracia 350km 10377-03781
Smyrna, Asia Ickham., Britannia 2600km 103251
?Lyon, Gallia Lugdunensis 2100km 101311
Izoarelc, Moesia Inferior 660km 10339-03471
Unknown site, Moesia Inlerior c.490km 10358-03601
Kalugerovo, Thracia 425km 10380-03851
Civitas Corieltauvorum, Britannia Thorpe in the Glebe, Britannia 10326]
Glevum, Britannia Cirencester, Britannia 20km 103241
?Vinoia, Britannia Buxton, Britannia 160km 103271
Apollonia, Galatia Kochennovo, Thracia 800km 103861
Tavium, Galatia Trier, Gallia Belgica 2000km 101291
Unknown city in Galatia Kalugerovo, Thracia c.600km 103731
?Pictaium. Gallia Aquilania Unknown site, Pannonia c. 1200km 103661
Rusicade, Numidia ?Lyon, Gallia Lugdunensis 1100km 103301
Skikda (i.e. Rusicade), Numidia 101641 (5+ cx.)
?Savaria, Pannoma Superior Strebersdorf, Pannonia Superior 25km 103651
Colonia Berylcnsis, Syria Beirut (i.e. Col. Beiylensis), Syria ------- 103671
Tyre. Syria Rome, Italia 2150km 103321
?Tyre, Syria Belosem, Thracia 1350km 103681
?Svishtov, Moesia Inferior 1450km (0352]
Philippopolis, Thracia Kalugerovo, Thracia 110km 103791
137
The average distance of travel (as the crow flies) of sealings from their places of
origin within each category can be calculated by dividing the total combined distance of
all sealings within that category by the number of examples, i e
Total of distances
number of examples
Thus, the average distance of travel for the above listed Imperial scalings would
appear to be:
4000 - 2000km
2
The average distance of travel for the above Taxation scalings, including the
Ostia example, is:
2920 = 292km.
10
The average distance of travel for the above Provincial scalings from the border
of their province, including the 7 examples found within the issuing province as having
travelled 0 miles, is:
8945 = 496.9km.
18
The average distance of travel for the above Civic sealings, including 10 examples
which could only be measured to the nearest provincial border, is.
43288 =709.6km.
61
138
Therefore, the average distance travelled by the sealings listed above in all four
categories is:
4000 + 2920 + 8945 + 43288 = 59153 = 650km

2+10+18+61 91
This figure can only act as an indication to remind us of the distances which other
sealings may have travelled. It is possible that some of these sealings which named their
place of origin were intended to travel further afield than the others which did not and
have therefore produced an unnaturally high figure, although the British provincial
examples demonstrate that this is not always true [0306 - 0309; 0313]. In addition, I
believe that many of the Civic sealings were connected with the payment of customs dues
(see chapter 7 entitled 'Civic sealings - Evidence for their Use') and so would have been
attached to goods regardless of how much further they were intended to travel.
In order to study method no.2 (i.e. the spread of identical sealings across the
Empire) we shall examine the evidence of the entire Catalogue, omitting those examples
which fit the criteria but which have already been listed above. Since the majority of
places of application will be unknown, we shall list them in the order in which the first
example appears in the Catalogue. Again, distances are measured in a straight line since
not only do we not know which routes would have been followed but we cannot even
guarantee that the distribution pattern passed through one find spot before reaching the
other, i e. we could well be dealing with two completely different lines of distribution
emanating from the unknown place of origin.

RECORD NUMBER FIND SPOT OF SEALING DISTANCE
IMPERIAL1

1001 6J London, Britannia 450km

100521 ?Lyon, Gallia Lugdunensis
139
OFFICIAL

102841 Unknown site, Britannia c 400km

102851 Trier, Gallia Belgica
MISCELLANEOUS
106821 Brough under Stainmore, Britannia 44km

J0758J Piercebridge, Britannia
107331 Kirmington, Britaimia 90km

107621 ?Ravenscar, Britannia
107431 London, Britaniiia 450km

110841 ?Lyon, Gallia Lugtlunensis (2 examples)
107501 London, Brilannia (3 examples) 450km

112231 Lyon., Gallia Lugdimcnsis (but lacks arrows and interpuncts)
107511 London, Britannia 450km

108351 Lyon. Gallia Lngdtmensis
10835A1 Lyon. GaHia Lugdtrnensis (6 examples)
10835B1 Lyon, Gallia Lugdunensis (70 e"camples)
(0836J Lyon, Gallia Lugdiinensis
112301 ?L)on, Gallia Lugdunensis
112311 ?Lyon, Gallia Lugduncnsis
I1232J ?Lyon, Gallia Lugdunensis
112331 ?Lyon, Gallia Lngdunensis
(1596AJ Izvoarele, Moesia Inferior 285km

117311 Ka1ugero o, Thracia
1161 OAJ lrioarele, Moesia Inferior 285km

11730A1 Kalugcrovo, Thracia
140
The four occasions on which identical scalings have been found in Lyon and
London are probably evidence of the trade route from the Mediterranean (and further
afield) up the Rhône to the junction with the Saône at Lyon, up the Saône, overland to
the Moselle and then down the Rhine (or overland from the Saône to the Seine)
Whether or not all of the goods would have been unloaded at Lyon before some
continued to London or whether the goods bound for London stayed on board the river
craft on to which everything had already been transferred at ArIes is difficult to say
(King, 1990, 115). The way in which the scalings at Lyon appear to have been
immediately removed at the quayside and thrown into the river would, however, favour
the latter.
It is interesting that identical scalings have been found at the sites of Izvoarele in
Moesia Inferior and Kalugerovo in Thrace (identical examples of Ephesus and Smyrna
scalings have also been discovered at these two sites). While being only 285km apart
from each other they are supposedly very different in nature, Izvoarele probably being
the foriltown of Sucidava and Kalugerovo presumably being the mansio of Arzus. It has
in the past been suggested, however, that both of these sites may have acted as emporia.6
Regardless of this, we should probably see both sites simply as being settlements with
economic lives, which were thus open to receive goods from similar sources.
Conclusions
We can see from the information in this chapter that lead scalings are yet another
type of artefact, like amphorae and barrels, which allows us to trace the distribution of
(presumably) perishable items which would otherwise be unknown to the specialist in
trade and the economy.
141
Notes
I. however we should bear in mind that the goods themselves may have travelled from
further afield prior to having the lead sealing attached.
2. This method could be adopted for the scalings of military units of which we know the
bases. My argument against so doing is that, on account of the movement of garrisons,
we would have to impose dates on many of the scalings which they cannot support in
themselves.
3. The intended route from Alexandria as far as Cape Malea can be deduced from the
account of misfortunes given by Lucian in The Ship, 6-9. We can also glean information
for the route from Caesarea in Judaea to Rome from another voyage which did not go
quite as intended, that of St.PauI in the Acts of the Apostles 27.1 - 28.13.
4. It should be noted that the true nature of this sealing was recognised at a late stage
and it is therefore entered under Miscellaneous in the Catalogue. It is not included in its
rightful place in the chapter entitled 'Categories of sealings found in the provinces'.
5. It is probably safe to assume that these scalings originate from the same place on
account of their identical design, although I have omitted some other Imperial examples
due to their more general appearance which could not be so closely linked.
6. See Still, 1995, 354-355 for full information on both sites.
142
Chapter 11
IMPERIAL SEALINGS - TIlE STATISTICAL EVIDENCE
The number of Imperial sealings discovered in each province, in ascending order

of quantity, compared with the total of all categories for that province is as follows:
GALL LUG (65/1042)
FlAt IA (48/1 19)
MOESIA INF (45/158)
TIIRACIA (40/1 17)
BRITANNIA (35/372)
PANNONIA (25/61)
GAL LIA B1I. (16/53)
PANN SUP (15/30)
IJNKNOWN PRV (12/54)
NORICLJM (6/Il)
ASIA (1/22)
SYRIA (2/11)
AFR PROCON (2/2174)
(ii RM SIll' (1/5)
10 20 30 40 50 60 70
Fig 2 Actual numbeTs of Impei-ial scalings foirnd in each province (there are no exact figures for Numidia)
The figures represented in this histogram can be seen to be extremely susceptible

to the overall numbers of sealings of all categories found in these provinces. For
example, the six top provinces shown here are also the six largest collections of sealings
(although not in the same order).
143
It would be more productive to view these figures as percentages of the total
number of sealings of all categories for each separate province (see chapter 12 on
provincial statistics for the details of these figures):
NORICIJM (54.55%)
PANNONIA SUP (50%)
PANNONIA (40.98%)
I FALIA (40.34%)
TIIRA(lA (34 19%)
GALLIA DEL (30.19%)
MOESIA INF (28.48%)
UNKN PROV (22.22%)
GPRMANIA SUP (20%)
SYRIA (18.18%)
ASIA (13.64%)
BRITANNIA (9.4 1%)
(',Al Lilt I (JG (6.24%)
ArRl('A PROC (0.94%)
EMP(4 (ii) (13.73%)
EMP (- CiL) (19.94%)
5 10 15 20 25 30 35 40 45 50 55
Fig. 3 Imperial scalings as a percentage of the total of all scalings in their respective provinces compared with the
total of imperial scalings as a percentage of the whole empire total (including and excluding Gallia Lugdunensis)
In this way we see the imperial sealings set in their true context within their own
province, thus rendering the evidence of the histogram immune from the influence of any
difference in the size of provincial totals. Obviously, high percentages of imperial sealings
may be caused by a low number of sealings in other categories but, to a certain extent,
this is what the exercise is all about.
If we study the histogram we find that, of the seven provinces with over 25°
imperial sealings in their provincial totals, only Italy, Thrace and Gallia Belgica could not
be described as frontier provinces. In fact Thrace often resembles a frontier province in
144
many ways' and Italy (most sealings being from Rome) obviously occupies a special
place as the hub of the empire. Gallia Belgica (i.e. Trier) assumes this role to a certain
extent in the fourth century but many sealings precede this, perhaps connected with the
fact that the procurator of Belgica and the two Germanies had his headquarters there
(Wightman, 1985, 62).
However, out of the remaining six identifiable provinces (i.e. those with less than
25% imperial sealings in their provincial totals), three (Germania Superior, Britannia and
Aflica Proconsularis) with a possible fourth (Syria - depending on the date of the
scalings) are also frontier provinces.
At first glance this negates the equation between frontier provinces and a high
level of activity using imperial scalings. But closer inspection suggests that the true link is
between frontier provinces on the Danube and the use of imperial scalings. This may in
part be due to my in-depth research in this area but, in favour of my suggestion, the
publication of imperial scalings would be more likely than of private examples in any
country. It would appear from chapter 3 ('Imperial Scalings - The Evidence for their
Use') that the majority represent the organisation of provisions for the army by the
emperor, although it is unclear as to whether or not this was only during campaigns. One
reason that the Danube area has this high proportion of imperial scalings could be the
incessant nature of tribal pressure on the border in this region.
When we compare the percentages for the provinces with those for the whole
empire (the inclusion or exclusion of Gallia Lugdunensis makes little difference) we find
that the majority of provinces possess well above the normal amount of imperial scalings
as typified by the empire as a whole. However, this tells us little except that there are
several provinces without any imperial scalings which therefore bring down the
percentage figure for the whole empire.
145
In looking at the above figures, we should bear in mind that imperial sealings may
tend to cluster on certain types of site If this is so then the figures will be biased as to
whether the provinces possess such sites (or at least those which have been
excavated/metal-detected).
We shall now look at imperial sealings chronologically. We have no direct

evidence to suggest that the lead in any scalings was ever recycled in the Roman period
so the scalings as found should, in theory, be representative of those actually used.
Common sense, however, tells us that perhaps recycling is something that should be
borne in mind.2
Sealings assigned to emperors by inscription
If we compare the quantity of scalings of each emperor using only those scalings
in the catalogue which name the emperor (and excluding the scalings in the appendices)
then we arrive at the following histogram with the actual figures for each emperor shown
in brackets after the name (see Table I below for details of Catalogue numbers):
IX )MII IAN III

1 RAJAN 151
ANT.PItJS [2)
M.AIJR['I ItJS (2j
GAl 111 NtiS Ill
CONS1ANTINE I 151
(RISP1JS 121
(ONS1ANTENE 11111
(ONSIANS 121
Jill IAN 151
I IJJX)XIA 121
JULiuS NFPOS Ill
.5 1.0 1.5 20 25 30 5 40 45 50 55
Fig 4 Actual numbcr of linperial sealing identifed ny inscnption
146
This is perhaps better seen as the average number of sealings discarded in each
year of each of 21 separate periods out of a putative 1000 Imperial sealings from
throughout the empire. This will allow us to compare it directly with a later histogram
showing less securely dated sealings which have been identified by portrait alone This
method is based on that used for coin finds (Casey, 1984, 28) although I have adapted
the length of some of the periods in order to fit the nature of our evidence. The basic
calculation to obtain such a figure is:
imperial scalings per pcnod X 1000

length of period total of imperial scalings for empire
However, this can be adapted as necessary and in this case we will be using:
imperial scalin gs id. by inseT., per period X 1000

length of period total of imperial scalings for empire
id. by inscr.
It should be stressed that this method is very unsafe when used in conjunction
with such small amounts, but it is necessary for a direct comparison with other
histograms since it omits the variable of period length.
I have indicated which Records are being used and recourse to the Catalogue will
provide exact details of the emperor involved.
The figures (to two decimal places) and division of periods used in the histogram
are as follows:
147
Table I

Period Actual quantity & Record Nos. %
A (3OBC-AD4I) 0 0
B (41-54) 0 0
C (54-69) 0 0
D (69-96) 1 [0098] 1.33
E (96-117) 5 [0046; 0068; 0162-164] 8
F (117-138) 0 0
G (138-161) 2 [0089; 0117] 3
H (161-180) 2 [0050; 0099] 3.67
1(180-193) 0 0
J (193-222) 0 0
K (222-23 8) 0 0
L (238-260) 0 0
M (260-284) 1 [0194] 1.33
N (284-310) 0 0
0 (3 10-337) 9 [0010; 0020(2 ex.); 0029-0030; 0088; 0135- 11
0136; 01601
P (337-360) 2 [0017; 0028] 3
Q (360-375) 5 [0011-0014; 0138] 11
R (3 75-400) 0 0
S (400-425) 0 0
T (425-455) 2 [0001; 0236] 2.33
U (455-475) 1 [0114] 1 67
Total 30
148
The above figures produce the following histogram:
iig impenai seaiing per nousarni iaernuIeu Dy inscnption
The histogram shows high figures (%o) for Period E (effectively the reign of
Trajan), Period 0 (the reign of Constantine and his Caesars) and Period Q (Julian and
Valentinian 1). The figure for Period E is heavily influenced by four sealings [0068,
0162-0164] which are purely epigraphic with no bust - the only ones of their kind (i.e.
referring to a specific emperor without showing him). The reason for The absence of a
bust is unclear and need not concern us here (although, judging by the lettering, lack of
artistic ability may be a contributory factor for [0162-0164]). However, the conscious
decision to produce purely epigraphic sealings at that time explains the high figure shown
in the histogram.
The high figure for sealings identified by inscription in Period 0 may be due to
the fact that the number of Caesars under Constantine had brought about the need for
149
close identification of emperors whose portraits were so similar. This problem had not
arisen under Diocletian since the majority of imperial scalings in his period were 'issued'
by the Tetrarchy as a whole, with four portraits [0180, 0200-0202, 0225; 0261]. This
prevalence of contemporaneous rulers in the time of Constantine has probably also
nudged the total higher.
The high figure for Period Q is the result of five sealings, four apparently from
the same die [0011-0014]. Unfortunately, the discovery of these four identical sealings
on one site renders the figure in our histogram of less use than others. This also
demonstrates the dangers of using such small numbers of scalings.
Sealings assigned to emperors by portrait identification
We shall now compare the quantity of sealings which have been assigned to
certain emperors by virtue of portrait identification alone. The evidence presented here is
obviously less secure than that given in the previous histogram. The division into periods
enables us to enter some of the doubtful sealings in the correct place according to the
style of the portrait despite the fact that they could not be assigned to the reign of a
particular emperor. While this loses some of the detail it hopeflully renders the work more
broadly correct and thereby usable. The details of which Records are being used are
included in the table. For the purpose of this histogram I have had to narrow down
identifications where the index suggested two different emperors for the same
impression. I have also omitted three scalings, [0095, 0102 and 0122], which were
originally inscribed with an unknown emperor's name but lost it due to damage (i.e. they
do not belong within our body of deliberately unnamed scalings)
The calculation being used here is:

imperial scalings id from portrait, per period X 1000

length of period total of imp scalings for empire
Id. from portrait
150
First, here are the figures and the Record numbers used to obtain them:
Table II

Period Actual quantity & Record Nos.
A (3OBC-AD4I) 1 [0076] 0.09

B (41-54) 0 0
C (54-69) 5 [0127; 0230; 0239; 0257; 0276] 2.82
D (69-96) 1 [0183] 0.34
E (96-117) 1 [0085] 0.43
F (117-138) 4 [0248-0249; 0252; 0274] 1.62
G (138-161) 4 [0097A; 0101; 0104; 0280] 1.1!
H (161-1 80) 9 [0069(2 ex.); 0070; 0071(2 cx.); 0092; 0100 4.02
0115; 0128]
1(180-193) 2 [0078; 0082] 1.28
J (193-222) 56 [0003; 0008, 0015; 0021(15 cx.); 0022; 16.84
0032; 0042; 0047 - 0049; 0053-0054;
0058-0059; 0072(2 ex.), 0077; 0081;
0083-0084; 0086-0087; 0091; 0093-
0094; 0105-0106; 0119; 0129-013 1,
0185; 0190-0191,0196; 0199; 0240;
0258; 0262; 0271-0272; 0277]
K (222-23 8) 6 [0079(2 ex), 0080; 0124; 0134; 0184] 3.25
L (23 8-260) 12 [0043; 0075(2 cx.); 0096; 0165; 02 10- 47
0211; 0228, 0255; 0260; 0278-0279]
M (260-284) 2 [0004; 0182] 0.68
N (284-3 10) 9 [0074; 0180, 0200-0202; 0219; 0225; 0232, 2.99
0261]
0(310-337) 1 [0125] 0.34
151
P (337-360) 0 0
Q (360-375) 3 [0073; 0197; 0203] 1 71
R (3 75-400) 0 0
S (400-425) I [0060] 0.34
1 (425-455) 0 0
U (455-475) 0 0
Total 117
These figures produce the following histogram:
rig ij uiipiiiu ciIug'i pJ UIUIEidilu IUCULIIICU LJUIII tAll 11d11 WUIIC
This histogram shows one outstanding period of recovered sealings, Period J

(Pertinax, Severus and his family, Elagabalus). This is basically a real difference not just
152
an apparent one. The only doubt is that Severus and Caracalla are usually easily
identifiable, perhaps more so than some other emperors who have not been recognised.
Perhaps we should also consider the possibility of a harder lead alloy being used at this
time which has resulted in better preserved portraits?
The second highest figure, although a long way behind the first, belongs to Period
L (Gordian 111, Philip I and Philip II, Trajan Decius and family, Trebonianus Gallus),
with Periods H (Marcus Aurelius and family) and K (Severus Alexander and Maximinus)
a close third and fourth respectively.
Comparison of sealings assigned to emperors by inscription and

portrait
If we put the information from Figs. 5 and 6 together in one histogram it will help
us to see the differences at a glance:
18
16
14
12
10
LI INSCRIPTION
8
El PORTRAiT
6
ABCDEFGHI JKLMNOPQRSTU
Fig 7 Imperial scalings per thousand identified from (i) mscnptlon and (u) portrait alone
153
If we compare the totals of actual numbers of scalings in Tables I (30 scalings)
and 11(1 17 scalings) we see that there are nearly four times as many identified by portrait
alone as by inscnption. This difference is unlikely to be due to selective publication since
epigraphic examples would be preferred (although that is not to say that all scalings
identified by portrait are anepigraphic). It is also unlikely to be due to busts of private
individuals or deities being included accidentally since, while this is possible in itself, I
have also had to omit from these figures many undoubtedly imperial scalings on which
the portraits could not be recognised, thus hopeftilly balancing any inaccuracies.
The main reasons that I can see for the prevalence of scalings which do not name
the depicted emperor(s) are: (i) there are too many busts shown for the names to be
fitted in (this applies to 60 out of 117 examples); (ii) other information is included e.g.
R(alio) C(asfrensis) or Anabolici, thus limiting the space for names (this applies to 21
examples out of the 57 remaining scalings after reason (i) has been taken into account);
(iii) if a bust was virtually a prerequisite then the only way to make matrices more
quickly and cheaply was to omit further detail in the form of a tiny inscription; (iv) lead
might not always give a clear enough impression to warrant the inclusion of a small
inscription whereas a bust could still be recognised.
If we take reasons (i) and (ii) into consideration and only use the 36 scalings
which depict one emperor without naming him, we now have much closer numbers of
scalings identified by portrait alone and those identified by inscription. However, the real
value in doing this lies in the fact that we are now comparing like with like - these 36
scalings were not left with the emperor unnamed because there was no space but as a
conscious decision.
154
The calculation being used here is:
imperial scalings of single emperor

without other inscription, id. from portrait, per period X 1000
length of period total of imp. scalings for empire
of single emperors without other
inscription id from portrait
In this case the figures and records used to obtain them would look like this:
Table III

Period Actual quantity & Record Nos. %
A (3OBC-AD4I) 1 [0076] 0.28

B (41-54) 0 0
C (54-69) 1 [0276] 1.94
D (69-96) 1 [0183] 1.11
E (96-117) 1 [0085] 1.39
F (117-138) 2 [0248-0249] 2.78
G (138-161) 1 [0280] 1.11
11(161-180) 0 0
1(180-193) 2 [0078; 0082] 4.17
J (193-222) 9 [0047; 0077; 0081; 0083-0084; 8.61
0105; 0185; 0196; 0277]
K (222-23 8) 6 [0079(2 ex.); 0080; 0124; 0134; 0184] 10.56
L (23 8-260) 7 [0075(2 ex.); 0165; 0210; 0228; 8.89
0278-0279]
M (260-284) 1 [0182] 1.11
N (284-3 10) 1 [00741 111
0(310-337) 1 [0125] [11
P (337-3 60) 0 0
155
Q (3 60-3 75) 2 [0073; 0197] 3.61
R (3 75-400) 0 0
S (400-425) 0 0
T (425-455) 0 0
U (45 5-475) 0 0
Total 36
15
14
13
12
11
10
9
8
LI 1NSCRIVflON
7
6 [1 PORTRAIT
5
4
3
2
ABCDEFGHI JKLMNOPQRSTU
I ig S hnperial scalings per thousand idcnti(ied from (i) insuiption and (ii) portrait alone, the latter omitting
examples bearing any other inscription or multiple busts
This histogram would appear to show that their is a distinct division between
periods when sealings with and without an identifying inscription of the emperor were in
use. 3 Eight periods contain only sealings identified by portrait alone while four periods
have only sealings identified by inscription. The remaining six periods (ignoring three
156
blank periods) have a mixture, although in these periods the named examples are usually
in the majority, never the minority. The exceptions to this are Periods D and M where the
two varieties have virtually the same figure.
Looking at this joint histogram we can also see that the sealings identified by
portraits alone are spread more widely over the years than those identified by inscription
(14 periods out of 21 have sealings identified by portrait alone whereas 10 out of 21 have
those identified by inscription). This is particularly noticeable in the Periods 1-L
(180-260). Sealings identified by inscriptions are completely unrepresented in these four
periods which include the highest, second, third and fourth highest figures for those
identified by portraits alone. It is true that many of the identifications in Periods K and L
are debatable although hopefully they are broadly correct, some possibly being
interchangeable between the two periods. Interestingly, Period J has been toppled from
its significant lead in the previous histogram. This is almost entirely due to the fact that
so many of the Severan examples bore the multiple busts of Severus, Caracalla and Geta.
Conclusions
While it is true that the number of Imperial sealings which were deliberately
struck without the name of the emperor (i.e. not for reasons of lack of space due to extra
busts or other information being included) is similar to the number which were struck
with his name inscribed, there is an apparent dichotomy between the periods in which
these two types of sealing were in use. On present information it is impossible to tell
whether this difference is just one of artistic preference or if it suggests differing uses or
needs in certain periods. We should also consider whether this difference could just be a
geographic one, although this is difficult to prove since it would require dividing the
sealings into even smaller samples (separate provinces within each period) than the
already dangerously low numbers on which the above work was based. On a slightly
different note, I have examined the findspots of the examples used in Tables I and II and
157
there does not seem to be any obvious division of sealings identified by inscription or
portrait between certain sites or type of site so this possibility can probably be ignored.
158
Notes
1. Possibly due to its close relationship with Moesia Inferior, forming a hinterland to that
frontier zone, and/or because of the constant movement of troops through it, from east
to west and vice-versa.
2. The amount of sealings known in all categories from all over the empire (a little over
2000) suggests that there must have been many more originally which have been lost to
us. Byzantine scalings were apparently re-used to a certain extent (Oikonomides, 198Th,
101).
3. However, when working with so few scalings, such a polarisation should be expected,
since the limited number of scalings can only fit into one category or another.
159
Chapter 12
PROVINCIAL DISTRIBUTION OF SEALINGS

- TIlE STATISTICAL EVIDENCE
In this chapter we shall use the evidence of the main part of the catalogue to
examine the relationship between the different categories of sealings within each
province. We shall do this by looking at the numbers of sealings in each category as a
percentage of the total of all categories for that province Unfortunately, the limited
number of examples puts too much emphasis on just a few sealings in some provinces.
Another drawback is that, as explained at the beginning of chapter 9, in addition to
sealings bearing Iria nomina and anepigraphic sealirigs with the images of deities and
animals, the Miscellaneous category also contains sealings with unidentifiable inscriptions
which may actually belong in other categories.
By way of comparison, we shall first look at a histogram produced from figures

which represent a breakdown of sealings in the catalogue from the empire as a whole
into their correct categories. The first set of figures and histogram will include the large
collection from Gallia Lugdunensis but we shall then view the same information without
the strong influence of this material.
Table IV: Whole empire including sealings from Callia Lugdunensis

(Total: 2304 examples)
Number of sealings within each category Percentage of total

Imperial (318 examples) 13 8%

Official (13 examples) 0.56%

Taxation (12 examples) 0.52%
160
Provincial (18 examples) 0.78%
Civic (69 examples) 2.99%
Legionary (72 examples) 3.13%
AIae (10 examples) 0.43%
Cohorts (161 examples) 6.99%
Beneficiarii Consularis (3 examples) 0.13%
Equites Singulares (2 examples) 0.09%
Pedites Singulares (3 examples) 0.13%
Miscellaneous (1623 examples) 70.44%
Fig 9 Petcentagec of scalings in dilferent categones for whole empire including (jallia I ugduncnsis
161
Table V: Whole empire excluding sealings from Gallia Lugdunensis
(Total: 1262 examples)
Number of scalings within each category Percentage of total
Imperial (253 examples) 20 05%

Official (10 examples) 0.79%
Taxation (3 examples) 0.24%
Provincial (17 examples) 1.35%
Civic (67 examples) 5.3 1%
Legionary (53 examples) 4.2%
AIae (10 examples) 0.79%
Cohorts (150 examples) 11.89%
Beneficiarii Consularis (3 examples) 0.24%
Equites Singulares (2 examples) 0.16%
Pedites Singulares (3 examples) 0.24%
Miscellaneous (688 examples) 54. 52%
162
Fig. 10 Percentages of scalings in different categories for whole empire excluding GnIIia Lugdunensis.
Both Figs.9 and 10 present us with a similar story, although Fig. 10 avoids the
excesses of the Gallia Lugdunensis totals. However, we can see that the basic hierarchy
of the categories is hardly affected by the presence of the large Miscellaneous collection
from Gallia Lugdunensis (although the percentage of Legionary sealings was slightly
greater than that of Civic) since Miscellaneous sealings are the most common anyway.
Having thus satisfied ourselves that the inclusion of Gallia Lugdunensis will not ruin our
results, we will include that province in our future calculations for the empire total of
each category.
163
We shall now examine the sealings found in each separate province in order to
see whether this can tell us anything about the nature of that province.
The column entitled 'Percentage of provincial total' gives us the scalings in each
individual category found in that province expressed as a percentage of the total number
of scalings of all categories found in that province.
The second column, entitled 'Whole empire category total as percentage of

adjusted whole empire total' gives us the scalings in the whole empire categories (Table
1) expressed as a percentage of the whole empire total, but, in order to compare like with
like, this total ignores the other categories not found in the province in question, e g.
Africa Proconsularis contains only Imperial and Miscellaneous scalings so we refer back
to Table I and add together the whole empire totals of Imperial (318 examples) and
Miscellaneous scalings (1623 examples) to make an artificial whole empire total (1941
examples). This is then used to calculate what percentage the two separate whole empire
categories are (16.38% and 83.62% respectively). This gives us a percentage figure
which can be compared directly with the percentage figures listed under the individual
province, without the interference of other categories which are not found in that
province. I believe that this technique is valid since direct comparison with the figures in
Table I would have led to the ratio of percentages being tainted by the other categories
which were not applicable to the province in question. It is possible to separate the
categories since scalings in the individual categories would have entered, or circulated in,
provinces for separate reasons and would not necessarily have been affected by increased
use by one organisation or another. Therefore this calculation and comparison seems fair.
Since the figures in the first and second columns are comparable, I have also
noted the occasions on which the ranking of percentages by size within a particular
province (e g Miscellaneous (70%) - first; Imperial (20%) - second etc.) differs from
those which the whole empire figures would suggest is normal for a collection of scalings
consisting of those categories.
164
Obviously, it is useless to include provinces which have produced sealings from
only one category in these calculations since both columns will consistently show 100%
These provinces have been noted in the appropriate places but no further action has been
taken.
Africa (only one Miscellaneous sealing)
Africa Proconsularis (Total: 217+ examples)
Number of sealings Percentage of provincial Whole empire

within each category total category total as
percentage of
adjusted whole
empire total
Imperial (2 examples) 0.92% 16.38%
Miscellaneous (215+ examples) 99.08% 83.62%
The two columns of figures produce the following histogram:
165
Fig II Percentages of scalings in different categories in Africa Proconsularis (tmshaded) compared with category
total for the whole empire expressed as a percentage of the whole empire total for those two categories (shaded).
Comparison of the first and second columns of figures (and inspection of the
resulting histogram) shows that Africa Proconsularis has a seventeen and a half times
lower than normal percentage of Imperial scalings and a higher than normal percentage
of Miscellaneous scalings.
When the actual number of scalings from the province is taken into consideration,
the percentages do appear to show a real difference to those based on the whole empire.
The ranking of the percentages for the province is also the same as for the whole empire.
166
Asia (Total: 22 examples)

percentage of
adjusted whole
empire total

Miscellaneous (19 examples) 86.3 6% 83.62%
Fig 12 PercenIiges oreaIings in dillerent categories in Asia (unshaded) compared with the category total for the
whole empire expressed as a percentage of the whole empire total for those two categories (shaded)
167
resulting histogram) reveals that Asia has a slightly lower than normal percentage of
Imperial sealings and a slightly higher than normal percentage of Miscellaneous.
Although the total number of actual sealings from this province is not great, the
percentages are very close to those based on the whole empire The ranking of the
percentages for the province is also the same as those for the whole empire.
Britannia (Total: 372 examples)

percentage of
adjusted whole
empire total
Imperial (35 examples) 9.4 1% 13.8%

Official (2 examples) 0.54% 0.56%
Taxation (1 example) 0.27% 0.52%
Provincial (8 examples) 2.15% 0.78%
Civic (4 examples) 1.08% 2.99%
Legionary (46 examples) 12.37% 3.13%
Alae (10 examples) 2.67% 0.43%
Cohorts (150 examples) 40.32% 6.99%
Beneficiarii Consularis (3 examples) 0.8 1% 0.13%
Equites Singulares (2 examples) o 54% 0.09%
Pedites Singulares (3 examples) 081% 0 13%
Miscellaneous (108 examples) 29 03% 70.44%
168
Fig 13 Percentages of scalings in different categories in Britannia (unshaded) compared with the category total for
the whole empire expressed as a percentage of the whole empire total for those twelve categories (shaded).
Britain is the only province to have sealings belonging to all categories and
therefore the figures in the second column are exactly the same as those in Table IV.
Comparison of these figures with those in the first column (and inspection of the
resulting histogram) shows that Britain has an almost one and a half times lower than
normal percentage of Imperial sealings, the normal percentage of Official sealings, half
the normal percentage of Taxation sealings, three times the normal percentage of
Provincial sealings, half the normal of Civic, four times the normal of Legionary, six
times of those of Mae, nearly six times of those of Cohorts, six times the normal
percentage for those of Beneficiarii Consularis, Equites Singulares and Pedites Singulares
and nearly two and a half times less than the normal percentage of Miscellaneous.
169
If it were not for the high percentages of scalings of Cohorts and Legions then
the figures for Britain would look virtually normal. In fact, other military categories have
also distorted the figures since several of the categories have only been identified in
Britain. This all sits well with the general perception of Britannia as a military frontier
province. When we compare these figures with those for the whole empire we can also
see that Provincial sealings and those of Mae are also well represented. The ranking of
the percentages of the province differs from those suggested by the figures for the whole
empire in that Miscellaneous should be first but is second, Imperial should be second but
is fourth, Cohorts should be third but are first, Legionary should be fourth but are third,
Civic should be fifth but are seventh, Official should be seventh but are eighth, Taxation
should be eighth but are ninth and Ala should be ninth but are fifth.
Dalmatia (Total: 5 examples)

percentage of
adjusted whole
empire total

Legionary (I example) 20% 4.25%

Miscellaneous (4 examples) 80% 95.75%
The two columns of figures produce the following histogram.
170
I ig. 14 Percentages of scalings in different categories in Dalinatia (unshaded) compared with the category total for
the whole empire expressed as a percentage of the vhole empire total for those two categories (shaded).
Comparison of the first and second columns (and inspection of the resulting
histogram) shows that the percentage of Legionary sealings is almost five times the norm
(bearing in mind of course that we are talking about just one sealing) while that of the
Miscellaneous scalings is below the norm.
The facts that the actual numbers involved here are so small and that there are
only two categories means that it is almost impossible to say anything constructive about
the sealings in Dalmatia. The absence of Legionary scalings over much of the empire has
thrown undue weight on to this one example. It is interesting that even one Legionary
sealing has been found here, considering their general empire-wide paucity. The ranking
of the percentages for this province is the same as that suggested by the figures for the
whole empire.
171
GaHia Belgica (Total: 53 examples)
Number of scalings Percentage of provincial Whole empire

percentage of
adjusted whole
empire total

Miscellaneous (3 1 examples) 58.49% 79. 52%
rig i Percentages of scalings in different catcgoncs in Gaiha IJelgica (unshaded) compared with the catcgory total
for the whole empire expressed as a percentage of the whole empire total for those five categories (shaded)
172
resulting histogram) shows that for Gallia Belgica the percentage of Imperial sealings is
twice the norm, of OfficLal sealings almost six times the norm, of Provincial scalings four
and a half times, roughly normal for Civic and nearly one a half times lower than normal
for Miscellaneous scalings.
The fact that that the percentage of Imperial scalings is higher than is normal for
the whole empire is presumably connected with Trier's role as an administrative
centre/imperial capital. The similar effect for the percentage of Official scalings could be
explained as due to Trier's place as the headquarters of the procurator of Belgica and the
two Germanies and subsequently of the Prefect of the Gauls. These points may also
count for the higher than normal percentage of Provincial scalings. It may be that the
lower than average percentage of Miscellaneous scalings is caused by these higher
figures mentioned above. The ranking of the percentages for the province differs from
that suggested by the figures for the whole empire in that the Civic, Provincial and
Official categories should all be joint third whereas they are in fact third, fourth and fifth
respectively.
Gallia Lugdunensis (Total: 1042 examples)

percentage of
adjusted whole
empire total


Taxation (9 examples) o 86% 0.53%

Provincial (I example) 0.1% O.79%
173

Civic (2 examples) 0 19% 3 04%

Legionary (19 examples) 1 82% 3 17%

Cohorts (11 examples) 1 06% 7.08%

Miscellaneous (935 examples) 89. 73% 71.4%
The two columns of figures produce the following histogram
Fig 16 Percentages of scalings in different categories in (lallia I ugdunensis (unshaded) compared with the category
total for the vholc empire expressed as a percentage of the whole empire total for those six categories (shaded)
Comparison of the first and second column of figures (and inspection of the
resulting histogram) reveals that for Gallia Lugdunensis the percentage of Imperial
sealings is half the norm, of Taxation sealings one and a half times above the norm, of
Provincial sealings eight times less, of Civic sealings sixteen times less, of Legionary
174
scalings nearly half, of scalings of Cohorts nearly seven times less than normal and of
Miscellaneous scalings just above normal.
The percentage of Imperial scalings in Gallia Lugdunensis appears small against

that of Miscellaneous scalings and this is confirmed by the figures for the whole empire
It should be remembered that the figures for the whole empire include this unusually
large collection of scalings from Gallia Lugdunensis and so, for example, the huge
number of Miscellaneous scalings cancel themselves out and tend to make Gallia
Lugdunensis look more normal than it really is. The ranking of percentages for the
province differs from that of the whole empire in that Cohorts should be third but are
fourth, Legionary should be fourth but are third, Civic should be fifth but are sixth,
Taxation should be sixth but are fifth and Provincial should be sixth but are seventh.
Gallia Narbonensis (only two Miscellaneous sealings)
Germania Inferior (only four Miscellaneous sealings)
Germania Superior (Total: 5 examples)

percentage of
adjusted whole
empire total

Imperial (I example) 20% 15.8%
Legionary (1 example) 20% 3.58%
175
Fig 17 Percentages of sealings in diflérent categories in Gennania Superior (unshaded) compared with the category
total for the whole empire expressed as a percentage of the vhole empire total for those three categories (shaded)
Comparison of the first and second column of figures (and inspection of the
resulting histogram) tells us that Germania Superior has a slightly above normal
percentage of Imperial sealings, of Legionary sealings five and a half times the normal
percentage and nearly one and a half times below normal of Miscellaneous sealings.
Yet again, here we see small numbers of actual sealings taking on inflated
importance as percentages when there is really too little information to assess. The
176
ranking of percentages for the province differs from that suggeited by the figures for the
whole empire in that the Imperial category should be second and the Legionary third
whereas they are both joint second.
Italia (Total: 119 examples)

percentage of
adjusted whole
empire total

Taxation (2 examples) 1.68% 0.59%
Civic (1 example) 0.84% 3.4%
Miscellaneous (66 examples) 55.46% 79.75%
177
Fig 18 Percentages of sealings in different categories in Itaia (unshaded) compared with the category total for the
whole empire ecpressed as a percentage of the whole empire total for those five categories (shaded)
resulting histogram) shows that for Italia the percentage of Imperial and Official sealings
is two and a half times above the norm, nearly three times the norm of Taxation sealings,
four times below the norm of Civic sealings and nearly one and a half times below the
nonn of Miscellaneous scalings.
The figures for Italy show that the percentage of Imperial scalings is rather high
in relation to that of the Miscellaneous scalings. The figures for the whole empire
confirm this and also reveal the higher than normal percentages of Official and Taxation
scalings. Again, the percentages of these other categories profit at the expense of the
Miscellaneous scalings. The ranking of the percentages for Italy differ from those
suggested by the figures for the whole empire in that Official should be fourth and
Taxation fifth whereas they are both joint third with Civic, which should be third, in fifth
178
position. The higher than average percentage of Imperial, Official and Taxation scalings
fits well with the Rome (or ?Rome) provenance for most of the Italian material.
Melita (only 3 Miscellaneous examples)
Moesia Inferior (Total: 158 examples)

percentage of
adjusted whole
empire total

Official (1 example) 0.63% 0.62%
Civic (30 examples) 18 99% 3.27%
Legionary (4 examples) 2.53% 3.4 1%
179
rig i 9 Percentages of scalings in different categories in Moesia Inferior (unthaded) compared with the category
total for the whole empire expressed as a perccntagc of the whole empire tota1 for those six categories (shaded).
resulting histogram) reveals that for Moesia Inferior the percentage of Imperial sealings
is nearly twice the norm, of Official sealings normal, of Provincial sealings three times the
norm, of Civic sealings nearly six times, of Legionary scalings just below normal and of
Miscellaneous sealings over one and a half times less than the norm.
Looking at the percentages for the province we can see that the Civic scalings are
more important than usual. Comparing these figures with those for the whole empire
shows that they are indeed special but that this also applies to the Imperial scalings. The
higher than average percentage of Imperial sealings may be caused by the frontier nature
of the province with the campaigning that this brings about or, if we have misunderstood
the function of these sealings, it could represent the delivery of produce from imperial
180
estates in, for example, Asia Minor. We know that a great deal was imported from that
area and this is demonstrated in the much higher than average percentage of Civic
sealings, the majority of which come from Asia Minor. The ranking of the percentages
for the province are the same as those based on the whole empire with the exception of
the Civic sealings which should be fourth but have moved into third position, displacing
Legionary.
Noricum (Total: 11 examples)

percentage of
adjusted whole
empire total

Miscellaneous (5 examples) 45.45% 83 .62%
181
Fig 20 Percentages of scalings in different categories in Noricum (unshaded) compared with the category total for
the whole empire expressed as a percentage of the whole empire total for those two catcgones (shaded).
resulting histogram) shows that Norictim has over three times the normal percentage of
Imperial sealings and nearly half the normal percentage of Miscellaneous sealings.
Again we have small numbers of actual sealings divided between only two
categories which produce misleadingly impressive differences in percentage. However,
the strange phenomenon of a higher percentage of Imperial than Miscellaneous sealings
is highlighted when the provincial figures are compared with the normal figures for the
whole empire. On account of these figures, the ranking of the percentages of the
province have been reversed when compared with the whole empire
182
Numidia (Total: 16+ examples)

percentage of
adjusted whole
empire total
Imperial (?3+ examples 18. 75% 115.82%
- source says 'several')

Civic (5+ examples) 3 1.25% 3.43%
I ig 21 ['cicentages of scalings in dillcrent categories in Numidia (unshaded) compared with the category total for
the 1ioIe empire expressed as a percentage of the 'lole empire lo(aI for those tso categones (shaded)
183
resulting histogram) tells us that Numidia has nine and a half times the normal percentage
of Civic scalings and one and a half times less than the normal percentage of
Miscellaneous scalings. If our assumption is correct as to the approximate number of
Imperial scalings, then Numidia has about the normal percentage.
The percentage of Civic scalings is much closer to that of Miscellaneous than

normal and comparison with the figures based on the whole empire show that they are
much more important than is usual. This is of interest because, rather unusually, they all
name the city in which they were found [0364]. The ranking of the three categories in the
province still has Miscellaneous at the top but Civic has swapped places with Imperial.
Pannonia inferior (only 1 Miscellaneous example)
Pannonia Superior (Total: 30 examples)

percentage of
adjusted whole
empire total
Imperial (15 examples) 50% 15.72%

Official (2 examples) 6.6 7% 0 64%
Civic (I example) 3.33% 3.41%
184
Fig22 Percentages of scalings in different categories in Pannonia Superior (unshaded) compared th the categoly
total for the whole empire expressed as a percentage of the whole empire total for those four categories (shaded).
resulting histogram) shows that Pannonia Superior has three times the normal percentage
of Imperial sealings, ten and a half times the normal percentage of Official scalings, the
normal percentage of Civic sealings and half the norm of Miscellaneous scalings.
The figures for the province show that Imperial scalings were important here,
more so than Miscellaneous scalings. This difference is even more marked when
compared with the figures based on the whole empire. Official scalings also show a high
percentage. The ranking of percentages based on the whole empire suggests that
Imperial and Official should be second and fourth whereas they are first and third
respectively. The majority of the scalings from Pannonia Superior were found in the
frontier area along the Danube, which raises the question of whether or not these
185
Imperial and Official scalings were connected with military campaigns and/or trading
with the barbarians.
Pannonia - unspecified (Total: 61 examples)

percentage of
adjusted whole
empire total

Civic (I example) 1.64% 3.4 1%
Miscellaneous (31 examples) 50. 82% 80.23%
186
Fig.23 Percentages of scalings in different categories in Pannonia - unspecified (unshaded) compared with the
categoy total for the whole empire cxpressed as a percentage of the whole empire total for those four categories
(shaded).
resulting histogram) reveals that for Pannonia (without any further clarification) the
percentage of Imperial scalings is over two and a half times the norm, of Official sealings
over ten times the norm, of Civic scalings half the norm and of Miscellaneous scalings
one and a half times below normal.
While Pannonia (without any further clarification) has the same range of four
categories of scalings as Pannonia Superior, it does display some slight differences. The
Miscellaneous scalings display a slightly greater percentage than the Imperial although
comparison with the percentages based on the whole empire show that the Imperial
scalings are still much more important than is nonnal. This again holds true for the
Official scalings. In fact, the Official scalings are the only category which, for this
187
province, break the ranking of percentages based on the whole empire. They should be
fourth (i.e. lowest) but are actually third. Despite the fact that we do not know the exact
findspots of any of these sealings it could be suggested that they are from a similar region
to those of Pannonia Superior i.e. the frontier area of the Danube, either in Pannonia
Superior or Inferior.
Syria (Total: 11 examples)

percentage of
adjusted whole
empire total

Civic (1 example) 9.09% 3.3 1%
Legionary (1 example) 9.09% 3.46%
188
Fig.24 Percentages of sealings hi different categories in Syria (unshaded) ccanpared with the categoty total for the
whole empire expressed as a percentage of the whole empire total for those four categories (shaded).
resulting histogram) shows that Syria has a slightly above normal percentage of Imperial
sealings, over two and a half times the normal percentage of Civic and Legionary sealings
and below the normal percentage of Miscellaneous sealings.
The low numbers of actual sealings involved make it difficult to say anything
usefi.iI here, although the ranking of percentages within the province is virtually the same
as for those based on the whole empire.
189
Thracia (Total: 117 examples)

percentage of
adjusted whole
empire total
Imperial (40 examples) 34. 19% 15.68%

Provincial (2 examples) 1.7 1% 0.89%
Fig 25 Percentages of scalings m different categories in Thracia (unshaded) compared with the category total for
the whole empire expressed as a percentage of the whole empire total for those four categones (shaded.
190
resulting histogram) shows that Thrace has over twice the normal percentage of Imperial
sealings, twice the normal percentage of Provincial sealings, five times the normal
percentage of Civic scalings and nearly half of Miscellaneous scalings.
The percentage of Imperial scalings in Thrace is not far behind Miscellaneous and
that of Civic is not far behind Imperial. When we compare the provincial figures with
those based on the whole empire we see that both Imperial and Civic scalings are more
important than is usual. The ranking of the percentages of the province is the same as for
those based on the whole empire.
Unknown Provinces (Total: 54 examples)

percentage of
adjusted whole
empire total

Provincial (1 example) 1.85% o 89%
Civic (2 example) 3 7% 3.4%
191
Fig.26 Percentages of sealings in different categoiies from unknown provinces (unthaded) compared with the
category total for the whole empire expressed as a percentage of the whole empire total for those four categories
(shaded).
resulting histogram) shows that this collection of scalings from various unknown
provinces has nearly one and a half times more than the normal percentage of Imperial
sealings, over twice the normal percentage of Provincial scalings, the normal percentage
of Civic scalings and roughly the normal percentage of Miscellaneous scalings.
These figures, particularly the last two, for scalings from various unknown
provinces are not too dissimilar to the average and as such should not cause any alarm.
As expected from a random collection, they seem to reflect the overall trends noticeable
in the figures for the whole empire. The ranking of the percentages is the same as for
those based on the whole empire.
192
Conclusions
As mentioned earlier, the quantity of sealings in most provinces is insufficient to

provide reliable results in work of this kind It is interesting to attempt the comparison of
the nature of usage of scalings in these provinces but no detailed hypotheses should be
built on such unstable ground
193
Chapter 13
TIlE ICONOGRAPHY OF ROMAN LEAD SEALINGS
This chapter will examine the types of designs found on lead scalings of all
categories
Imperial portraits
Many imperial portraits are found on lead sealings. These are sometimes, but not
often, of a similar quality to those seen on coins. An overview of the composition of
portraits found on scalings reveals an important difference, in that there appear to be
relatively more scalings bearing multiple busts (two, three or four) than is the case for
coins. I have covered this more ftilly in chapter 11 but some of the main points should be
mentioned here.
The examples in question are as follows: those showing two busts are [0036 -
0040]; [0042]; [0069 - 0071]; [0086]; [0094]; [0096]; [0100]; [0127 - 0133]; [0151];
[0159]; [?0178]; [0197 - 0199]; [0202A]; [0205 - 0206]; [0213](jugate busts); [0214 -
02231; [0231]; [0233 -02341, [0239- 0240]; [0252-0259]; [0266]; [0268]; [02811.
Those with three busts are: [0007A - 0008]; [0015]; [0021 - 0022]; [0053 -
0054]; [0058]; [0060]; [0087]; [0108]; [0152 - 0156]; [0166]; [0203 - 0204]; [0224];
[0227]; [?0239]; [0260]; [02631, [0273]; [0282].
Those with four busts: [0043]; [0157 - 0158]; [0167]; [0180]; [0200 - 0202];
[0225]; [0261 -0262].
The presence of only one sealing showing jugate busts [0213]' serves to highlight
the more common technique of portraying busts facing each other, presumably in order
194
to ensure that the impressions were recognisable on lead. The sealings with four busts
are extremely interesting since the various compositions are apparently unknown on
coins. The most important of these are the impressions depicting the emperors of the first
tetrarchy in two registers with each bust facing its partner [0180], [0200 - 02021, [0225],
[02611. This is certainly never seen on their coins which usually show a single emperor.
Coins with different obverses would have been produced for all four members but could
have been used regardless of the bust whereas it appears that a need was felt for the
goods sealed to be despatched in the name of the tetrarchy as a whole.
There are many other imperial portraits which deserve special mention by virtue
of the fact that there are other objects in the field. Some simply show items being held by
the person portrayed, for example [0004] with a hand holding a ?flower in front of the
bust, [0005] with a caduceus behind the bust and an unidentified shape in front, [0226]
also with caduceus, [0177] with a sceptre behind the bust and [0182] with a shield at the
side of the bust and a horse in front, as seen for example on a bronze medallion of
Maximianus Herculeus (Vermeule, 1978, 182 and p1.19, 2). Other impressions bear
symbols, for example [0084] with crescent and star, [0174] with a six pointed star, [0220
- 0221] with a star and [0176] with a round ornament consisting of a central sphere
surrounded by dots
One bust is accompanied by a long-legged bird (?Phoenix) [02801 while another

sealing shows four busts with a vertical dolphin placed centrally between them [0262].
Gerasimova-Tomova states that the busts are of Septimius Severus and his family and
that the dolphin is of the same type with which the coinage of Byzantium was
countermarked and that therefore this must be where the sealing originated (1994, 377-8,
no.16). I have not, however, been able to find a dolphin counterniark pertaining to
Byzantium (Howgego, 1985).
195
Several busts are depicted with Victories who are crowning them with wreaths
[0095], [0229], [0245 - 0246], [0251]. Another example shows two busts separated by a
Victory [0281].
Not all of the objects can be identified with certainty, including those found in the
following examples. [0206] (unless D and K ligatured), [0242] with shape in front of
bust (could be a quiver of arrows), [0269] with objects above, in front and behind and
[0270] with an object in front (possibly a club).
Religion/mythology
Deities, Personifications and mythological figures
Enough has been written in the past on the detailed iconography of

Graeco-Roman deities and so this section will be restricted to factual information on
which deities are present, unless there is any direct relevance to the figures' use on
sealings. As in other sections, the evidence of sealings produced from purpose-made dies
has not been separated from those impressed by ring intaglios. The Catalogue, however,
will, in many cases, provide information as to size, shape and sometimes inscriptions
which will enable the reader to gain some idea of the likely nature of the original matrix.
It can easily be seen that the deities depicted on lead scalings cover practically the
whole classical pantheon and a few more besides. By and large these are the same gods
and goddesses found on Roman imperial coinage and there are very few ostensible
representations of specifically local or pre-Greek gods as are sometimes encountered on
the provincial coinage (Butcher, 1988, 50). The reason for this may be that the majority
of civic sealings are purely epigraphic and therefore do not provide us with any
illustrations of the cities' tutelary deities
196
Of the originally Eastern religions which gained acceptance in Rome, that of
Serapis [08401, [1019], [1268], [1270-72], [12741, [1518], [1541-3], Isis [0840], [1263
- 64]. [1274], [1518] and 1-Jarpocrates [1274], [1543], [1582], [1794] is shown on
several sealings, as is that of Asciepius [1383], [1630], [1659]. The god Bacchus appears
on two examples, [0953] and [1768].
The personifications shown on sealings are Fortuna, Pietas, Salus, Spes, Tyche (I
have used the Greek form to denote examples of Fortuna wearing a mural crown) and
Victory. Most of these are common on coins and gems especially Fortuna and Victory,
both of which are also the most popular on scalings. Out of the 15 Fortuna examples
only four possess inscriptions. One of these is imperial [0018] whereas the other three
clearly belonged to private individuals, [0921], [0986] and [1134]. There are 48
examples depicting Victory, Ii of which bear inscriptions. Nine of these inscriptions are
imperial [00011, [0009], [0042], [0095], [0110], [0114], [0121], [0188 - 0189], while
three are apparently private, [0856], [1073] and [1120]. The mural crowns on the figures
which I have referred to as Tyche may suggest that these were intended to represent
particular cities, although only in the case of one example, [0332], do we have an
interpretable inscription to verify this. The shape referred to as a mural crown, however,
may actually be a modius or calathus which Henig says is 'a Graeco-Egyptian conception
which may link her with Isis' (1974, 99). The majority of Tyche scalings are
anepigraphic but there are some others which bear inscriptions although these are
incomplete [0657] and [1723].
The representations of the traditional divinities are generally typical of those

encountered on intaglios and coins. Apollo is an exception to this in that there are no fill
length figures of him, the only ostensible image being a bust ([0530]).
Diana is also shown as a bust, with quiver over her shoulder ([1734 - 1736]) and
standing with bow ([0662], [1627]).
197
Jupiter is depicted by a bust in two examples [12591 and [1574], the latter
apparently placed on the back of an eagle. He is also shown as a standing figure holding
various combinations of sceptre, thunderbolt and patera, sometimes with accompanying
eagle [0459], [1280 - 1282], [1287 - 12881, 11318], [1348], [1583], [1657] Two
sealings show him with other deities. in [0709] he is standing facing Mars and in [1677]
he is with Hercules. One sealing depicts him as Zeus Ammon [0294].
Mars is the subject of many sealings, although I have included numerous figures
referred to as Virtus by Dissard (1905, 83-6). Only one sealing shows a bust of him
[1105], the majority depicting him as a standing figure holding spear and shield [0842],
[1298], [1322], [1349], [1351 - 13621, [1799] and, more rarely, a parazonium [1319],
[1363 - 1366], [1705]. Two sealings show him holding a small Victory (Victoriola) and
spear [1320], [1350], while another two show him with palm and spear [1299], [1368].
Only one example shows him as Mars Gradivus with a trophy over his shoulder [1367].
One impression shows Mars facing Jupiter, as mentioned above [0709].
There are slightly fewer sealings depicting Mercury than Mars. This is surprising
in light of the fact that Mercury was the god of traders. Thus one might have expected to
find merchants protecting their goods with his image. Several of these impressions show
busts of Mercury with or without his petasus [0226], [06321, [0660], [0839], [0849],
[1042], [1088], [1267] while two examples show him seated, [1289] and [1290]. The
majority, however, are standing figures with purse and caduceus [0379], [0639], [0804 -
0805], [0919 - 0920], [1300 - 1302], [1323], [1369 - 1370], [1742]. Two of these
figures are accompanied by a cockerel [1301] and [1323] Three impressions show
Mercury facing Fortuna [1751 - 1753].
Minerva is represented by one helmeted bust [09611 and three standing figures
holding spear and shield [1198], [1324] and [1371]. However, seated figures with similar
accoutrements are often described as Roma and I have followed this
198
Neptune is depicted on five sealings, all of which show him holding a dolphin and
trident [0821], [09411. [0957], [0994] and [1121] They all bear an inscription which, in
at least two cases, appears to be the name of the merchant.
Venus is represented on four sealings, all of which show her as a standing figure.
One is facing the front with objects on either side [0661], two show Venus Anadyomene
wringing out her hair [0769] and [1305], and the fourth is Venus Victrix [1373].
The lesser divinities and figures from mythology are also found on sealings.
Bonus Eventus is always depicted as a standing figure, usually holding ears of corn and a
dish of fruit [0841], [1314], [1341], [1577] and probably [0707 - 0708] unless these
represent satyrs. It appears to be one of the last two to which Henig refers as a goddess
standing right (1974, no.816). He had not examined it and was apparently following
Richmond (1936, 122, no.4). Bonus Eventus' female counterpart, Ceres, is shown seated
with ears of corn and cornucopiae [1279].
One impression can be identified as Britannia by virtue of the accompanying

inscription [0311].
Cupid is shown in several ways, often as a minor character in the scene [13051,
[1340] and [1631]. On four impressions he is riding on a dolphin [0733], [0762], [0807]
and [1166], on one he is riding a hippocamp [1343] and on another he is driving a
quadriga [1548]. Two of the dolphin examples are almost certainly from the same matrix,
[0733] from Kirmington and [0762] possibly from Ravenscar. Cupid is shown standing
on at least two impressions, holding an arrow [13421 and embracing Psyche [1748].
Genii are shown with patera and cornucopiae on three private sealings which also
name the owner [1076], [1079] and [10811 and on nine impressions which apparently
refer to the customs station at ArIes [02931, [0295 - 0297], [0299] and 0301]. Two other
sealings show a Genius next to an altar.
199
Hercules is represented as a bust in six examples, often with a club and
sometimes wearing a lion skin [0270], [0838], [09621, [1265 - 1266] and [1540]. He is
shown standing with club and lion skin in two examples [1317] and [1575] and
apparently without the lion skin in another [1345]. Only three sealings show the Labours,
one with the Lernaean Hydra [1810] and two with the Nemean lion [1346] and [1514].
One impression shows him accompanied by another deity, Jupiter [16771.
One sealing shows Leda and the Swan [0710], following the identification by
Martin Henig (1974, 103 & no.821). This is an almost abstract composition which Henig
passes over lightly, despite his inspired identification.
Pan is possibly represented by a bust on one example [1767] and as a single

standing figure on another [1325]. Three examples show Pan butting with a goat [0711 -
0713], while one shows him with two Cupids [1631]. Another, probably depicting the
thiasos, may show him dancing, holding a ribbon above his head, with a satyr standing
behind [0665].
Roma is depicted as a seated figure on two sealings [0274Aj and [1700]. It

should be pointed out that standing figures with similar equipment are usually described
as Minerva and I have followed this.
Two sealings show the she-wolf suckling Romulus and Remus [06701, [1810].
On both of these the wolf is facing right. The most extensive use of the wolf and twins
on the standard coinage is apparently the fourth century Urbc Roma issue where the wolf
is facing left. On several 'Greek Imperials' from various Roman colonies in Asia Minor
and Thrace, however, the wolf is facing in the same direction as on our sealings.Despite
this information, we are still left with the possibility that the die-cutter was copying an
Urbs Roma coin and simply neglected to cut it in reverse.
200
Salus is depicted as a single figure on two sealings, one showing her holding a
flask and snake [0640] and another with an altar and snake [0664]. She also appears with
Asciepius on three sealings [1383], [1630] and [1659].
Satyrs are represented by one sealing showing a bust [1740] and another showing
three dancing together [1384]. There is a possibility that two of the scalings listed above
under Bonus Eventus may be satyrs [0707 - 0708].
Sol appears as a radiate bust on two identical sealings [0705 - 0706] and as a
standing figure on four others [0882], [1657], [1703] and [1743]. One sealing shows the
bust of Sol facing the bust of Luna with an eagle, lion and snake also in the field [12731.
There are many impressions showing single or multiple figures which probably
represent deities or characters from mythology. Among the more interesting are a
draped, bearded figure seated right, resting his chin on his right hand with his elbow on
his knee and a possible scroll in his left hand [1722]. He is facing a small figure (or
statue?) standing left, which appears to be holding a palm branch in its raised right hand
and an unidentified object in its left hand. The main, pensive figure in this scene is
extremely well delineated. Another sealing has a similar composition but probably refers
to a different story. In that impression the main seated figure is facing left and reaching
out towards a small statue (?) standing right inside an arch [17411. This scene may be
connected with Diomedes seizing the Palladium from Troy, although the iconography is
slightly different to that normally seen on intaglios.
Christianity
The Good Shepherd was a iepresentation of Christ which had pagan antecedents
but the sealings all follow the accepted Christian iconography of a figure, with a sheep
over his shoulders, standing facing the front between two sheep [1655], [17021 and
[1745] A similar composition, dating to the early third century, can be seen in a fresco in
201
the crypt of Lucina in the Catacomb of St Callixtus (Strong, 1988, 258). The sealings,
however, probably belong to the fourth if not the fifTh century.
A similar composition is found in sealings depicting Daniel in the lions' den

where a central figure, this time with his arms raised, is placed between two animals
[1620], [1687], [1729], [1746], [1798]. There is a possibility, due to the amorphous
nature of the animals, that some of these scenes represent St.Menas with two camels, as
shown on terracotta flasks from Alexandria (Metzger, 1981, passim). The localised
sphere of influence of the St.Menas story and the universal knowledge of the Daniel
story suggest that the majority, if not all, of the scalings represent the latter. It is not
impossible, however, for some of these sealings to have originated in Alexandria
although we should remember that representations of St Menas are usually limited to
these flasks. Similar compositions of Daniel in the lions' den can be seen in a fresco in
the crypt of Lucina in the Catacomb of St.Callixtus, dated to the early third century
(Strong, 1988, 258-9) and also in two frescoes in the Via Latina Catacomb dated to the
first half of the fourth century (Ferrua, 1991, 64 & 145). The scalings probably date to
the fourth or fifth century.
There is a possibility that [1619] and [1629] may show David killing the lion,
although this is far from certain.
Other Christian symbols appear on lead scalings, such as fish and anchors. Since
these are also open to more prosaic interpretation, I shall limit the list to those examples
which show the more obviously Christian symbol of anchor and fish together [0651],
[1122] and [ 1499 J.
The Agnus Del may possibly be seen on three sealings with the Chi-Rho above,
although neither the identification of the lamb nor of the Chi-Rho are certain [1632 -
1633], [1647]
202
On the subject of the Chi-Rho, it would seem that in the majority of cases in
which this definitely appears on scalings it is being used as an official symbol for the
fourth or filTh century government, and not as a simple declaration of faith. One
exception to this is an example with an anchor placed below the Chi-Rho [0783].
None of the peacocks, doves or dolphins which are shown on scalings are
ostensibly Christian symbols, with the possible exception of one example which places a
small cross above the head of a dove [0864].
Two scalings actually have a cross as the main feature [1590], [1643] and a third
shows a hand holding a cross between the thumb and forefinger [1593]. These all appear
to be fifth century, if not later.
Gnosticism
There are at least four sealings bearing Gnostic impressions. One of these depicts
the cock-headed and snake-footed lao [0982] while another names him while showing a
horned, bearded bust [0981]. The third and fourth bear the same impression which
consists of a vertical line of text crossed halfway down by a horizontal line of text which
shares one letter [08551. This was classed as Gnostic by Dissard perhaps because its
incomprehensibility appeared to label it as a vox magica (1905, 114). There is another
sealing which may possibly bear a Gnostic inscription, 6uvcqut, 'through power' [0764].
Judaism
Two scalings show a menorah [1696], [1765] and a third probably does [0621].
One of these [1764] bears what is presumably Hebrew script on the obverse.
203
Human figures
There is one particular sealing which shows a man fishing [1796]. This is not the
standard representation of a fisherman seated on a rock, but appears to show a man
perched on a pole, a method of fishing still seen today in some countries. In attempting
to research this (fruitlessly) I came across a very similar intaglio design of the
third/second century B.C. which shows a man fishing from inside a snail-shell (Brandt,
1970, no.843, p1.97). The lead sealing is not as clear as the gem design but there are
some very slight differences (such as a foot apparently sticking out from the line of the
pole) which persuade me not to accept the snail-shell identification whole-heartedly.
Animals
The range of animals, real or imaginary, found on sealings is extensive. We find

the following creatures, mostly as single figures but some in multiple groups: antelope,
ape/baboon, bear, bee, bull, camel, capricorn, chimera, cockerel, crab, crayfish, crow,
deer, dog, dolphin, dove, eagle, elephant, fish, fox, goat, griffin, hare, hippocamp,
hippopotamus?, horse, hydra, ibis, lobster?, mouse, owl, ox?, panther?, peacock,
pegasus, phoenix?, pig, ram, scorpion, sheep, shrimp, snail, snake, stag, swan, triton and
(she-) wolf. In addition to this list there are many birds and several animals which cannot
be identified any further.
From these, the numerically outstanding groups are eagles and lions. There are 41
scalings depictmg an eagle as the main subject and 75 showing a lion. Many of the eagles
are in poses unlike those found on intaglios and coins. The eagles on scalings are often
squarely facing the front with wings raised and head turned to one side, whereas those on
coins and gems are usually (although not always) in more realistic poses The prevalence
of the more simplistic pose may be due to a desire to ensure that the device was
recognisable when impressed in the lead. An eagle in a similar pose is found on the lid of
204
a seal box from London (Hall & Merrifleld, 1986, 10, fig.21). This came from a first
century context and was believed, without any particular evidence, to have been used for
government documents.
The majority of the scalings showing lions are from Lyon (58 examples) and all
bear similar impressions. Many include crescents and differing numbers of stars in the
field above the lion. Most just have a crescent, some have a crescent and three stars
[1474 - 1475] and [1477], one has a crescent and seven stars [1476] while another has a
crescent with a single star [1473]. This variety of stars may have some astrological
significance (even the position of the crescent changes in relation to the stars) although
the wide range of numbers may suggest that detailed investigation (e.g. Britannia xxi
(1990), 369, note 30) would be fruitless.
Some of the animal scalings bear scenes which deserve specific mention. One
shows an ape or baboon riding on an elephant [0752]. This may be a parody on Bacchus'
Indian Triumph. Another bears the representation of what appears to be a camel carrying
a load on its back [1708]. The device of a mouse with a whip driving a chariot drawn by
two eagles (or cocks?) is found on three scalings [14811 and is also known from gems.
The representation of a bull on the provincial scalings of Britannia Inferior [0306 - 0307;
0310; 0313] is thought to show that this was the symbol of the province although the
origin of this is uncertain. 2 The similar image of a stag on the scalings of Britannia
Superior [0308 - 0309] is also of uncertain origin.3
Combinations
Many scalings bear combinations of heads and sometimes legs. There are several
different types although many show Silenus heads with cockerel legs and a horse's head
on top. Others simply show conjoined heads
205
An interesting type, not strictly a combination, shows a cock with a human head
dressed as Mercury (1 e. holding a caduceus and wearing a petasus) [0714 - 07171. The
surviving examples are rather faint and it appears to be one of these which Henig has
understandably interpreted as a 'long legged bird standing right' (1974, no.810). The
relevant details, however, can be recognised when compared with an undated example
from Berlin which is described as 'em Hahn als Hermes mit dem Kopfe des Gottes
(geflugelter Petasos), das Kerykeion unter dem FlUgel' (Furtwangter, 1900, vol 1, pl.xlvi,
no.29,; vol.11, 222, no.29).
Inanimate objects
Some scalings have inanimate objects as their main image. One of the most
interesting of these shows four sacrificial implements [1759] Gerasimova-Tomova says
that these implements are found on coins throughout the Republic and up to the third
century (1994, 387, no.55) but I believe that the exact composition found on this sealing
appears only on coins of Nero as Caesar (AD5O-54). Several scalings show ships, [1501
- 15021, [1714], [1760 - 1761]. The last two ships are of unusual form with strongly
curved bows. These are said to be similar to those seen on coins of Lucius Verus from
Philippopolis (Gerasimova-Tomova, 1994, 382) but I have been unable to verify this.
Chariots are also seen on scalings, [0675], [1500], [1548], [1661], [1763]
One sealing [0524] has previously been published as being purely epigraphic with
no design (RIB 2411.246) but there is a possibility that it may in fact bear a very
interesting decoration
This is a sealing marked BFC referring to the beneficiaril consularis. There

appears to be a shape after all three of the letters. The first, between the 'B' and the 'F'
is unidentified but between the 'F' and the 'C' and again after the 'C' there is a shape
which could be some sort of floral decoration. These consist of two spherical objects
206
either side of a stem which opens out into a flower with a tall thin centre piece. However,
with some imagination, these shapes could be interpreted as the special spear
('Benefiziarierlanze') carried by heiieficiarii as the insignia of their rank (Bishop &
Coulston, 1993, 126). These sometimes take the form of flattened spearheads with two
circular holes, symmetrically placed either side of the shaft, cut out. I have not been able
to discover any bearing the flower-shape/crown? (although cf op.ciL, figs.12 & 91) but I
feel that the presence on the sealing of the two circular shapes with a possible spear point
above may be important. The use of this representation on a sealing is not as strange as it
may seem since Bishop and Coulston note that "the spearhead shape was also applied to
baidric fastening plates, decorative belt appliqués and strap terminals". Therefore it was
an accepted decoration for equipment used by the benejIciarii and, were any symbol to
be used on their scalings, this would be the most logical.
Designs possibly from outside of the empire
Some designs are unlike anything encountered on Roman gems or coins and
parallels have had to be sought elsewhere. There are two scalings with similar designs
depicting a plant with three flowers/buds [1713], [1758] These bear some resemblance
to the representation of three pomegranates on the shekel of the First Jewish Revolt
(AD66-70) but the war and subsequent siege would seem to preclude the distribution of
goods from Jerusalem at that time In fact, we find closer parallels in Sassanian
stamp-seals of the fifth century AD which are also thought to show pomegranate plants
(or tulips) (Bivar, 1969, nos.LA6-7; lIenig, 1994, no.1023).
As mentioned in the section on imperial busts, several are depicted with Victories
who are crowning them with wreaths [0095], [0229], [0245 - 0246], [0251]. While they
may have been copying actual Roman examples, it should be pointed out that the
Bosporan kings produced coins with similar designs showing, for example, Constantine
being crowned (Butcher, 1988, 77). This may be important in view of the fact that the
207
last three scalings mentioned above are from Thrace. Many of the Bosporan coins
definitely copied Roman originals showing various joint Augusli facing each other, and
again we are reminded of the quantity of scalings bearing such designs which have been
found in Thrace and Moesia Inferior. The majority of these scalings are probably from
inside the empire but it may be best to keep an open mind
Similar impressions on other items
There are extremely few impressions on other objects which immediately bring to
mind those on our scalings. There is one legend however which is probably worth
mentioning here since it appears on at least two other types of item. The sealing [0209]
apparently bears the inscription SPES IN DEO placed around a bust. This legend is also
found on one side of a tessera with a bust on the other (Rostovtzeff & Prou, 1900, p1.1,
no.4) and, around a Chi-Rho, as a stamp on pewter ingots (RIB 2406.1-8). It seems
likely that the phrase was not a real affirmation of belief and was probably just being used
as an official motto in all of these cases.
Epigraphic styles
It would be extremely helpfi.il if we were able to date lead scalings by the style of
the inscriptions found on them. This, however, is virtually impossible and even
dangerously misleading. On the other hand, epigraphy can provide some information.
One of the most illegible inscriptions on a lead sealing is that on [0235J which
reads AVG / IMP with the last three letters ligatured. On account of its rough execution
this looks almost like an abstract pattern which is unusual for a sealing apparently
referring to the emperor. It is difficult to say which imperial agency would have produced
208
such a sealing, although perhaps we should consider an imperial estate with someone
using an improvised matrix.
Some scalings bear inscriptions which include strange letter forms. Two
examples, [0320] and [03681, include a heavily senfed letter 'T' which has a small tail
curving to the right, giving it the appearance of a retrograde 'J' with a crossbar. This
would appear to have been influenced by a cursive (either Old Roman Cursive or New
Roman Cursive4) or uncial form. Interestingly, the inscription on [0368] is in Greek
whereas that on [0320] is in Latin but refers to a Greek-speaking area (and was probably
engraved by a Greek speaker) so they may both take their form from the lower-case
Greek tau.
Another sealing with a strange letter form is [0321] which has strangely-shaped
'L's. One of these is based on part of the letter '0', extending from c. 11.00 clockwise to
c6.00, while another is similar but with a kink at about 3.00 thus rendering the letter as
the number 3. These are probably based on cursive forms, although the hooked top is
more closely paralleled in Old Roman Cursive which was comparatively rare in Egypt by
the AD28Os (Tomlin, 1988, 87). 1 can find no parallel for the example resembling the
number 3.
Possible examples of matrices
It would appear from the impression of surrounding bezels that many lead
scalings were impressed using gem stones. These are usually the anepigraphic scalings
which bear designs typical of the standard intaglio repertoire. Other dies were probably
made of metal and took the form of stamp-seals. The subject of boulloteria and their
possible use is covered in chapter 2 ('Typology of Roman lead scalings'). That chapter
also includes detailed study of the two-piece dies used for the majority of military
scalings.
209
Another type of matrix for lead sealings which has been posited is the group of
bronze cuboid dies which have been found in Britain. 5 The suggestion that they were
linked with the production of lead sealings was first made by Martin Henig in connection
with the example from Kingscote (1977, 319-21). The slight differences between the
designs on the cube and the reverse types of coinage of the third century led him to
believe that it had not been a die for making counterfeit coins. The intaglios on a later
discovery in Shropshire are, however, more like those on coins (although not exactly
paralleled) since three follow the obverses of coins, with busts and legends, while the
other three are in the style of the reverses of coins, including consulship numbers (Mills,
994, 72-4). Two of the obverses are of Antoninus Pius with the third showing Marcus
Aurelius as Caesar (the latter had been recut and Mills suggests that it may have replaced
another bust of Antoninus). Mills says that the cube could not have been used to make
coins since the striking would have damaged the uppermost die. I would add that it is
unlikely that these were intended to defraud since the high quality of the busts is let down
by the lack of parallel for the legends. If a counterfeiter was going to produce such fine
work one would expect that he would have taken the trouble to ensure that the legends
were correct. The pairing, however, of obverses and reverses is interesting.
Assuming that these two cubes possessed a similar fi.inction (which is far from
certain considering the time-span involved), could they have been used in the
manufacture of lead scalings? I do not believe so on account of the resulting shape of the
impression. These intaglios all have circular pelleted borders (except for one on the
Kingscote cube) but would leave a surrounding square impression in the lead (if the flan
were large enough) since they are engraved on each side of a cube. Virtually all sealing
impressions are the same shape as any border included in their design. In addition, it is
usually only purely epigraphic matrices which have a square or rectangular shape. For the
Shropshire example in particular, the reverses which include numbers of consulships and
salutations as imperator are completely unknown in the sphere of lead scalings. The only
allowance that I am willing to make is that, in the same way that these cubes are rare, so
might be any lead scalings which they produced. New lead scalings found in Britain quite
210
often possess previously unknown impressions and so it is not impossible for sealings
produced by this method simply not to have been discovered yet
There are two interesting matrices which were not for use with lead sealings but
which should be mentioned here.
The first was found in the theatre of Philadelphia in Lydia. It is a bronze matrix,
4.5cm high and 6cm wide. It shows busts of Septimius Severus (facing right) between
Caracalla and Geta and dates to AD209- 11 since Geta is laureate. The inscription reads
KX(cn&oc) BarnXt6ic öckaitp(otoç (Dedeoglu & Malay, 1986, 101-2; AE 1986,
no.672).
The second matrix, also of bronze, has a diameter of 8 6cm and was found in the
Roman customs house in Savaria (modern Szombathely), although I have not been able
to discover whether or not the identity of the building is based solely on the evidence of
the matrix (AE 1968 no.423; Hainzmann & Visy, 199!, 164-5). It bears a retrograde
inscription which Mócsy gave as C(aius) TIT(ius) ANT(onius) PECVLIARIS
COND(uctor) VECT(igalis) OCT(avae) PANN(oniarum) II ANN(o) (conductionis) XII
while Hainzmann and Visy have C(aius) TIT(ius) I ANT(onius) PECV I LIARIS CO I
ND(uctor) VECT(igalium) OC / 1(o) PANN(oniarum) (duarum) I ANN(o) XIII. ørsted

dates this to the second half of the second century AD (1985, 259). It is difficult to say
what this matrix was used to stamp although the letters appear to be raised, suggesting
that it was used in conjunction with ink rather than any fictile material.
On the subject of matrices, we should also look at the evidence found on lead
sealings for the alteration of dies. We have seen above the suggestion that one side of a
bronze cuboid die was completely recut but scalings can only reveal the alteration of
individual letters. The main example is [0271] but this is not a simple matter of recutting.
The 'B' on the rev, seems to be of an inferior quality to the other letters and appears to
have an incuse rectangle around it. If the rectangle on the sealing had been raised then
we could say that the letter on the mould had been recut by having the area around it
211
lowered. The fact that it is incuse, however, must mean that the rectangle on the mould
was raised, possibly suggesting that a new piece of metal, bearing the letter, had been let
into the mould. This may have been done to cover a mistake at the time that the die was
originally being cut, although we would expect the inset letter to be of the same quality
as the others The time spent on the high-quality engraving of the bust could explain the
desperate measures needed to correct the subsequently-discovered spelling mistake.
A similar effect is seen on [0017] which has individual depressions around three
of the letters (0, N and S) forming the inscription Constan[s] P F Au[g], again
suggesting that the matrix had been raised in these places. These depressions, however,
follow the shape of the letters carefully, especially the 0 and the N, and so may have
been left raised on the die in order to emphasize the letters.
Another possible example is [0421] for which Frere says 'the die has made a
depression around R, not all of which survives' (RIB 2411.43) although this does sound
rather different.
212
Notes
1. Another possible example [02691 is probably just double-struck.
2. Frere suggests that it was the badge of leglo VI, based at York, the capital of Britannia
Inferior but admits that there is no other evidence to support this (RIB 2411.34). His
reference to [0416], whether or not that does show a bull, is of no use here since several
other legionary sealings have reverses impressed by intaglios which were the personal
property of the soldier responsible and not intended to refer to the legion itself
3. Hassall's suggestion (quoted by Frere under RIB 2411.37) that the stag on [0309] was
a misinterpretation of the capricorn of Iegio II Augusa can now be disregarded in light
of the more recent discovery of [0308] which clearly shows a stag.
4. Unfortunately our sealing [0320] covers the period of changeover between the two
styles and so we cannot suggest which was the model for this letter.
5. There are at least two of these cuboid dies in existence in Britain. In 1992 Martin
Henig (pers. comm.) told me that a member of staff at the Ashmolean Museum knew of
another example but when I contacted her some time later she had little recollection of
this although she thought that it may have been brought in for identification. She was
unable to turn up any further information following a detailed search of several
departments in the museum and for any possible notes that she may have made at the
time. Nigel Mills (1994, 73) says that there is a third example in the British Museum but
Catherine Johns has informed me that this is not in the Department of Prehistoric and
Romano-British Antiquities. She feels that it may have been brought in for identification
and that the story has arisen from that. The member of staff at the Ashmolean Museum
says that it is conceivable that Mills' own example is the one which was submitted there
for identification.
213
Chapter 14
CONCLUSIONS
It is clear from the group of chapters concerning the evidence for use among
different categories that, despite their convenience for cataloguing, most of these
categories of scalings are not as clear cut as they may at first appear. For example, we
have seen that some Miscellaneous examples may in fact be Military, some Imperial
sealings may have been used on behalf of the army (as may some Official examples),
many Civic examples were connected with municipal customs dues and the majority of
Provincial scalings are probably Official.
Not only does the use of sealings have to be considered separately for each
category but even for examples within the same category, not just on account of the
contamination mentioned above but because of the varied uses that these sealings could
be put to.
Our findings can be summarised as follows:
Imperial sealings - All Imperial sealings would presumably have guaranteed the physical
integrity of the contents and would have acted as proof of exemption from custom
duties. Many of these scalings may have been attached to supplies for the army but need
not suggest the presence of the emperor.
Official scalings - Despite the lack of hard evidence, it is likely that Official scalings
usually indicated exemption from customs duty, in addition to any guarantee of physical
integrity of the contents. Only the suggested scalings of the Praefeclus Praelorio
Galliarum may have anything of interest to tell us.
214
Taxation sealings - The fact that we only have single-sided examples of scalings in this
category suggests that they were used to ensure that no untaxed goods were added later
to the packages for which dues had already been paid.
Provincial scalings - We cannot assume that these scalings would have provided
exemption from customs dues but we are probably correct in saying that they advertised
the authority by which they had been sealed (i.e. the governor of that province) and
would also have deterred any tampering
Civic scalings - The majority of these scalings were probably used to show payment of
customs dues. There is no direct evidence that any were used as labds to advertise the
place of origin or to guarantee quality, although it is not impossible.
Military scalings - The majority of Military scalings (i.e. those from Britain) were used to
protect their contents and, quite often, to name the person responsible. The Military
scalings from Lyon probably signified exemption from customs dues but this would not
have been necessary for the majority since they were only used within Britain.
Miscellaneous scalings - The main group within the Miscellaneous category, that of
scalings belonging to private individuals would almost certainly have provided security
but may also have acted as a label to expedite recovery of one's goods from a ship's
cargo. They are not, however, connected with the payment of customs dues since they
are purely for the benefit of the merchants named.
The use of the work in chapters 10, 11 and 12 to form any detailed conclusions is
rather dangerous on account of the limited quantities of scalings involved in the
calculations. It is hoped that the work is interesting pr se and that the methods can be
used in the future but, at present, there are insufficient numbers of scalings to make any
safe claims from the results.
215
From the above summaries, it can be seen that the basic rule of thumb for the use
of Roman lead sealings is that virtually all scalings probably provided security for the
contents of their packages (Taxation sealings and most Civic sealings doing so in order
that nothing was added rather than taken away) while any sealing from a government
agency (i.e. Imperial; Official; Military, but not necessarily Provincial) would have an
added purpose of denoting exemption from dues at any customs point passed en route.
The Taxation scalings and most Civic scalings would show that payment of dues had
already been made on the goods thus sealed, while only Those of private individuals
within the Miscellaneous category (and possibly the Provincial scalings) would possess
nothing more than a security/recognition significance.
216
ABBREVIATIONS

AAITAB Ada Associa!,o,,is internalionalis Terra Ai:Iiqua
Baka,,ica.

ACM ihe Acts of the Christian Mailyrs, ed H
MusuriHo, 1972,Oxford
Ada Arch. Ada Archaeologica A cademiae Scientiarum

Hiingaricae.
AE L 'Am,ée ipigrap/:ique.
Amm. Marc. Ammianus Marcellinus, Hisloria.

AMN Ac/a Mi, cei Napoceiisis.
Ann. d. his!. A,,,,ali d 'Ins/un/a di ('orri.spondenza

Archeologica.

A US Annuaire de l'U,iiversilé de Sofia (Toô,,uuuIuK iia
Coç!weiaia Yuiueejcmnem).
Ann. Nass. Not located - reference from ('IL XIII 10029.151
An(iq. Jour,:. The Antiquaries Journal.
A,,z. Wie,,. Anzeiger tier ()sierreichisc/,e,, Akademie tier

Wisseiischaflen (Philosophisch-His/orische
Kkisse).
Arch. Ael. Archaeo!ogia Aeliana.
Archaeol. PoLsk. Akad. Archaeologia - Rocznik In sty/ui/u His!orii Kul/my

Materialnef Polskiej Akademii Nauk.
Arch. Cant. Archaeologia Cantiana.
Arch. Jour,,. The Archaeologicalfournal.
ASFN Annuaire de Ia Sociélé defra,icaise de

IumisrnaIique
BA CT/-IS Bulletin archeologique dii ('omuté des Travauux

Iii sloriques et Scienlifiques.
217
B.A.R. British Archaeological Reports.
BCACR Bulleilino del/a ('o,nmiss,one archeologica

coiminale di Roma
13/A Bulletin de / instil:,! d 'Arthèologie.
BillS The British Journal for the His!oiy of Science
BMC British Museum Catalogue of Greek ('oinc,

1873-1927, London
Bonn. Jahrb. Bonner Jahrbücher
Britan,iia Brilaimia. A .Journal of Romano-British and

Kindred Studies.
Bull. 1)al,n. Builetino di archaeoiogia e sloria Dalmata.
BSNAF Bullet::: de la socièlé nationale des anliquaires de

France.
Bull. d. Istit. ?Builleltino deli 'Isliiuilo Siorico lialiano
BVJ3KABG Bulletin i'an de I 'ereeniging tot Bevordering der

Kennis i'an de Anlieke Beschaving le
's-Gravenhage.
('IL Corpus Inscriptionum Lailnarum, 1863-, Berlin
Cod Just. Codex Juslinianus.
Cod. Theod. ('odex Theodosianuuc
Compt. -Rend d'Hipp. Comptes-Rendus d'Hipp .. Not located - reference

from ('IL VIII 22656.17-18.
('W (umber/and and Wectnioriand Antiquiariaui and

Archaeological Soieiy, iransaci,o,,.s of
Daremberg-Saglio Daremberg, C, & Saglio, E., Diclionnaire des

Anliquiulés Grecquie et Romainec d 'après iec
lexies el les Monuments, Paris, 1877-.
218
Denkschr. Wien. Denkcc/irfien der Osterreichischen Akademie der
Wksenschaflen (P/lilosophisch-Historkche
Klase).
Derbyc. AJ. 1)erbyshire Archaeological Journal.
1)1g. Digesta In Iiniani Angus/i, ed T.Mommsen,

1870, Berlin.
Disserl. Arch. Not located - reference from ('IL XV 'Sigilla Varii

Generis', pa.ssiln.
Diz. Epigr. Dizionario Epigrafico di ,4ngichiià Roniane,

de Ruggiero, E., (ed), 1886-, Rome.
EE Ephemeris Epigraphica.
Epigr. A natal. Epigraphica A,iaIoIica.
Gal/ia Gal/ia.
Gaz. arch. Gazette archéologique. Recueil de monuments

pour servir a Ia connais.sauice ci a / 'histoire de
tart antique.
HA Ilis/oria Augusta.
1GB Juiscriptiones Graecae in Bulgaria Reperiae I, ed.

G. Mihailov, (2nd ed.), II-V, Sofia, 1958-70.
IGRRP Inscripliones Graecae ad Res Romanas

Periinenies, ed. R. Cagnat, J. Toutain, P. Jouguet,
0. Lafaye, Paris, 1906-27.

IJNA The Iniernational Journal of Nautical
Archaeology
1./NA UE The International .Journal of Nautical

Archaeology and (Jnden4'aier I.xploralioii.
ILN Illustrated London News.

ILS Inccripiiones La/mae Seleciae, ed. H.Dessau, 3
vols. in 5 parts, 1892-1916.
219

1/ui. Au!. Aritonine Itinerary (Ilinerarium Aniou,,ni A ugu c/i)
lii,,. Third. Bordeaux Itinerary (Iliuierarium Burdigalense .sive

Huerosolymilainim)
Jcthreshefle des L)5k'rreichischen Archdologischen

Instituis in Wien.
.J.JP Journal of izirislic Papjrology.
JOM fain huch des OherOslerreichi.schen

Museal'ereines

JRA Journal of Roman Archaeology.
IRS Journal of Roman Sludie c.
Kurir. Jahrb. Kvririerisches Jahrbuch.

Liddell & Scott A Greek-Engli.sh Lexicon, New Edition (rev.
Stuart Jones and Mackenzie), 1968, Oxford.
Lit!. Numismal. Vindol,on. Litlerae Numismaticae I 'mdoho,iensis.
Lucian, The Ship Lucian, The Ship.
MAIBL Mémoires de 1 'Académie des Inscriptions ci

Belles-Leilres.
MEFRA Mélanges de 1 'École Fran caise de Rome.
Mi/telL aits Ost-Ung. A rchaologisch-epigraphische Mitteihingen azis

()cierreich- ( Jngaun.
Mitiheil. des arch. Insliiutc, rOm. Mitiheilungen des deulschen archäologischen

Insliluits, A b/h. rOmische A htheiluing.
Mill. d. (1)si. Numisma. Ges. Milleilungen der Os'terreichische,,

Numismal:schen Gesellschafl.
MSNAF Jtlemoires de Ia Société Ncuio,iale des Anfiquaires

de Jrance.
No!. I)ig. No/ha l)ignita/um ommum lam civilium quam

,nuliiarium.
220
Not. d. Scan Notizie degli Scavi di aniichilà.
Num.Kôz. Numizn,aiikai KOzlOny.
OC]) Oxford ('lassical Dictionary, 1970, Oxford.
l'Amh Amherst Papyri, Grenfell, B P., & Hunt, A S.,

(edd.), 1900-01, London.
Philostratus, Vii. Apol. Philostratus, Vita Apollinii
Philostratus, Vii. Sop/i. Philostratus, I 'llae Sop/iistaruni.

PIR Prosopographia Imperil Romani.
Pliny, Ep. Pliny, Epistulae.

PLRE Prosopography of the Later Roman Empire.

Pontica Ponlica - Ada Musei Tomitani.
Ptolemy Ptolemy, Geographia.
RESEE Revue des Etudes Sud-Est Européennes.
Rev.Afr. Revue Africaine. Journal des travaux de Ia soclélé

hislorique A Igerienne par les membres de Ia
sociéié.
Rev.Num. Revue Numismatique.
RIB The Roman Inscriptions of Britain.
RAID Roman Military Diplomas, Roxan, M.M., 1978-,

London.

RA'IR Roman Military Records on Papyrus, Fink, R.O,
1971, Cleveland
RNMSAC Recuei! des Notices el Mémoires de Ia Sociélé

Archéologique dii dépariemeni de Constantine.
ROm. Osi. ROmishes ()sterrekh, Jahresschr?/i der

o.sierreichischen GesellschafiJuir Archaologie
221
SCN S/rid/i si ('erceIri de Nzimi.sinaiica.
Seneca, Brev. vii. Seneca, I)e Brevitate V//ac.
5K; LSylloge !nscriplioimnl Graecarum, ed. W.

Dittenberger, 3rd ed., 1915-24, Leipzig.
Stephanus B. Stephanus Byzantinus, El/rizika.
Strabo Strabo, Geographus.
Suetonius, Nero Suetonius, Nero.
Tab. Petit. Tabula Peulingeriana.
7ib. Vindol. 1 indolanda: The La/i,, Writing-tablets, Bowman,

A.K., & Thomas, J.D., 1983, London.
Themistius Themistius, Oralio,,es.
i: Lanes. & ('hesh. H. S.2 i'ra,isac/io,, of (he Lancashire and ("/,es/,ire

Historic Society, series 2.
T. Lond. Middi. A.S. Transaclions of/he Loiidon and Middlesex

Archaeological Society.
irier. Zeil. Trierer Zeilschrfl.
Vegetius Flavius Vegetius Renatus, Epiloma rei mi/hans.
Zeiischr. für Os!. Gym. ZeitschrJijuir aqerreichische Gymnasien.
222
BIBLIOGRAPhY
This bibliography also covers references in the Catalogue Periodicals have been
referred to here by their abbreviations as noted in the Abbre;ialions section.
Anon, 1879, No!. d. scavi
AlfOfdi, A, 193 1/2, 'Einige Denkmäler des pannonischen Urchristentums',

Num. KOz. xxx/xxxi.
Alfbldy, G, 1974, Nor,a,m, London.
Ailason-Jones, L. & Miket, R, 1984, The ('alalogue of smalifinds from South Shields
Roman 1'orl, Newcastle upon Tyne.
Babelon, E.C.F., & Blanchet, J.A., 1895, Catalogue des bronzes antiques de Ia
Bibliolhèque Nalionale, Paris(more info - add date to 0112)
Barkóczi, L., 1980, 'History of Pannonia' in Lengyel, A. & Radan, G.T.B., (edd.), op.
cit.
Barnea, I., 1969, 'Plombs Byzantins de Ia Collection Michel C. Soutzo', RESEE vii
Bamea, 1, 1982, 'Sigilii Bizantine de Ia Durostorum-Dorostolon', Pontica xv
Barnea, I., 1985, 'Sigilii Bizantine Inedite din Dobrogea (II)', Ponlica xviii
Becker, 1868, Ann. Nass. 9
Benndorf, 0., 1875, 'Beitrage zur Kenntnis des attischen Theaters', Zeifschr.fur act.
Gym. 26.
Berbrugger, 1864, Rev Afr. 8
Berbrugger, 1865, Rev. Afr. 9
van Berchem, D., 1937, 'L'annone militaire dans l'Empire romain au 111 siècle',
MSNAF xxiv
Besnier, M. & Blanchet, P. 1900, ('ollection Fcirges, (vol 9 of Musées el ('ollections

ArchéoIogiqiiei de l'Algérie e de Ia l'unicie), Paris
223
Binsfeld, W., 1988, 'Galatische Ware in Trier', Kurir. .Jahrb. 28; Same article also in
Fzinde zindAusgrahzingeii im Bezirk Trier - Aus der Arheit des Rheinischen
La,,desnn,seums Trier 20
Birk, D.A., 1975, 'Recent Underwater Recoveries at Fort Charlotte, Grand Portage
National Monument, Minnesota', LJNA 4.
Birley, A.R., 1988, The African Emperor: Septimius Severus, London.
Birley, E., 1958, 'The Roman fort at Brough-under-Stainmore', ('W2 lviii, 31-56.
Bishop, M.C. & Coulston, J.C.N., 1993, Roman Military Equipment, London.
Bivar, A DII., 1969, catalogue oft/ic Western Asiatic Seals in the British Museum:
Stamp Seals, II The Sassanian Djvasty, London.
Blagg, T.F.C., 1989, 'Richborough' in Maxfield, V.A. (ed.), op.cil.
Boon, G., 1991, 'Plumbum Britannicum and Other Remarks', Britannia xxii.
Bormann, 1887, No!. d. Scavi
Brandt, E., 1970, Antike Gemnien in Deutscheii Sammiungen I Siaatliche

Munzsammlung Miinchen. Teil 2 Italische Gemmen etruskisch his
roniisch-repiiblikanisch, Munich.
Breeze, D.J., & Dobson, B., 1987, Hadrian's Wall, London.
Bruzza, L., 1870, Ann. d. Inst.
Buli, Fr., 1890, Bull. DaIm. 13.
Bu1i, Fr., 1893, Bull. DaIm. 16.
Burnham, B.C. & Wacher, J., 1990, The 'Small Towns' of Roman Britain, London.
Bushe-Fox, J.P., 1949, Fourth Report on the excavations of the Roman for! at
Richborough, Oxford. (Society of Antiquaries Research Report XVI).
Butcher, K., 1988, Roman Provincial ('oi,is; An introduction to the 'Greek Imperials',
London.
Caruana, 1., forthcoming, The Southern I)efencec and ihird Century Barracks (?1 the
Roman Forts at C'arlisle: Excavations on the Tullie House Exten.sion Site, 1989.
224
Casey, P.J., 1984, Roman Coinage in Brila,,,, Aylesbury.
Cayfus, (Comte de), 1752, Recueil d 'antiqilifés /gjplle?nie', Eiriisques Grecqzies,

Romainec, Paris.
Chapman, H P. A., 1978, The Archaeological arid other evidence for the organisa/lon
and operatic!? of/he Cursti.s Publicus, Ph.D thesis, June 1978, University of
London.
Chevallier, R., 1989, Roman Roadc, London.
Chonchev, D., 1956/7, 'Die Schifbarkeit des Flusses Hebros im Altertum',

Wissenschafihiche Zeitschr/i der Ernst Morllz Arndi - Universiiät Grefswald
1/2, vi.
Clay, P., 1980, 'Seven Inscribed Leaden Sealings from Leicester', Britannia xi, 317-20.
Collingwood, R.G., 1927, 'Maiden Castle in Stainmore', CW xxvii, 170-77.
Collingwood, R.G., 1931, 'Objects from Brough-under-Stainmore in the Craven

Museum, Skipton', CW2 xxxi, 81-84.
Cornell, 1. & Matthews, J., 1982, A i/as of the Roman World, Oxford.
Cowen, J.D., 1937, 'A Celtic Sword-Pommel at Tullie House', CW2 xxxvii, 67-71.
Cruikshank Dodd, E., 1961, By:anthie Silver Stamps, Dumbarton Oaks.
Culic, V., 1971, 'Plumburi ale legiunii a X1-a Claudia gsite in sud-vestul Dobrogii',
SCN5
CuIic, V., 1975, 'Plumburi Comerciale din Cetatea Romano-Bizantin de Ia lzvoarele

(Dobrogea), I-Il!', Ponlica viii
Cu1ic, V., 1976, 'Plumburi Comerciale din Cetatea Romano-Bizantin de Ia Izvoarele,

IV-V', Po,,tica ix
Culicá, V., 1978, 'Cu privire Ia lagru1 Iegiunii a XI-a Claudia Ia Dunrea de Jos',
Ponlica xi
Culic, V., 1979, 'Plumburi Comerciale din Cetatea Romano-Bizantin de la Izvoarele

(Addenda et corrigenda)', Po,s(ica xii
Cuppers, H., 1974, 'Ausgewahlte rOmische Moseiflinde', Trier. Zeil. 37
225
Curie, J., 1911, A Ronian frontier post and its people: the fort at Newsiead, Glasgow.
Daniels, C.M., 1987, 'Africa' in Wacher, J. (ed.), op.cil.
Davies, R.W., 1971, 'The Roman Military Diet', Br, lannia ii (also published in Service
in the Roman Army, Breeze, D, & Maxfield, V. (edd), Edinburgh)
Davies, R.W., 1989, 'Joining the Roman Army', Service in the Romai Army,
Breeze, D., & Maxfield, V., (edd.), Edinburgh (also published in Bonn. Jahrb.
169, 208-32).
Dedeoglu, H., & Malay, H., Lpigr. A,,atol. 8.
Delattre, A.-L., 1892, ('ompl. -Rend. d'Hipp.
Delattre, A.-L. 1898, BACTHS
Delbrueck, R., 1932, Antike Porphyrwerke, Berlin & Leipzig.
Dembski, G., 1975, 'Römische Bleisiegel aus Osterreich (Eine Materialvorlage)',

ROm. Ost. 3.
Dembski, G., 1979, 'Eine UngewOhnliche Verwendung eines Bleisiegels aus

Carnuntum', Lilt. Numismal. Vindobon., Vienna.
Deringer, H., 1965, 'Em rOmisches Bleisiegel aus Enns', JOM 110.
Dessau, H. (ed), 1892-1916, liiscripliones Lalinae Seleclae, 3 vols. in 5 parts.
Detsicas, A. P. (ed.), 1981, ('ollectanea Historica - Essays in memory of Stuart Rigold,

Maidstone.
Devijver, H., 1989, 'Eine Lateinische holztafel in Leiden', The equesiriaFi officers of the
Roman imperial army (Ma yors: Roman army researches 6), Amsterdam
Diaconu, P., 1963, 'Urine vechi crestine descoperite In sud-vestul Dobrogei', Biserica
Oriodox2z Romángi 81.
Diaconu, P., 1971, 'n cutarea Dafnei', l'ontica iv
Dissard, P., 1905, La Collection Récamier. Catalogue des Plombs antiques ('weaur,
tessères, monnaies et objels diver. ․), Paris-Londres.
Domaszewski, A, 1890, 'Studien zur Geschichte der Donauprovinzen. 1. Die Grenzen

von Moesia superior und der illyrische Grenzzoll', Miuheil. ares O'J-Ung. 13
226
Dore, J.N., & Gillam, J.P., 1979, ilie Roman Foil at South Shields - Excavations
18 75-1975, Newcastle
Doublet, G., 1890, Mu,sée d'Alger, Paris.
Doublet, G., & Gauckler,P., 1892, Musée de Constantine, Paris.
Draganov, D., 1993, Monenioeeenenso na Ka6ui7e (The Coinage of Cabyle,), Sofia
Draganov, D. (ed), 1994, Jloce.iuu4elI vcucoin e àpeeiia Tpaiaia - ill Me.côyuiapoe,,

(uAmOjuV.4, "KalJu.ie" (Studies on Settlement life in Au,cient Thrace -
Proceedingc of the ilird international Sjmposium "Cabyle ",), Jambol.
Dressel, H., Notice stir le musée 1)othi'ell (not located - reference from ('IL XV 7940)
Egan, G., 1990, 'Leaden seals - evidence for EIC trade in textiles', JJNAUE 19.
Egan, G, 1994, Lead (lot/i Seals and Related Items in the British Museum, B. M.
Occasional Paper 93, London.
Egger, R., 1967, 'FUnfBleietiketten und eine Guf3form: die neuesten

Magdalensbergfiinde', Anz. Wien. 104, 195ff.
Endrei, W. & Egan, G., 1982, 'The Sealing of Cloth in Europe, with Special Reference
to the English Evidence', Textile History, 13, No.1.
Engelmann, H., & Knibbe, D., 1989, 'Das Zollgesetz der Provinz Asia. Eine neue
Inschrift aus Ephesos', Epigr. Anatol. 14.
Esmonde Cleary, A.S., 1989, The Ending of Roman Britain, London.
Ferrua, A., 1991, The Unknown Catacomb - A unique discovery of early Christian art,
New Lanark.
Ficoroni, F., 1740, IPiombi Antic/il, Rome.
Fourtier, 1853, A,,,,. de Consianli,,e
Fourtier, 1865, Rev. Afr. 9
Frere, S S., 1987, Britannia - A History of Roman Britain, London
Frere, S.S., & StJoseph, J K.S., 1983, Ronsan Britam from the Air, Cambridge.
Froehner, W., 1890, ASEN xiv
227
Frost, H., 1969, The Mortar Wreck iii Mellieha Bay, London.
Fulep, F., 1980, 'Pannonia Research in Hungary', in Lengyel, A, & Radan, G.T.B.
(edd), op.cit.
Furtwangler, A., 1900, Die An/iken Gemmen, Leipzig/Berlin.
Garnsey, P., & Sailer, R., 1987, The Roman Empire - Economy, Society and Culture,
London
Garrucci, R., 1866, '1 Piombi Scritti', Dis.ert. Arch.
Garrucci, R, 1847, 1 Piombi antic/il raccolti dali 'eminentissimo principe ii cardinale

Lodo'ico Altieri, Rome.
Garrucci, R., 1862, 'Notice sur quelques plombs antiques', Rev. Num. vii.
Gatti, 1891, Not. d. scavi.
Gerasimov, T., 1974, 'Miii mt exa 1L1,oMra C HMC1 o na ipas Ecec B Maima Amtq'
('A customs sealing with the name of the city of Ephesus in Asia Minor'),
Bulletin de ('Ins/limit d'Archeologie xxxiv
Gerasimova-Tomova, V., 1982, 'Tipt OBCIUI floMGn 14 MapKn Or Ka6nje' ('Trade

sealings and stamps from Cabyle'), I1zpeu IIaiuoiia.ieii Cu.umoauju -
IIoce.iuuen )Kueom B Tpaicwi (First National Symposium - Settlement life in
Thrace,), Jambol.
Gerasimova-Tomova, V., 1992, 'Bleiplomben aus Durostorum' in I)ie

Schwarzmeerkiiste in der Spaian!ike mind imfruhen Mile/al/er, (edd. Pillinger,
R., PUIz, A. & Vetters, H.),Vienna.
Gerasimova-Tomova, V., 1994, 'Tipronct<irre BpiwIt na Apioc (111! B. OT n.e.)' ('The

Trade Connections of Arzus (1st - IlIrd Century AD)'), in Draganov, D (ed),
op. cii.
GObI, R., 1969, 'Spatantike Bleibulle eines Maximinus aus Carnuntum' in Miii. d. Ost.
Numicmat. Ges. 16
GObi, R., 1978, A,,tike Niimismaiik vol 2, Munich
Goodbum, R., & Bartholomew, P. (edd), 1976, A Wects of the Nolitia Dignitaizim, BAR
S- 15, Oxford.
Goudineau, C, 1981, 'Informations Archeologiques - Côte D'Azur', Gallia 39.2
228
Gren, E., 1941, KIeinasien iiiid der Os/balkan in der H irl.schafilichen Enm'icklung der
rOmischen Kaiserzeit, Uppsala.
Grenier, A., 1934, 'Les plombs de commerce' in Maiiueld'Archéologie I'rehi.slorique

('ei/iqiie el Gab-Romaine vi.2 (ed Déchelette, J)
Guérin, M., 1900, 'Etude sur les Quattuor Publica Africae', RNMSAC xxxiv.
Hainzmann, M. & Visy, Z. (edd.), 1991, liisirumenta Inscripla La/i,ia - Das rOmische
Leheii im Siegel der K!einiiischrfien, Pécs.
Hammond, N. G L. (ed), 1981, Alias of the Greek and Roman World in Antiquity, Park
Ridge.
Harster, 1888, Calal. (presumably of Speyer Museum but not located - reference from
(YL XIII 10029.32)
Hassall, M.W.C., 1976, 'Britain in the Notitia', in Goodburn, R., & Bartholomew, P.
(edd.), op.dll., 103-17.
HassaH, M.W.C., 1983, l)erhys. A. .1. ciii
Haverfield, F., 1890, 'Roman Inscriptions in Britain 1888-1890', Arch. .Journ. xlvii
Henig, M., 1974, A Corpus of Ronian Engraved Gemstones from British Sites, B.A R.
(British Series) No.8, Oxford.
Henig, M., 1977, 'A bronze medallion and cube from Kingscote, Gloucestershire', Ant.
Journ. lvii.
Henig, M., 1980, 'An intaglio and sealing from Blackfriars, London', Anliq. Jour,,. lx, H
Henig, M., 1984, Religion in Roman Britain, London.
Henig, M., 1994, Classical Gems - Ancient and Modern Intaglios and Cameos in the
Fitzwilliam Museum, Cambridge, Cambridge.
Hoddinott, R.F., 1975, Bulgaria in Antiquity, London.
Holz, H., 1969, The Thomas 011ive Mabboit Collection Part!: Coins of the Greek
World, New York.
I-Iowgego, C. J., 1985, Greek Imperial ("ounlermarks, London.
229
Jarret, M G., 1994, 'Non-legionary Troops in Roman Britain Part One, The Units',
Britannia xxv, 35-77.
Johnson, S., 1979, The Roman Forts of the Saxon Shore, London.
Johnson, S., 1989, Hadrian's Wall, London
Jones, A H.M., 1964, The Later Roman Empire, 284-602, Oxford.
Jones, B., & Mattingly, D., 1990, An Atlas of Roman Britain, Oxford.
Kaibel, G, 1890, Jnccripliones Graecae VoLXIV - Inscripliones Jtaliae el Siciliae,

Berlin.
King, A., 1990, Roman Gaul and Germany, London.
King, C.W., 1872, Antique Gems and Rings, London.
Kiebs, E., & Dessau, H., 1897-8, Prosopographia Imperil Romani Saeculi I, II, III,
(1st ed.).
de Laet, S .J., 1949, Portorium - Etude cur 1 'Organisalion l)ouia,uière chez les Romains,
surlout a 1 'epoque dii Haul-Empire, Brugge.
Laurent, V., 1958-9, 'Un sceau inédit de l'Impératrice Aelia Eudoxia, femme de
Valentinien 111', Cahiers de Byrsa.
Lengyel, A., & Radan, G.T.B., 1980, The Archaeology ofl?oman Pannonia, Budapest
Lequément, R., 1975, 'Etiquettes de Plomb sur des Amphores d'Afiique', MEFR4 87.
Lewis, N., & Reinhold, M., 1966, Roman ('ivilization, Sourcebook H. The Empire, rev.
ed., New York.
Liou, B., 1980, 'Les amphores a huile de I'épave Saint-Gervais 3 a Fos-sur-Mer:

premieres observations sur les inscriptions peintes', Producción y Comercio del
Aceile en ía Aniiguedad, Madrid.
Maaskant-Kleibrink, M, 1971, 'Cachets de Terre de Douche (7)', BVBKABG xlvi.
Mc Dowell, R H., 1935, 'Stamped and Inscribed objects from Seleucia on the Tigris',
The University ofMichigan Studies xxxvi.
Maddoli, G, 1965, 'Le Cretule del Nomophylakion di Cirene', Anin,ario della regia
Scuola archeologica di Atene xli-xlii
230
Mann, J C., 1976, 'What was the Notitia Dignitatum for?', in Goodburn, R., &
Bartholomew, P. (edd ), op.cit.
Mann, 3 C., 1985, 'Two 'Topoi' in the Agricola', Britannia xvi.
Matthews, 3., 1989, 77ie Roman Empire of Ammianuc, London.
Maxfield, V.A., (ed), 1989, The Saxon Shore - A Handbook, Exeter.
Meiggs, R., 1960, Roman Os/ia, Oxford
de Mello, U.P., 1979, 'The shipwreck of the galleon Sacramento - 1668 off Brazil',
IJNA 8.
Metzger, C., 1981, Les ampotiles a eu/ogle dii musée dii Louvre, Paris.
Middleton, P., 1980, 'La Graufesenque A question of marketing', Athenaeum 58,

186-9 1.
Mills, N., 1995, Roman Artefactsfound in Britain, Chelmsford.
Mime, G., 1985, The I'orl of Roman London, London.
Mitchell, S., 1983, 'The Balkans, Anatolia, and Roman Armies across Asia Minor',
Armies and Frontiers in Roman and Byzantine Anatolia, ed. S. Mitchell, Oxford.
Mitchell, S., 1993, A nato/ia - Land, Men, and Gods in Asia Minor. Vol.!: The Cells in
A nato/ia aizd the Impact of Roman Rule, Oxford.
Mócsy, A., 1953, 'Das Terrilorizim Legionis und die ("aiiabae in Pannonien',
Ac/a Arch. 3
Morrison, C., 1987, 'Numismatique et Sigillographie: parentés et méthodes'

in Oikonomides, N., (ed), op.cIl.
Mowat, R K., 1895, BSNAF.
Mrozewicz, L., 1981, 'Les Plombs de Novae', Arc haeol. Polck. Akad. xxxii
Nicolaou, K., 1971, ILNcclix (May 1971).
Oikonomides, N, 1985, Byzantine Lead Seals, Dumbarton Oaks
Oikonomides, N., 1986, A Collection of Dated Byzantine Lead Seals, Dumbarton Oaks.
231
Oikonomides, N., (ed), 1987, S/nd/es in Byzantine Sigillography, Dumbarton Oaks.
Oikonomides, N., 198Th, 'The Lead Blanks used for Byzantine seals' in Oikonomides,
N., (ed), op.cit.
Oikonomides, N., 1987c, 'Le Boullôtèrion du César Constant (336/337), trouvé a

Beaumont-sur-Oise' in Oikonomides, N, (ed ) op.cit.
Orsted, P., 1985, Roman Imperial Economy and Romanizatio,,, Copenhagen.
Parker, A.J., 1987, 'Trade within the empire and beyond the frontiers', in Wacher, J.
(ed), op.cit.
Patsch, C., 1898, 'Piombo der legio XI Claudia P.1 aus Gardun', JO141 I.
Peacock, D.P.S., & Williams, D.F., 1986, Amphorae and the Roman Economy, London.
Pflaum, H.-G., 1940, 'Essai sur le Cursus Publicus sous le Haut-Empire romain',
MAIBL 14.1.
Polaschek, E., 1943, Review of Publicun; Porlorium Illyrici by A. Dobô, His! orische
ZeitschrVi 167, 144-49.
Poulter, A.G. (ed.), 1983, Ancient Bulgaria, Nottingham.
Poulter, A.G., 1987, 'Townships and Villages' in Wacher, J., op.ci!.
Reader, 1903, Arch.Journ. lx
Rheinisches Landesmuseum Trier, 1984, Trier, Kaiserresidenz und Bischofssitz, Trier
Richmond, l.A., 1934, 'The Roman fort at South Shields'(Appendix Lead Seals from
South Shields), Arch. Ael. xi
Richmond, l.A., 1936, 'Roman Leaden Scalings from Brough-under-Stainmore',

CW2 xxxvi (1936)
Richter, G.M.A., 1968, Engraved Gems of/he Greeks and the Etrnscc,ns, London.
Rickman, G E., 1971, Roman Granaries and S/ore Buildingi, Cambridge
de Ridder, A., 1915, Les Bronzes antiques du Louvre, vol. 11 - instrumentum, Paris.
Riley, D N, 1977, 'Roman Defended Sites at Kirmington, S.Humberside and Farnsfield,

Notts, recently found from the air', Britannia viii.
232
Rivet, A L.F., 1988, Gal/ia Narhonen.ci.c - Son/hem Gaul in Roman Times, London.
Roger, J., 1860, (]atalogue di, musée archéologique de Philippeville ('A/gene,) conlenant
en deux sections, le detail explicatfdes objels d 'anüqui/é, avec no/ice
hisiorique de chaqiie objet, Philippeville.
Römer-Martijnse, E., 1990, 'ROmerzeitliche Bleietiketten aus Kalsdorf, Steiermark',

l)enkschr. Wien. 205.
RostovtzefF, M. I., 1896, B till. d. Islil. - Mitiheil. dec arch. Insilhids, TOm. A bth.
Rostovtzeff, MI., 1896, Mitlheilung des arch. Instiluls, rOut. Abth. II.
Rostovtzeff ML, 1897, Rev. Num.
Rostovtzefl', M.I., 1899, Rev. Num.
Rostovtzeff M.1., 1900, Elude stir lesplombs antiques in Rostovtzeff & Prou ca(al. des
plomhs. Also published in Rev.Num. 1897-8 but with different pagination.
Rostovtzefl MI., 1904, 'Geschichte der Staatspacht in der rOmischen Kaiserzeit big
Diokietian', Phi lologus, SuppL, ix, 329-5 12.
Rostovtzeff, MI., 1932, 'Seleucid Babylonia: Bullae and Seals of Clay with Greek
inscriptions', Yale Classical Studies III.
Rostovtzeff, M.I., 1957, Social and Economic History of/he Roman Empire, 2nd ed.,
Oxford.
Rostovtzeff, MI. & Prou, M., 1900, ('alalogue desplomhsde l'Anliquulé, dii Moyen
Age el des temp.s modernes conserves au dépanlement des Médailles el Antiques
de Ia Bibliothèque Nationale, Paris.
RCHM, Roman London (not located)
Rubensohn, 0., 1907, Elephantine Papyri, Berlin.
Salinas, A., 1883, No!. d. Scavi.
Sandrovskaya, V.S., 1975, Bioauinuucucne lIe'ianiui B CotJpanuu 9p.4m,na.*a

(Byzantine Sealings in the Hermitage collection), SiPetersburg
Schreiber, 1842, Verzeichniss (ms) - not located.
Schuize, W., 1904, Zur Gecchichie lafrmnischer Eigennanien
233
Schwinden, L., 1985, 'ROmerzeitliche Bleietiketten aus Trier. Zum Handel mit Pfeffer,
Arznei und Kork,' Trier.Zeit. 48, 121ff.
Sear, DR., 1982, (;reek Imperial Coins and their values, London
Sena Chiesa, G., 1966, Gemme delMuseo Na.zionale di Aquileia, Aquileia.
Sharpe, R R. (ed), 1907, Calendar of Letter-Books Preserved among the Archives of

the (]orporat ion of the City of Londo,, at the Guildhall, Letterbook H, London.
Smetius, J., Pater et Filius, 1678, Anliquitates Neomagenses, Nijmegen.
Smith, C. Roach, 1848-80, Colfectanea Antiqiia, 7 vols., London.
Smith, H.Ecroyd, 1866, T. Lancs & ('hesh. H.S. 2 vi, 137-52.
Spier, J., 1992, Ancient Gems and Finger Rings, Catalogue of the Collections - The J.
Paul Getty Museum, Malibu.
Stead, 1.M., 1976, Excavations at Winier/on Roman Villa and Other Roman si/es, North
Lincolnshire, 1958-67, H.M SO.
Steyert, A., 1895, Nouvelle histoire de Lyoii I, Lyon.
Stieglitz, C.L., 1820, Archaologische Unierhaftungen, part II, Leipzig.
Still, M C.W., 1993, 'Opening up Imperial lead scalings', .JRA 6
Still, M.C.W., 1994, 'Some Roman lead scalings from Arzus', in Draganov, D. (ed.),
op. cit.
Still, M.C.W., 1995, 'Parallels for the Roman lead sealing from Smyrna found at Ickham,
Kent', Arch. Cant. cxiv (1994).
Strong, D., 1988, Roman Art, Harmondsworth.
Sutherland, C H, 1974, Romai Coins, London / Monnaies Romaines, Paris
Tancheva-Vasileva, N., 1990, 'Kabyle Military Camp', AAJTAB v.
Thibault, F., 1888, Les Donanes chez les Romains, Paris.
Thompson, D J, 1987, 'Imperial Estates', in Wacher, J (ed), op. cit.
Tomassetti, G., 1887, BCACR.
234
Tomassetti, G, (ed), 1892, BCA('R.
Tomlin, R.S.O, 1988, 'The Curse Tablets' in CunlifTe, B (ed), The Temple of Siths
Minen'a at Bath, Vol.2 The Finds from the Sacred Spring, Oxford.
Tóth, E., 1991, 'Bleibullen im Ungarischen NationalMuseum' in Hainzmann,

M. & Visy, Z. (edd ), op. cii.
Toynbee, J.M C., 1973, A,,imals in Roman Life a,,dArt, London
Turcan, R., 1987, Nigra Monela. Sceaux, felons, Iessères, amuleties, plombs nionélaires
on monélformes, objeis divers enpiomb ozi en elain d'epoque romaine
conserves an M,isée des Beaux-Arts de Lyon (Palais Saint-Pierre), Lyon.
Tylecote, R.F., 1964, 'Roman lead working in Britain', BJHS vdl II, part 1, no.5.
Velkov, V., 1983, 'The Thracian City of Cabyle' in Poulter, A.G (ed.), op.cii.
Vermeule, C., 1978, 'The imperial shield as a mirror of Roman art on medallions and
coins', in Scripia Nummaria Romana - Essays preseiwed to Humphrey
Sutherland, (edd. Carson, R.A.G., and Kraay, CM.), London.
Verneaux, J., (ed), 1966, Pseudo-Kodinos. Traité des Officec, Paris.
Villefosse, A.H., 1882, BSNAF
Visconti, 1879, BCACR.
Visconti, 1847, Memorie Niimismaiiche per I 'anno 1847 pubblicale da 1). Diamilla I
Vitucci, G., undated, 'Libertus', Diz. Iipigr. IV., Rome.
Vlassa, N., 1981, '0 Categorie Nesemnalata de pietre gravate magice', AlvIN xviii
Vons, P., 1980, 'Roman Lead Sealings from Velsen (N.H.)', Helmium xx
Wacher, J., 1974, The Towns of Roman Britain, London.
Wacher, J. (ed), 1987, The Roman World, London.
Warmington, B.H, 1954, The North Africa,: Provinces from Ekioclelia,, to (he Vandal
Conquest, Cambridge.
Wiegand, T. & Schrader, H., 1904, Priene
235
Wightman, E.M., 1985, Ga//ia Be/gka, London
Wilkes, J J., 1969, I)almalla, London
Wilkes, J J., 1981, 'Map no.24: The Balkan and Danubian Provinces' in Hammond,
N G.L. (ed), op.cil.
de Witte, J.J.A.M., 1884, 'L.Munatius Plancus et le Genie de Ia yule de Lyon?', Gaz.

arch.
Young, Ci., 1981, 'The Late Roman Watermill at Ickham, Kent, and the Saxon Shore',
in Detsicas, A. (ed), op.ciI.
Zacos, G , & Veglery, A, 1972, Byzanaiiie Lead Sea/s Vol. 1, Part 1, Basel.
Zhouglev, K., 1966, 'La situation économique en Thrace et Mésie et leurs relations
commerciales avec l'Italie aux Id_IVs den é.', A(JS.
236
(i. )

Vol 1

Uploaded by

Copyright:

Available Formats

Vol 1

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Vol 1

Uploaded by

Copyright:

Available Formats

Roman Lead Sealings

MICHAEL CHARLES WILLIAM STILL

SUBMITTED FOR TIlE DEGREE OF PILD.

UNIVERSITY COLLEGE LONDON

This thesis is based on a catalogue of c. 1800 records, covering over 2000

List of Maps, Figures, Tables and Illustrations 10

2. Typology of Roman Lead Scalings 35

Type 1: Two-sided with different sized obverse and 36

Main Indices 609

Quick Reference Guide to Classification and Provenance of Sealings 715

Nomina of individuals and Families (except members of Imperial Family) 727

Cognomina of Individuals (except members of the imperial Family) 730

Members of Imperial Families (where mentioned by name) 734

Members of Imperial Families identified from portraits alone 734

List of previously unpublished examples 738

List of previously published examples for which I have offered 739

Supplementary Material (bound in at back of Volume II)

Still, M.C.W., 1993, 'Opening up Imperial lead sealings', Journal of Roman

p.88, line 6: for 'manslo' read 'mutalio'.

p.120, note 16: for 'Janet' read 'Jarrett'.

p.203, bottom line: for '[1764]' read '[1765]'.

p.230, top line: for 'Jarret' read 'Jarrett'.

Map of Britain showing findspots (inset 3 of map of empire) 22

Fig.2 Actual numbers of Imperial scalings found in each province 143

Fig.4 Actual numbers of Imperial scalings identified by inscription 146

Fig.5 Imperial scalings per thousand identified by inscription 149

Fig.9 Percentages of sealings in different categories for whole 161

Fig. 10 Percentages of sealings in different categories for whole 163

Fig. 11 Percentages of scalings in different categories in Africa 166

Fig. 12 Percentages of scalings in different categories in Asia (unshaded) 167

Fig. 13 Percentages of scalings in different categories in Britannia (unshaded) 169

Fig. 14 Percentages of scalings in different categories in Dalmatia (unshaded) 171

Fig. 15 Percentages of scalings in different categories in Gallia Belgica 172

Fig. 17 Percentages of scalings in different categories in Germania Superior 176

Fig. 18 Percentages of sealings in different categories in Italia (unshaded) 178

Fig. 19 Percentages of scalings in different categories in Moesia Inferior 180

Fig.20 Percentages of scalings in different categories in Noricum (unshaded) 182

Fig.2 I Percentages of scalings in different categories in Numidia (unshaded) 183

Fig.22 Percentages of sealings in different categories in Pannonia Superior 185

Fig.24 Percentages of sealings in different categories in Syria (unshaded) 189

Fig.25 Percentages of sealings in different categories in Thracia (unshaded) 190

Fig 26 Percentages of scalings in different categories from unknown 192

Table V. [Percentages of scalings from] Whole empire excluding sealings 162

Imperial - Gallia Lugdunensis (0042-0069); Italia (0111) 770

Imperial - Pannonia (0226-0227); Thracia (0230-0241) 774

Imperial - Thracia (0268-0270); Unknown Provinces (0271-0273) 777

Provincial - GaHia Belgica (0314), Thracia (0320-0321) 780

Civic - Thracia (0386-0387); Unknown Province (0388) 782

Miscellaneous - Asia (0668-0675); Britannia (0705-0709) 784

Miscellaneous - Britannia (0739-0779); Germania Inferior (151 1-1512); 786

Miscellaneous - Unknown Provinces (1794-1809); App 1. 13 - Italia 793

Replies concerning an absence of sealings in their museum collections were

Lothar Schwinden of the Rheinisches Landesmuseum Trier informed me that

No. on map Site Pros ince Modern Country

Aldborough Britaiinia Great Britain

24. Corbridge Britannia Great Britain

31. Gigen Moesia Inferior Bulgana

65. Ravenglass Britannia Great Britain