i « * * Z+Z *1*2 **2*2

HHHMi

Jfijj jiajjto M3 ifllHiiSI

! '5r<s*!’!i}ii)}iM jf }?f«? «*<}«$<$« If of {Jfcifli

CENTENNIAL PUBLICATION

CENTENNIAL

PUBLICATION

THE AMERICAN NUMISMATIC SOCIETY Edited by HARALD INGHOLT

1958

THE AMERICAN NUMISMATIC SOCIETY • NEW YORK

ALL RIGHTS RESERVED BY THE AMERICAN NUMISMATIC SOCIETY

ERRATA

Page 66, 11. 5 and 7 from top: For “Brettiens” read “Bruttiens”.

Page 66, 1. 11 from top: For “Planche VII, 7” read “Planche \ II, 4”. Page 67, 1. 3 from bottom: For “Brettiens” read “Bruttiens”.

Page 80, 1. 12 from top: For “Plate II, 10” read “Plate IX, 10”.

PRINTED IN GERMANY AT J.J. AUGUSTIN • GLOCKSTADT

FOREWORD

Born a century ago with no resources except hope and enthusiasm, our Society has now reached a great milestone in its career. It has today all the hope and enthusiasm with which it began life. It also has something more: its physical plant, its collections, its place among the world’s learned societies,- all the result of the generosity, the vision and the wise service given by a long line of devoted members.

Many suggestions for the proper celebration of this anniversary were offered and considered. In 1908 when the Society first occupied its new home on Audubon Terrace, it inaugurated the Archer M. Huntington Medal “in Commemoration of the Fiftieth Anniversary of The American Numismatic Society.’’ Copies of that medal have been awarded from time to time since then for distinguished service to numismatics. Another means of celebration seemed desirable today, the publication of two special volumes, one a history of the Society, the other this collection of short papers covering the entire range of numismatic interest.

This particular volume brings together articles contributed by friends through¬ out the world. Their papers will, we believe, be a permanent memorial to the anniversary which they and we now commemorate.

Louis C. West President

The American Numismatic Society

EDITOR’S NOTE

Xhe Council of the Society at its meeting of April 15, 1955, entrusted the preparation of the publication marking the Society’s Centennial to a committee composed of Caspar J. Kraemer, Jr., Professor, New York University, George C. Miles, Chief Curator, Sawyer McA. Mosser, Executive Director, and Louis C. West, President of the Society, and myself.

With the work now coming to a close I want to express my thanks to Miss Rose E. Mangini and to Miss Vivian D. Kline of the Society’s clerical staff for the care with which they have read the proofs. For valuable advice freely given I am deeply grateful to my colleagues on the committee, especially to Mr. Mosser who not only gave me the benefit of his wide experience in numismatic publi¬ cations, but also temporarily took over for me last summer when I was excavat¬ ing in the Near East. Finally, may 1 humbly express the hope that the Tyche of Numismatics may have so guided the reading of the page proofs that mistakes will be found to be neither too numerous, nor too serious.

Harald Ingholt Centennial Publication

TABLE OF CONTENTS

Foreword. By louis c. west . v

Editor’s Note. By harald ingholt . vii

Abbreviations . xi

Silver Currency and Values in the Early Byzantine Empire. By Howard

L. ADELSON . I

The Portrait of Caesar on the Denarii of 44 B.C . and the Sequence of the Issues.

By ANDREW ALFOLDI .

A Knotty Problem Connected with the Coinage of Azes. By a. s. altekar . 45

Le Roi Pyrrhos. By jean babelon . 53

A Hoard of Silver Coins of the Empire of Nicaea. By Alfred r. Bellinger . . 73

Introduccion del “ Alancuso” en la Economia Carolingia. By pio beltran ... 83

Der Munzenfund von Werl ( Westfalen ) 7955, vergraben urn 1240. By peter

berghaus . 89

XVII-XIX Century Manuscript Material on Anglo-Saxon Coins. By

c. e. blunt . 125

Brasher & Bailey: Pioneer New York Coiners, 1787-1792. By Walter h.

breen . 137

Questioni ponderali. By laura breglia . 147

Chemical Composition of Alexandrian Tetradrachms. By earle r. caley . . 167

The Coinage and Chronology of A.D. 238. By r.a.g. carson . 181

La Nike funebre delle monete di Elis. By ettore gabrici . 201

Notes on the Danish Mint-Masters in the Middle Ages before 1377. By georg

GALSTER . 209

Von der N umismatik zur Geldgeschichte . By hans gebhart . 233

Les plus ancicns tarifs monetaires illustres des Pays- B as. By H. enno van

gelder . 239

The Alexandrine Tetradrachms of Cabyle in Thrace. By th. gerassimox . . . 273

Trois monnaies parthes inedites. By R. ghirshman . 279

ylsscs of Orichalcum. By michael grant . 285

Some Modern Forgeries of Carolingian Coins. By philip grierson . 3°3

Eine Medaille des Andrea Doria von Christoph Weiditz. By p. grotemeyer 317

IX

X

TABLE OF CONTENTS

Richard Wagner in Medallic Art. By henry grunthal . 329

"Cooperation ’ : Some Thoughts and Suggestions for the Intensification of Numis¬ matic Research. By Walter havernick . 349

Rechnung und Zahlung. By eduard holzmair . 353

A Note on Corinthian Coins in the West. By G. k. jenkins . 367

Der Ostfriesische Miinzmeister Dietrich I den. By karl kennepohl . 381

A propos des representations de boucliers sur les monnaies grecques. By lkon

lacroix . 401

Tresor de monnaies carolingiennes decouvert dans le Jura. By jean lafaurie 407 L’Embleme du lis dans la numismatique byzantine: son origine. By v.

LAURENT . 417

Ein Vierteljahrhundert Wiener Munzkabinett. By august loehr . 429

A Coinage of the Revolt of Fregellae ? By harold mattingly . 451

Le Grand Sceau de France ( depuis ij8g) . By jean mazard . 459

The Early Islamic Bronze Coinage of Egypt. By george c. miles . 471

La technique des revers partiellement incus de monnaies pheniciennes . By

PAUL NASTER . 503

A Hoard of the First Silver Coins of Nuevo Reino de Granada [Colombia) . By

R. I. NESMITH . 513

A Recently Discovered Coin Solves a Vermont Numismatic Enigma. By

ERIC P. NEWMAN . 531

A Lycian Hoard. By Sydney p. noe . 543

Die Medaillonprdgung unter Cams und seinen Sohnen. By karl pink . 553

Store-Cards or Tokens of Mexico. By a. f. pradeau . 563

Sur des types de monnaies imperiales d’Asie Mineure. By louis Robert . . 577

Some Electrum and Gold Greek Coins. By e. s. g. robinson . 585

The Ring of Polycrates. By Charles seltman . 595

Parion au 3e siecle avant notre ere. By henri seyrig . 603

Diocletian s Reformed Coinage in Britain and Related Problems. By c. h. v.

SUTHERLAND . 627

The Grain-Ear Drachms of Athens. By Margaret Thompson . 651

A Numismatic Study of a King without a Coinage. By bengt thordeman 673 L’ Atelier monetaire d’ Anvers des temps merovingiens au X 'IP siecle. By

VICTOR TOURNEUR . 683

A Unique Medal of the Seljuk Tughrilbeg. By John walker . 691

Coins and Indian History. By R. b. whitehead . 697

ABBREVIATIONS

AJA

American Journal of Archaeology

AJN

American Journal of Numismatics

AJP

American Journal of Philology

ANS

American Numismatic Society

ANSMN

American Numismatic Society Museum Notes

Babelon, Traite

Babelon, Traite des monnaies grecques et romaines

BCH

Bulletin de correspondence hellenique

BMC

British Museum Catalogue. Individual volumes quoted deal

with Greek, Byzantine, Vandal, Indian, or Anglo-Saxon coins.

BMCRE

Coins of the Roman Empire in the British Museum

BMCRR

Coins of the Roman Republic in the British Museum

BNJ

British Numismatic Journal

BSFN

Bulletin de la Societe Frangaise de Numismatique

BZ

B yzantinische Zeitschrift

CAM

Cambridge Ancient History

CIL

Corpus Inscriptionum Latinarum

CRAI

Comptes rendus de V Academie des Inscriptions et Belles-

Lettres

ESM

Newell, The Coinage of the Eastern Seleucid Mints

HN 2

Head, Historia Numorum, 2. edition

IG

Inscriptiones Graecae

IBS

Dessau, Inscriptiones Latina e Selectae

JDAI

J ahrbuch des deutschen arch dot ogischen Institute

XI

XI 1

ABBREVIATIONS

JHS

Journal of Hellenic Studies

JIAN

Journal international d’archeologie numismatique

JNG

Jahrbuch jiir Numismatik und Geldgeschichte

JRS

Journal of Roman Studies

MBNG

Mitteilungen der Bayerischen N umismatischen Gesellschaft

MGH

Monumenta Germaniae Historica

NC

Numismatic Chronicle

NNM

Numismatic Notes and Monographs

NZ

N umismatische Zeitschri ft

RBN

Revue beige de numismatique

RE

Pauly- Wissowa, Real-Encyclopadie

REA

Revue des etudes anciennes

REG

Revue des etudes grecques

Rev. Arch.

Revue archeologique

Rev. Phil.

Revue de philologie

RIC

Mattingly and Sydenham, Roman Imperial Coinage

RIN

Rivista italiana di numismatica

RN

Revue numismatique jrangaise

RSN

Revue suisse de numismatique

SHA

Scriptores Historiae August ae

SIG

Dittenberger, Sylloge Inscriptionum Graecarum

SNG

Sylloge Nummorum Graecorum

WSM

Newell, The Coinage of the Western Seleucid Mints

ZDMG

Zeitschrift der deutschen morgenldndischen Gesellschaft

ZfN

Zeitschrift fur Numismatik

SILVER CURRENCY AND VALUES IN THE EARLY BYZANTINE EMPIRE

Howard L. Adelson

Robably about the year 292 or 293, Diocletian and his colleagues began striking good silver coins of i/96th of a Roman pound (3.41 grammes). The terminus ante quem is determined by the obverse legends on some of the coins which appear to be struck in this standard, and the weight is clearly indicated by the mark XCVI which is found on other coins of this standard.1 The average weight of 560 coins weighing between 1.9 and 4.4 grammes is 3.09 grammes with the best preserved specimens varying between 3.2 grammes, the mode of the frequency curve, and

1 Theodor E. Mommsen, Histoire de la monnaie romaine, trans. Due de Blacas (Paris, 1873-4), III, p. 73, n. 2, hereafter cited as Mommsen-Blacas. Some silver pieces of Carausius of the period 287-293 were struck on the same standard. Karl Pink, “Die Silbcrpragung der Diocletianischen Tetrarchie,” NZ, Neue Folge, XXIII (1930), p. 38, argues in favor of 293 A.D. Otto Seeck, “Die Miinzpolitik Diocletians und seiner Nachfolger,” Z/N, XVII (1890), p. 61, suggests that it is pos¬ sible that the new silver coin was introduced in 286 A.D. at the same time as the aureus of i/6oth of a pound. On the mark of value and the literature explaining it see E. Babelon, Traite des mon- naies grecques et romaines (Paris, 1901), I, pt. I, col. 565, n. 2, hereafter cited as Babelon, Traite, and Mommsen-Blacas, III, p. 74, n. 2. Cf. C. H. V. Sutherland, “Flexibility in the 'Reformed Coinage of Diocletian,” in R. A. G. Carson and C. H. V. Sutherland, Essays in Roman Coinage Presented to Harold Mattingly (Oxford, 1956), pp. 175-6. C. H. V. Sutherland, “Diocletian's Reform of the Coinage: A Chronological Note,” JRS, XLV (1955), pp. 116-8, dates the silver reform, at least at Treveri, to 294 A.D.

i Centennial

I

o

HOWARD L. ADELSON

3.4 grammes.2 Thus the difference between the average weight and the theoretical weight is only about eleven percent.

These new silver coins differ in alloy from those of the third century. The Romans were fully capable of obtaining virtually pure silver from the ores at their disposal, and this they did for many purposes.3 In the Life of St. Theodore of Sykeon there is mention of a chalice and paten of silver bearing a five-fold stamp of purity. These objects were purchased in Constantinople and brought to the Saint. In answer to the prayers of the Saint, who thought them to be of unholy origin, they turned black before they were used in the church. Further investigation revealed that they came from an improper source and were therefore not fit for use in the Divine service.4 This five-fold stamp of purity was probably of the same variety as that found on several ingots bearing notations such as PRI(S)CI TR PS. PI. which Evans has expanded as Prisci(anus) Tr(everorum) p(u)s(ulati) p(ondo) u{num)P It was silver of this sort which was used in the new currency, and one estimate of the fineness of the coined metal places it at over 98% pure.6

The usurping British Emperor Carausius had, in fact, begun to strike fine silver currency at a standard of about i/5oth of a pound, but even his coinage contained some plated pieces while the main body of third century silver was quite base. Diocletian’s new silver issues were of fine silver, and his successors continued to

2 Gunnar Mickwitz, “Die Systeme des romischen Silbergeldes im IV. Jhdt. n. Chr. : Ein Beispiel zum Anwendung der variationsstatistischen Methode in der Numismatik,” Societas Scientiarum Fennica ( Finska Vetenskaps Societeten), Comment ati ones Humanarum Litter arum, VI, 2, p. 42, hereafter cited as Mickwitz, Die Systeme. Cf. ibid., pp. 51ft'.

3 On the refining methods see J. and L. Sabatier, Production de Vor, de V argent cl du cuivre chez les anciens et hotels monetaires des empires romain et byzantin (St. Petersbonrg, Paris, and London, 1850), pp. 47-9. Cf. J. Hammer, “Der Feingehalt der griechischen und romischen Mtinzen ,” ZfN , Neue Folge, XXVI (1926), p. n, and R. J. Forbes, Metallurgy in Antiquity. A Notebook for Archae¬ ologists and Technologists (Leiden, 1950), pp. 169-231, as well as R. J. Forbes, “Metallurgy”, in5 S. Singer et ah, A History of Technology (Oxford, 1956), II, p. 42, notes 1, 2 and 3, and pp. 43-45. In my own article (“The Bronze Alloys of the Later Roman Empire,” ANSMN, VI (1954), pp. 111-129, and esP- P- I21) I have shown that after the mid-fourth century great care was taken to remove all of the silver from the ores used for the bronze coinage.

4 Life of St. Theod. of Sykeon, in E. Dawes and N. H. Baynes, Three Byzantine Saints. Contemporary Biographies Translated from the Greek (Oxford, 1948), pp. 117-8.

5 A. Evans, “Notes on Coinage and Silver Currency in Roman Britain from Valentinian I to Con¬ stantine III,” ArC4, XV (1915), pp. 494ff.

6 A. von Rauch, “fiber die romischen Silbermunzen und den innern Werth derselben. Ein Bei- trag zu den altern metrologischen Untersuchungen,” Mitteilungen der Numismatischen Gesell- schaft in Berlin, III (1857), p. 307. Hammer, “Der Feingehalt der griechischen und romischen Mtinzen, ” ZfN , NeueFolge, XXVI (1926), pp. 109k, lists a series of analyses of coins of this period.

CURRENCY IN EARLY BYZANTINE EMPIRE

3

strike good silver.7 The XCVI coin was in fact a restoration of the Neronian denarius, and though the average weight of the fourth century coins is slightly below that of the denarius both were meant to be i/96th of a pound of pure silver.8 This is important because the Germanic tribes outside of the Empire had continued to use the Neronian coin as currency during the third century. The new silver piece was the first struck since Septimius Severus which was also used by the free barbarians.9

In the period from the abdication of Diocletian (305 A.D.) until about 330 A.I). there is surprisingly little evidence of silver minting.10 In this same period, however, a new coin which was one-half the restored XCVI piece in weight was introduced, but the new coin of i/i92nd of a pound of silver is relatively rare. During the reign of Constantine further additions were also made to the monetary system in the form of heavy silver coins of i/72nd and i/6oth of a pound.11 The date of the introduction of these coins is uncertain, but 324 A.D. has been suggested as the most likely date because large coins are customarily viewed as commemorative issues or for use in the distribution of largesses, and the final defeat of Licinius in that year provides a satisfactory motive for minting them.12

These are the basic silver coins of the early fourth century, and the changes that they underwent may be briefly recorded. Both heavy pieces were consistently struck by the later emperors of the fourth century save for Julian, who minted only the lighter of the two.13 The smallest coin, that of i/i92nd of a pound,

7 Mommsen-Blacas, III, pp. 88-9, 91-2, and esp. p. 91, n. 4, where the point is explicitly made that plated coins were not accepted as silver. Thus in the aes hoard of Wroxeter a plated coin was found which was obviously circulating as a bronze piece. Cf. J. Maurice, Numismatique constan- tinienne. Iconographie et chronologic , description historiqne des emissions monetaires (Paris, 190S-12), I. PP- 397-8-

8 Mickwitz, Die Systeme, p. 43, presents a frequency curve of 278 Neronian denarii with a mean weight of 3.27 grammes.

9 Mommsen-Blacas, III, pp. 121, 132. In the first century the Germans had a manifest preference for silver as opposed to gold. Tacitus, Germania, V.

10 H. Mattingly, Roman Coins from the Earliest Times to the Fall of the Western Empire (London, 1928), p. 222, n. 1; Maurice, Numismatique const antinienne, I, p. xxxviii.

11 See the tables based on F. Gnecchi, I Mcdaglioni Romani, I (Milan, 1912), given by G. Dattari, “Del Miliarense e della Siliqua neU'Epoca Costantiniana,’’ RIN, XXXI (1918), pp. 209ff.

12 Maurice, op. cit., I, p. xxxvii. Cf. Mickwitz, Die Systeme, pp. 6, 31, and 61-3, who suggests that a series of i/66th of a pound also existed. Coins at i/72nd of a pound can be distinguished from those of i/6oth of a pound, but pieces of i/66th of a pound would be virtually indistinguishable from those at i/72nd and at i/6oth of a pound.

13 Mommsen-Blacas, III, pp. 72-3; Mickwitz, Die Systeme, pp. 6 and 32. Cf. H. A. Grueber, 1 he Southsea Find of Fourth Century Coins,” A’C6, XVI (1936), pp. 293-303.

4

HOWARD L. ADELSON

was struck as late as the reign of Constantius II, but it was not issued after¬ wards.14

The most important change in the fourth century silver coinage is the replace¬ ment of the piece of i/96th of a pound by a coin of approximately i/i44th of a pound. Coins of the period 337-361 A.D., when plotted on a frequency curve, show a lack of definite peaks, which fact is directly attributable to a change of standards.15 Maurice, who had been influenced by the work of Mommsen, believed that at the latest by the year 324 a new coin was introduced.16 Mattingly has stated that at some time about the year 348 the new coin of i/i44th of a pound was introduced.17 Actually Mommsen noted that the coin of i/96th of a pound disap¬ peared only during the second half of the fourth century. Most of the coins of Constantius II and those of Julian as Caesar, and even some of his as Augustus, are of this standard, but about the year 360 A.D. the coin of i/i44th of a pound replaced the older one.18

The average weight of the new coins, as is to be expected, was low. The 261 coins minted in Treves from the Growely Wood hoard and the British Museum Collection showed an average weight of only 1.93 grammes, but it must be re¬ membered that the average weight of coins is generally below the theoretical weight and the coefficient of variation in this instance is 12.85. H is therefore quite safe to postulate a theoretical weight of i/iqqth of a pound, ca. 2.27 grammes, for these pieces.19 Thus the new coin would be in an exact weight relationship of 2:1 with the commonest heavy piece, that of i/72nd of a pound. The new coin, as compared with the earlier one of i/qbth of a pound, is quite common, and its variation from the theoretical weight is not much greater (fifteen percent).

14 Mommsen-Blacas, III, p. 73; Mickwitz, Die Systeme, pp. 7 and 60.

15 L. C. West and A. C. Johnson, Currency in Roman and Byzantine Egypt (Princeton, 1944), p. 180, infra.

16 Maurice, Num. const., I, p. cxxxi; Mommsen-Blacas, III, pp. 76-7.

17 Mattingly, Roman Coins, p. 223, places the change early in the reign of the sons of Constantine the Great. Mattingly, “The Monetary Systems of the Roman Empire from Diocletian to The¬ odosius I,” NC 6, VI (1946), p. 116, gives the year of this change as 348 A.D., when the 1,100th year since the foundation of Rome occurred.

18 Mommsen-Blacas, III, pp. 73-7. Ibid., Ill, p. 79, n. 2, mentions the fact that it cannot be mere chance that the Holwel silver hoard as well as others are generally confined either to coins struck before 360 or to coins issued after 360. Mickwitz, Die Systeme, pp. 60-1, has since confirmed this. Cf. Seeck, “Die Munzpolitik Diocletians und seiner Nachfolger,’’ Z/N, XVII (1890), p. 66.

19 Mickwitz, Die Systeme, pp. 61-3. Ibid. , p. 13, argues against this. Cf. Mommsen-Blacas, III,

P- 77-

CURRENCY IN EARLY BYZANTINE EMPIRE

5

There may perhaps be a reminder of the introduction of this new reduced denomination in the story of the mutinous uproar in the army after the siege of Pirisobora (363 A.D.). Julian promised to each soldier argenteos nummos centenos. If paid in the new coin this would amount to quite a bit less than if paid in the old i/96th of a pound coinage. The soldiers, in any event, thought that it was a small sum, and they voiced their opposition. There is no way of proving which coins are meant, but since the date of the event is after 360, it would seem as though the new denomination was signified.20

For some time after the reign of Julian this coin of i/iqqth of a pound remained the principal silver denomination. The slow decline in the effective weight of the coin during the period to the reign of Honorius was not sufficient in magnitude in the larger hoards, such as that of North Mendip, to be of any real significance. It is only some time about the years 392-397 A.D. that a sharp drop in the weight of the coinage occurs. The new pieces issued after that date weighed approximately 1.3 grammes. These new coins are in direct continuation in the hoards of the 1.9 gramme pieces. Indeed, the evidence of the clipping of the older coinage to the new size is quite manifest, and it seems probable that the new coin was not an addition to the monetary system but a replacement of the older piece. In the period from Honorius to Anastasius, which encompasses the entire fifth century, this reduced silver denomination was virtually the only silver coin in existence. There are to be sure, some coins that appear to weigh over two grammes, but their number is so limited that the fifth century is as bleak a period in silver currency as it is in dated papyri.21

The heavy silver coins of i/72nd and i/6oth of a pound and even larger con¬ tinued through the fourth century, but in 384 A.D. a law of Theodosius and Valentinian prohibited those who gave the public games from distributing a coin heavier than i/6oth of a pound.22 After that date the large medallions which had

20Amm. Marc., XXIV, 3, 3. Cf. Zosimus, III, 18. Babelon, Traite, I, pt. I, col. 566, felt that the older coins of i/96th of a pound were meant. If these older coins are siliquae then the largess was 4 1/6 solidi per soldier. Mommsen-Blacas, III, p. 72, infra, felt that only siliquae could have been meant. Cf. W. Kubitschek, “Der Ubcrgang von der vordiokletianischen Wahrung ins IV. Jahrh. Randbcmerkungen zu Schriften von G. Mickwitz,” BZ, XXXV (1935), p. 342, who suggests that the miliarense, a heavier coin, is meant. The miliarense was certainly the coin commonly used for largesses, but the value of 100 miliarenses must have been about seven or eight solidi and could not have aroused complaints regarding the paucity of the gift.

21 Mickwitz, Die Systeme, pp. 15-9 and 61-3, is the source for the information in this paragraph.

22 C. Th., XV, 9, 1.

6

HOWARD L. ADELSON

survived the reform of about 360 A.D. became extremely rare. The fifth century presented a black picture with the absence of heavy silver pieces.

The economic revival which followed the end of the first onslaught of the bar¬ barians began during the reign of Anastasius (491-518), who reformed the bronze coinage. During his reign silver coinage was still extremely rare, and it was only during the reign of Justin I that the quantity of silver issued from the imperial mints began to rise appreciably. The various denominations, however, cannot be determined because of an insufficiency of weights until the reign of Justinian I, who, in addition to heavy coins of i/72nd of a pound, struck some light coins (below two grammes) marked CN, PKG, and PK. This series, which was struck only in Italy, was continued by Justin II, but it seems to have come to an end before the accession of Tiberius Constantine. The weights of these coins are too few in number to establish very meaningful curves. It is not at all improbable, however, on the basis of the recorded weights for these coins that those marped CN (250 units) were struck on a standard of i/288th of a pound, and that those marked PKG (125 units) and PK (120 units) were meant to be roughly one-half as heavy. There is also a series of small silver coins of approximately the same standard as the CN and PK€ pieces, but without marks of value, that was issued in Carthage and Ravenna during the period from Justinian’s reign to that of Heraclius. This short-lived rivival under Justinian and Justin II, which was followed by a period of small irregular silver coinage issued in very limited quantities, was only relieved by the Heraclian reforms.23

23 The known weights and sources of the CN, PK€, and PK pieces are collected by Mickwitz, Die Systeme, p. 19, n. 3. Babelon, Traite, I, pt. I, col. 579, gives some silver weights for the reigns of Anastasius and Justin I, but actually very few coins of these rulers are known, certainly not enough to identify the standard. It is most likely that the revival of the economic life of the Empire began in the reign of Anastasius, and that the reigns of Justin I and Justinian are to be accounted as a unit during which the revival reached its peak. Cf. W. Wroth, BMC, Vandals, pp. xlix, and 120-1, who classifies the small coins bearing marks of value and the name of Justin as being derived from the reign of Justin II. W. Wroth, BMC, Byz., I, p. lxxvi, gives a list of the weights of silver coins without marks of value from the reign of Anastasius through the remainder of Byzantine silver coinage. Babelon, Traite, I, pt. I, cols. 5 7°-i, lists some coins of i/72nd of a pound beginning with the reign of Zeno. Mickwitz, Die Systeme, p. 31, lists the weights of thirteen coins of this series from the reign of Zeno through that of Justinian.

CURRENCY IN EARLY BYZANTINE EMPIRE CHART I*

PERIOD AT WHICH THE COINS WERE ISSUED

Theoretical

292-

305-

337-

361-

363-

393-

450-

518-

565-

Standard

Wt. in grs.

305

337

361

363

393

450

5i8

365

610

fPK

576

0.568

\PKG

X

288

I*I4

CN

X

192

1.71

X

X

X

X

144

2.27

X

X

X

96

3-41

X

X

X

72

4-55

X

X

X

X

X

X

60

5-46

X

X

X

* This chart is a somewhat revised and expanded version of one found in West and Johnson, Currency in Roman and Byzantine Egypt, p. 106. Such charts are the simplest way of keeping track of what coins were issued during the different periods.

In determining the gold-silver ratio during the early Byzantine period we are fortunate in having the review article by West in which most of the earlier views and almost all of the sources are discussed and summarized. There is no point in reproducing this work, but it would be well to summarize very briefly West’s results and to add a short discussion of one or two of the sources omitted by him.24

West has concluded from his study of the sources, as they were known in 1941, that only three texts furnish a true indication of the gold-silver ratio. To these three given by West three more that seem to be of almost equal value can now be added. The complete list of six notices, counting C. 77?., XIII, 2, 1, as distinct from C. Just., X, 78, 1, because there are intervening notices of different ratios, is as follows:

CHART II

Source

Date

Ratio

SB 6086

early IV cent.

1:18

P. Oslo 162*

IV cent.

i:i4-4

C. Th., XIII, 2, 1

397 A.D.

1:14.4

C.I.L., V, 8734*

ca. 400 A.D.

i:i5

C. Th., VIII, 4, 27

422 A.D.

1 :i8

C. Just., X, 78, 1*

534 A.D.

i:i4-4

* The asterisk indicates that this source was omitted by West.

24 L. C. West, “The Roman Gold Standard and the Ancient Sources,’’ A JP, LXII (1941), pp. 289-301.

cS

HOWARD L. ADELSON

West has quite properly omitted the passage in Ammianus Marcellinus recording that Julian gave a donative of five solidi and one pound of silver to each soldier. Since this passage is often quoted, however, it would perhaps be best to undertake to demonstrate the correctness of omitting it from any discussion of the gold- silver ratio.

Mommsen quoted this passage in his work and then stated that the two sums were equivalent and showed a gold-silver ratio of 1:14. 4. 25 The passage itself merely says quinos omnibus aureos, argentique singula pondo promisit ,26 There is no indication that the two sums are equivalent. Later scholars, however, such as Sabatier and Kubitschek, to cite only one earlier and one later numismatist, have accepted Mommsen’s point of view.27 Segre has said of this passage that it “is doubtless a very unusual donativum,” and one which he reckoned at about twenty solidi.28 There is absolutely no reason for this assumption. Ammianus Marcellinus himself indicates that the donativa on the election of an emperor were fixed.29 The sources show that there was a custom of distributing five solidi and a pound of silver to each of the soldiers on the election of a new emperor. Leo I, in 457 A.D., Leo II, in 473 A.D., Anastasius I, in 491 A.D., and Justin I, in 518 A.D., all gave the same donative.30 In the case of Leo I our information is derived from a late ceremonial guide, and Bury comments, “This description is taken from an evidently contemporary document preserved in Constantine Porph. De ceremoniis, I, c. 91 [ed. Bonn, p. 412]. There can, I think, be little doubt that Constantine found it in the ceremonial book (Kcn-aa-rao-is) compiled by Peter the Patrician in the sixth century, from which we know that he derived other accounts of early ceremonies (see ib. cc. 84, 85). ”31 Ceremonies of the type

25 Mommsen-Blacas, III, p. 155.

26 Amm. Marc., XX, 4, 18.

27 J. Sabatier, Description generate des monnaies byzantines frappees sous les empereurs d’Orient depuis Arcadius jusqu’a la prise de Constantinople par Mahomet II, I (Paris, 1862), p. 50; Kubitschek, BZ, XXXV (1935), P- 348.

28 A. Segre, “Essays on Byzantine Economic History. I: The Annona Civica and the Annona Militaris,” Byzantion, XVI (1942-3), p. 413.

29 Amm. Marc., XXVI, 2, 10, on the election of Valentinian I, says, “ob nimcupationcm augustam debita protinus accepturi.”

M J. B. Bury, History of the Later Roman Empire from the Death of Theodosius to the Death of fustinian {A.D. 395 to A.D. 565) (London, 1931), I, pp. 316, 431-2; II, p. 17.

31 Ibid., I, p. 316, n. 2. Cf. the note in the Bonn edition of Constantine Porphyrogenitus, De Caenmoniis, II, p. 415. The passages are to be found in De Caerimoniis (ed. Bonn), p. 412 (Leo I) ; p. 432 (Leo II); pp. 424-5 (Anastasius); p. 429 (Justin I).

CURRENCY IN EARLY BYZANTINE EMPIRE g

that are written down in books of that order are not likely to respond to changes in the relative values of silver and gold. At one time the ratio of silver and gold may have been expressed in this gift, but there is no way of proving that such was the case during the reign of Julian, and the passage should certainly not be used as evidence in that connection. The custom of granting five gold pieces (solidi) to every soldier on every fourth year was enshrined in a law which was abrogated by Justinian.32 Possibly the addition of a pound of silver to this donative merely indicated the commencement of a new reign. Certainly the accession of Tiberius II Constantine in 578 was marked by a donative of nine solidi and no silver, which would seem to indicate that the pound of silver had been replaced by four solidi which may have been its equivalent in gold.33 If that were so the gold-silver ratio in 578 A.D. would have been 1:18.

The six acceptable quotations of the gold-silver ratio deserve a moment’s con¬ sideration. The dating of SB 6086 has presented a problem for many scholars because the papyrus itself is undated. The dating of Mickwitz, which places the document as post 324 A.D. but definitely in the early part of the fourth century, seems worthy of acceptance.34 The ratio of 1:18 which is given in SB 6086 is given in another instance which will be discussed later.

P. Oslo 162 cannot be dated closer than the fourth century. It records the purchase of one pound of silver bullion (novas) for five solidi. Both SB 6086 and P. Oslo 162 deal in terms of metal weights and what appears to be silver bullion. The ratio of P. Oslo 162 is confirmed by that of C. Th., XIII, 2, 1, of 397 A.D. That is to say that a ratio of 1:14.4 was in use in the latter half of the fourth century. Mickwitz has suggested that this might be only a temporary ratio caused by an unusual export of gold which was forbidden by a law of 374( ?).35 West, in

32 Procopius, Anecdota, XXIV, 27-9.

33 Ed. E. W. Brooks, lohannis Ephesini Historiae Ecclesiastici Pars Tertia, lib. Ill, c. xi, which is vol. 55 of the Scriptores Syri = Corpus Scriptorum Christianorum Orient alium (Louvain, 1952), vol. 106, p. 101. Cf. R. Payne Smith, The Third Part of the Ecclesiastical History of John Bishop of Ephesus Now Translated from the Original Syriac (Oxford, i860), p. 186, for a somewhat different translation than that of Brooks. This donative, which was given on accession to the throne, was apparently called the Augustaticum .

34 Mickwitz, Die Systeme, pp. 7 ff . He gives a full exposition of his reasoning there. Also see West and Johnson, Currency in Roman and Byzantine Egypt, pp. 185-6. J. G. Milne, “On P. Oslo 83 and the Depreciation of Currency,’’ Journal of Egyptian Archaeology, XXIII, pt. 2, pp. 258-9, and L. C. West, Gold and Silver Standards in the Roman Empire, NNM, No. 94 (New York, 1941), p. 187, n. 394, quite properly reject the use of P. Oslo 83 which is often cited to determine the gold-silver ratio.

35 Mickwitz, Die Systeme, p. 19, n. 1; C. Just., IV, 63, 2.

10

HOWARD L. ADELSON

his article, refutes this by referring to the some twenty year time lag between the law of 374 and that of 397. It can be shown that the Romans had a bullionist approach to economics, and this law of 374 A.D. may have been nothing more than a measure in that direction. It need not refer to a specific export of gold and a consequent shortage.

The bullionist approach to economics on the part of the Romans is patent. It was with a view to freeing the Empire from economic dependence upon Persia that Justinian sought direct routes, or routes through the lands of other peoples than the Sassanians, to the East.36 The organization of the silk monopoly and the price regulations for the import of that commodity illustrate the fact that the motive of Justinian in attempting to secure routes to the East, which lay to the north and to the south of Persia, was not only to ensure the continued access to eastern luxuries during the periods of strife with Persia, but also to prevent the Persians from making a profit on Roman trade. The history of Roman sumptuary legislation from the days of the Republic to those of Justinian indicates this bullionist view of governmental wealth and power. Many examples may be cited to prove this, but it is sufficient to note that Tiberius in the course of his letter to the Senate advocating stringent sumptuary legislation speaks of the purchase of luxuries which divert the wealth of Rome to strange and hostile nations.37 More important, however, is the fact that the text of C. Th., XIII, 2, 1, is repeated in exactly the same words in the Justinian Code.38 This would seem to indicate that the same ratio adhered in 534 that had existed in 397 A.D.

Silver values, however, must have been falling towards the end of the fourth century, if Mommsen’s expansion of the abbreviations in CIL, V, 8734 is accepted. The inscription contains the line SOL XX VIA/ L, which is expanded as sol(idos) XX v(el) i(n) a(rgento) u(ncias) L. Mickwitz accepts this expansion, but Kubitschek rejects it.39 Even if this expansion is rejected it can still be shown quite certainly that silver values fell, for in the year 422 there was issued the decree preserved in the Theodosian Code which grants the right of paying the sportulae to the duces of the pvimipilares in the form of either one pound of silver

36 A. A. Vasiliev, History of the Byzantine Empire (Madison, 1928), pp. 199 and 203 ff.

37 Tacitus, Annals, III, 53. Justinian refused permission to the grandmother of John the son of

Basilius to ransom him from Chosroes for 2,000 pounds of silver because he did not wish Roman

wealth to be transferred to the Persians. Procopius, Anecdota, XII, 8. Also see the speech of

Julian after the mutiny at Pirisobora quoted in Amm. Marc., XXIV, 3, 4-5.

38 C. Just., X, 78, 1.

39 Mickwitz, Die Syslenie, p. 18, n. 3; Kubitschek, BZ, XXXV (1935), p. 348.

CURRENCY IN EARLY BYZANTINE EMPIRE

1 1

or four solidi.40 The ratio of 1:14.4, which adhered in 397, was therefore replaced by that of 1 : 18. Mickwitz saw in this fact a change in the actual market price of silver, and he therefore attempted to refute the view that this edict referred ex¬ clusively to the case of the sportulae by citing the text of SB 6086 which shows the same ratio.41

In this discussion of the gold-silver ratio, in addition to those sources not ac¬ cepted by West, the very many ratios set forth by secondary authors based upon the actual weights of the coins have been omitted because, quite obviously, those ratios would be based upon the theoretical weight of the non-fiduciary silver coins, and there would also have to be absolute certainty regarding the identifi¬ cation of the denominations of the coins. Some scanty references such as those for the expedition against the Vandals in 467 are also disregarded. John Lydus gives the cost of that expedition at 65,000 pounds of gold and 700,000 pounds of silver.42 Procopius gives the cost of this same expedition at 130,000 pounds of gold.43 This yields a ratio of 1 :io.7. The sources of these two statements are dis¬ cussed by Bury, but their value from the standpoint of economic history of numismatics is questionable.44 The figure obtained from this data cannot be con¬ firmed from any other source, and one of the reports may well have been corrupted. The ratio, if correct, would be unusual. Even the statement of John Malalas that a general sent to Antioch just before the fall of that city to the Persians purchased silver from the inhabitants for two or three solidi per pound is not to be used in connection with establishing the mint ratio because it was a price peculiar to siege conditions.45 Malalas, as a sixth century author, is probably truthful when speaking of nearly contemporary events such as the siege of Antioch, but this price would reflect the desire of the fleeing Antiochenes to replace their silver with less bulky gold. The general, on the other hand, could make use of the mili¬ tary transport, and, therefore, the price of silver would be depressed.

That the basic silver coin, the siliqua, cannot have been a mere unit of account or a fiduciary coin has been proved by Mickwitz.46 Therefore it follows that the actual

40 C. Th., VIII, 4, 27.

41 Mickwitz, Die Systeme, pp. 17-8.

42 John Lydus, De Magistratibus, III, 43.

43 Procopius, Wars, III, 5.

44 Bury, History of the Later Roman Empire, I, p. 337, n. 3.

45 John Malalas, Chronographia, lib. XVIII, O 222 (ed. Bonn, p. 480).

48 Mickwitz, Die Systeme, pp. nff. ; Mickwitz, “Geld und Wirtschaft im rdmischen Reich des vicrten Jahrhunderts n. Chr.,” Societas Scientiarum Fennica ( Finska V ctenskaps Societeten), ( otnmen-

12

HOWARD L. ADELSON

weight of the siliqua should vary inversely with the rise and fall of the value of silver, since it is clear that it is always i/2qth of the value of a solidus. Obviously, on the basis of the evidence regarding the gold-silver ratio, the rulers from Con¬ stantine to Heraclius did not succeed in maintaining the mint ratio established by Constantine when he created the solidus and valued a certain silver coin at i/2qth of it.

What has been said indicates that actually the silver currency formed part of a seemingly bimetallic monetary arrangement in which the government was most interested in the gold portion of the system. The legal relationship of the silver coins to the gold, i.e., the number of siliquae equal to a solidus, remained unchang¬ ed. Gresham’s law should then have operated to cause silver to appear and to disappear from the market with regularity. The government might reduce the weight of the silver coin or raise it in order to counterbalance the effects of this law, but in the long run the arrangement could not operate successfully. The saving feature was that silver was unimportant in the Roman monetary system, and that the bronze was fiduciary. The Roman monetary system may actually be considered as monometallic in its important aspects, but a discussion of the silver series requires a clarification of the bimetallic aspect. The text of a law of 429 A.D. stipulating that the treasurer or tax collector had to accept either gold or silver without injustice in the matter of weights indicates that questionable practices involving Gresham’s law were known to these officials.47 Nevertheless consider¬ ation of this should be deferred until the identification of the various denomi¬ nations is completed.

All earlier theories regarding the so-called money of account siliqua have been adequately treated by Mickwitz. Any account of the fourth century denominations must begin by pointing out that since Diocletian struck only one silver coin that coin was known as the argenteus minutulus, argenteolus or more simply argenteus ,48

tationes Humanarum Litterarum, IV, 2, pp. 2iff., hereafter cited as Mickwitz, Geld und Wirtschaft. Also see the arguments of L. Blancard, “Une millares de Constantin a Heraclius,” RN 3, VI (1888), p. 428, for distinguishing between coined metal and bullion. This silver coinage, however, was of fine metal and would have been considered as bullion in transactions carried out by weight. P. Le Gentilhomme, “Le Monnayage et la circulation monetaire dans les royaumes barbares en Occident (Ve-VIIIe siecle),” RN5, VII (1943), pp. 5 iff . , doesnot accept the non-fiduciary character of this silver currency and argues strongly against accepting the views set forth by Mickwitz.

47 C. Just., X, 72, 15 = C. Th., XII, 6, 32.

48 Mommsen-Blacas, III, p. 81. Cf. P. Antinoopolis 38, of 301 A.D. where a pound of worked silver is priced at sixty-two denarii and a pound of cast silver at thirty-one denarii. Clearly the denarius was a unit of account.

CURRENCY IN EARLY BYZANTINE EMPIRE

13

Our knowledge of this term in connection with fourth century coinage is almost exclusively drawn from the Scriptores Historiae Augustae. The histories of the various rulers included in that work were composed in the fourth century, but the events referred to are of the third century. The numismatic material contained in these histories has, however, been adequately treated by Menadier with regard to its date.49 Only three quotations from this series of histories refer to the argen- tens minutulus. Vopiscus speaks of a coin called an argenteus philippeus minutulus .50 Aelius Lampridius mentions the reduction of the price of beef or pork from eight minutuli to two or even one minutulus in the reign of Alexander Severus.51

This argenteus was in common usage until the time of Julian the Apostate, but at some time prior to 323 Constantine changed its name to siliqua. It was then valued at i/2qth of a solidus. The terminus ante quern for the introduction of this term is furnished by an inscription from Feltre which uses the abbreviation SIL in speaking of coinage.52 The identification of the argenteus with the siliqua was first established by Mickwitz, who pointed out that the siliqua is known to have been i/24th of a solidus, and the ratio of gold to silver ca. 324 A.D. to have been 1:18. The siliqua auri or i/i728th of a pound of gold would thus be equal to 0.1895 grammes. If minted in silver at the above ratio the coin would weigh 3.4 grammes. This is exactly the theoretical weight of the argenteus of the years before 360. 53 Therefore in the period prior to 360 A.D. the argenteus-siliqua is the coin of i/96th of a pound of silver. The light coin of i/i92nd of a pound, which was struck in the early fourth century, was obviously the half-argenteus or siliqua.54

At some time after 355 A.D. the ratio of gold to silver changed. Silver rose in value. It must have risen much higher than a ratio of 1:14.4 would suggest, and it would seem from the weights of the new coins that the ratio should have been approximately 1:12. This would be about the ratio of the early empire in the period from Nero to Septimius Severus.

It is immediately evident that on this matter there is a divergence from the

49 Menadier, “Die Miinzcn und das Miinzwesen bei den Scriptores Historiae Augustae,” ZfN, XXXI (1914), pp. iff.

50 Vopiscus, Vita Aureliani, IX, 7; XII, 1.

61 Aelius Lampridius, Vita Sev. Alex., XXII, 9. Cf. A. Evans, “Notes on Coinage and Silver Cur¬ rency in Roman Britain from Valentinian I to Constantine III,” NC4, XV (1915), pp. 476 7.

52 ILS, 9420.

83 Mickwitz, Die Systeme, pp. 2-10.

54 West and Johnson, Currency in Roman and Byzantine Egypt, p. 108; Mommsen-Blacas, III, p. 73.

i4

HOWARD L. ADELSON

system set forth by Mickwitz. Mickwitz has proposed to see a drop in silver values to the ratio of 1:20 at this point and then to identify the 1.9 gramme coin as the half-siliqua.55 The error in this identification is twofold. Firstly, Mickwitz used the average weights of the coins introduced after 360 instead of the theoretical weight, and, as a result, he made an error in calculating the gold-silver ratio. On the basis of the theoretical weight of the coin (2.27 grammes), if it were a half- siliqua, the ratio would be 1:23 or 1:24 and not 1:20. The second error made by Mickwitz was to base his identification of the coin on the fact that one gold-silver ratio rather than another seemed closer to the recorded fourth century range of mint ratios which is 1 : 14. 4-1 : 18. This last error was aggravated by his mistaken use of the average weight of the coin. In addition it should be noted that our earliest notice of the half-siliqua comes from the fifth century (Nov. Valent., XN , 1, 1 (444/5)) when the 1.14 gramme coin (i/288th of a pound) was the common one.

If the mint ratio was, as we suggest, 1:12 about the year 361 A.D., then the siliqua would have a theoretical weight of 2.27 grammes which fits the new coinage of Julian most admirably. This new coin which replaced the piece of 1 /96th of a pound was struck a i/iqqth of a pound.56 Possibly Julian returned to the early empire mint ratio as part of his program of reviving the ideals and practices of that period, or perhaps he had an economic or financial motive in mind.

The next notice of a changed gold-silver ratio is of the year 397 A.D. In that year the mint ratio was specified as 1:14.4. The theoretical weight of the siliqua was then 2.73 grammes. The drop in silver values made the old siliqua too light to serve its purpose. The new coin of ca. 1.3 grammes, however, which replaced the siliqua, will serve admirably as the half-siliqua. This portion of the exposition of fourth century silver coinage, it will be seen, does not follow Mickwitz in viewing this coin of 1.3 grammes as the reduced weight of the same denomination as the 1.9 gramme piece.

The drop in silver values continued, and at one point in the early fifth century the market ratio may have been 1:15. By the year 422 the mint ratio was cer¬ tainly 1 : 18. This meant that the half-siliqua was now too light, and that it would not find general acceptance. The result was that though the half-siliqua was

55 Mickwitz, Die Systeme, pp. 14-5, 31-2.

56 Mattingly, “The Monetary Systems of the Roman Empire from Diocletian to Theodosius I, ArC6, VI (1946), p. 1 16, would also view the new coin as a siliqua. This change of standards put the siliqua in the ratio of 2: 1 with the miliarense of i/yznd of a pound which was the most common of the large silver pieces.

CURRENCY IN EARLY BYZANTINE EMPIRE

15

mentioned in two legal texts of the mid-fifth century, the government had ceased to issue silver coinage in any quantity.57 During the period when significant quantities of it were issued, however, the half-siliqua was also known as the nummus decargyrus. This identification rests upon the fact that in the only text mentioning this term the exchange of bronze coins for others is treated. These other coins must have been silver.58

Smaller coins of a quarter of a siliqua were also minted. No silver coins of less than the half-siliqua except the quarter-siliqua are known from the literary sources, and it is improper to presume that they existed.59 These quarter-siliqua coins are so small, and the weights vary so widely in view of their size, that it is unsafe to attempt to identify them with a specific standard of weight. They were probably used as largesses for the humbler classes of the population. There are aurei which show the emperor scattering handfuls of coin from a chariot, and these minute pieces may well have served that purpose.60

That the value of silver was falling in the period from 397-450 AT), and even afterwards can be shown easily by the evidence of the clipping of the older siliquae to approximately half-siliqua size. The British hoards such as Coleraine, Colerne, South Ferriby, Icklingham II and III, and Tuddenham show the most marked signs of open clipping.61 This open clipping cannot have been for fraudulent pur¬ poses because it is too easily recognizable, and there is no evidence of this practice in hoards dated prior to the reign of Honorius.62 The silver coinage of the fifth century, however, was not really in accord with the actual gold-silver ratio in the market-place, and the apparent experiment in bimetallism, which had been tried in the fourth century, came to a disastrous end.

57 The legal texts are Nov. Valent., XV, 1, 1, of 444/5 A.D., and Nov. Maioriani, VII, 1, 16, of 458 A.D.

58 C. Th., IX, 23, 2. Mommsen-Blacas, III, pp. 83-4; 84, note 1 ; and 164. Cf. J. de Petigny, “Ttudes sur l’histoire monetaire du quatrieme au septieme siecle,” RN, nouvelle scrie, II (1857), P- I27> n. 1, who identifies the maiorina pecunia with the nummus decargyrus. The law dates from 395 A.D.

r>9 F. Hultsch, Metrologicorum Scriptorum Reliquiae (Lipsiae, 1864-6), I, pp. 222, 231, 245, and 248, mentions the orrapiov which is defined ad loc. as the quarter-siliqua. This work will hereafter be cited as Hultsch, Scriptores. Cf. Babelon, Traite, I, pt. I, col. 578.

J. W. E. Pearce, “A Hali-Siliqua of the Treveran Mint,” NC6, III (1943), pp. 97-9. Cf. Evans, NC *, XV (1915), pp. 476-7.

61 H. Mattingly and J. W. E. Pearce, ‘‘The Tuddenham Hoard of Siliquae,” jVC6, VI (1946), pp. 172-173.

62 See J. W. E. Pearce, “A Second Find of Siliquae from Shapwick,” NC*, XVIII (1938), pp. 5 7-8-

HOWARD L. ADELSON

16

When silver coinage in the smaller denominations was resumed under Justinian an attempt may have been made to maintain a mint ratio of gold and silver. The notice which is extant of a ratio of 1:14.4 for the reign of Justinian is merely a copy of the earlier decree preserved in C. Th., XIII, 2, 1. The wording of this law remained unchanged. After 200 years, in tax receipts, one part of gold was still equivalent to 14.4 parts of silver. If this ratio is meaningful then the CN coins of i/288th of a pound can only be half-siliquae and those marked PKC, quarter- siliquae.63 If this could be proven then we should have established that the solidus of the sixth century contained 12,000 nummi. The PK€ coin would be i/q6th of a solidus or the quarter-siliqua and equal 125 nummi. This agrees remarkably well with the mint ratio and the marks on the coins, but it must be remembered that there is a statement in Procopius regarding the number of units in the solidus. This passage, which has often been discussed at length, states that the solidus of Justinian was worth 210 coins called folles, and that its value was reduced to 180 folles. The follis, virtually all authorities are agreed, was a large bronze coin. The highest denomination of bronze minted by Justinian was marked M which means forty nummi. This yields a solidus of 8,400 nummi. The CN piece would then equal 1/33. 6th of a solidus, but this is not a convenient simple relation¬ ship between gold and silver coins.64 The markings on the coins cannot be made

63 A. Segre, Metrologia e circolazione monetaria degli antichi (Bologna, 1928), p. 481; A. Lnschin von Ebengreuth, “Der Denar der Lex Salica,” Sitzungsherichte der Kaiserlichen Akademie dcr Wissenschaften in Wien, Phil. -hist. Klasse, CLXIII (1910), Abh. 4, p. 18, n. 1; Mommsen- Blacas, III, p. 90; A. Engel and R. Serrure, Traite de numismatique du moyen age, I (Paris, 1891), p. 4; and Mickwitz, Die Systeme, p. 20, view the CN coins as siliquae. Seeck, ZfN, XVII (1890), p. 67, identifies the CN coins as half-siliquae. Babelon, Traite, I, pt. I, col. 579, felt that they were one-third of a miliarense. Mattingly, NC6, VI (1946), p. 116, n. 8, accepts the CN coins as half-siliquae and equates them with the Vandal coins marked as equalling fifty units. On this basis he gets a solidus of 12,000 nummi, and five nummi then equal one Vandalic denarius. A. Segre, “Inflation and Its Implication in Early Byzantine Times,” Byzantion, XV (1940-1), p. 273, suggests that the CN coins were not reckoned as siliquae but as 250 nummi. This would be equal to 1/33. 6-1/28. 8th of a solidus. It would be calculated as i/30th of the gold coin.

61 Segrh, Byzantion, XV (1940-1), p. 273, follows this system despite this fact. The literary evidence on the number of nummi in a solidus is very contradictory. In Nov. Valent., XVI, 1 (445 A.D.), it is stated that the money-changers may purchase a solidus for 7,000 nummi and sell it for 7,200 nummi. Procopius, Anecdota, XXV, n-12, says that the solidus was valued at 210 folles and that Justinian lowered it to 180 folles, but Procopius, Anecdota, XXII, 38, accuses Peter, the treasurer of Justinian, of issuing gold coins of lower value, and it is likely that this is the reduction referred to in the first passage cited from that author. The number of nummi equivalent to a solidus was regulated by the government very strictly. See the request made to the emperor to raise that number. Symmachus, Relationes, X, 28 (384/5 A.D.). Cf. Mickwitz, Geld undWirtschaft,

CURRENCY IN EARLY BYZANTINE EMPIRE

17

to coincide with the statement of Procopius, if it is assumed that the CN coins are either siliquae or half-siliquae.

These silver coins bearing a mark of value were struck only in the newly re¬ covered province of Italy.65 They cannot be set in any simple relationship with a solidus of any possible number of units or nummi save 6,000 or 12,000. Possibly such a local standard existed in Italy, but more probably the answer is that in the time of Justinian these coins were minted in that province on a system of parallel standards. That is to say that in Italy they were an intermediate form of currency with the bronze coinage on the one side and the gold coinage on the other but without a simple relationship to the gold.66 A guinea in English monetary usage has no simple relationship to the pound of the crown.

It is only proper to point out that this solution is merely an admission of defeat because it remains impossible to establish a simple mathematical relationship between 250 nummi and any of the possible values for the solidus in terms of the statement of Procopius. If the marks on the bronze coinage are in terms of the same units as those on the silver pieces, then a system of parallel standards is necessary to explain those silver coins. If the marks of value are in terms of different units, then our problem must await the discovery of further evidence to indicate the unit expressed by CN, PK€ or PK.

p. 88, n. 34, on this passage. Cassiodorus, Variarum, I, x, 5, says that the “ancients” valued a solidus at 6,000 denarii. The passage is of the early sixth century. B. Hilliger, “Die Kupferrech- nung der spatromischen Kaiserzeit,” Numismatik, II (1933), pp. 56-7, says in an unconvincing statement that this passage from Cassiodorus refers to a valuation begun in the past and still current in the early sixth century. Seeck, ‘‘Die Miinzpolitik Diocletians und seiner Nachfolgcr,” ZfN, XVII (1890), pp. 73-4, felt that the wording of the passage indicated that Cassiodorus had drawn his information from books. He pointed out that St. Epiphanius (Hultsch, Scriptores, I, p. 276), had said that the talent of what seems to be New Testament usage was equivalent to a solidus. Possibly Cassiodorus had this passage at hand. All of these passages concerning gold coinage are discussed in my book on Light Weight Solidi and Byzantine Trade during the Sixth and Seventh Centuries, NNM, No. 138 (New York, 1957).

85 Mickwitz, Die Systeme, p. 20, n. 1, following Wroth, BMC, Byz., I, pp. 21-2, and Wroth, BMC, Vandals, pp. xlixf.

66 It is interesting to note that the Ostrogothic silver coins as well as those struck by Justinian at Ravenna have been found in southern and western Germany. Werner mentions a total of nineteen silver coins of Justinian and forty Ostrogothic silver pieces found in that area. Joachim Werner, Miinzdatierte austrasische Grahfunde (Berlin and Leipzig, 1935), p. 14, in Hans Zeiss, Gcrtnanische Denkmdler der Vdlkcrwanderungszcit, III, issued by the Romisch-Germanischc Kommission des Archaologischen Institutsdes Deutschen Reiches. Apparently the Germanic proclivity for the use of silver currency persisted in the more backward areas throughout the early mediaeval period.

2 Centennial

i8

HOWARD L. ADELSON

In the case of the miliarense, I have already set forth my views in a short note in ANSMN, VI I (1957), pp. 125-35, and it would be a needless repetition to restate them at length here. It need only be noted that the same methods that were applied to the siliqua may be used in that instance. At the start of the fourth century the coin of i/6oth of a pound was probably the miliarense, and that of i/72nd of a pound must have been valued at either U/3 or i1/* siliquae. After the change of the mint ratio the coin of i/72nd of a pound must have served as the miliarense, and the piece of i/6oth of a pound would then have been valued at two carats. In this later case, however, the coin was somewhat too heavy, and very few were issued. Of those that were issued many must have suffered the usual fate of such overweight pieces and been melted down. As a result the government soon ceased to issue them. When in the latter part of the sixth century the coinage of pieces of i/72nd of a pound was resumed, these must have been the miliarenses.

The relationship of the Roman monetary system, insofar as the silver coins are involved, to the systems of the Germanic invaders can now be set forth briefly. The barbarian systems of coinage contained denominations that bore Roman names, but the weights of these denominations did not coincide with those of the imperial issues. Until the individual Germanic people began to strike their own coinages they used Roman issues. When barbarian silver first appeared in the fifth century the Roman coins must have been the half- and quarter-siliquae. The barbarians seem to have struck siliquae, in the case of the Ostrogoths, at the same weight as the Roman half-siliquae of the fifth century or the CN coins of Justinian. The barbarian half-siliquae correspond in weight to the Roman quarter-siliquae or the PK€ coins minted in Italy by Justinian.67 Is it not likely that Justinian valued the Ostrogothic siliqua at 250 nummi and marked his coinage in Italy accordingly ? This would conform with the weights and marks on the coins as well as with the system of parallel standards which has been postulated for that period.

The Vandalic silver coinage approaches the Roman in having three coins; a reduced siliqua of approximately 2.5 grammes marked C; a half-siliqua of ap-

H7 Mommsen-Blacas, III, p. 90, based on J. Friedlaender, Miinzen der Ostgothcn (Berlin, 1S44), p. 17. Friedlaender is followed by most numismatists. Cf. Keary, Coinages of Western Europe from the Fall of the Western Empire under Honorius to Its Reconstruction under Charles the Great (Lon¬ don, 1879), PP- 46ft I Wroth, BMC, Vandals, p. xxix. Mommsen-Blacas, III, pp. 79-80, states that the coinage of the Ostrogoths and Vandals was struck on the Roman standard of i/i44th and i/288th of a pound. Still virtually the only imperial issues of light silver during the fifth century were of 1.3 grammes effective weight.

CURRENCY IN EARLY BYZANTINE EMPIRE

*9

proximately 1.25 grammes marked L; and a quarter-siliqua of approximately °-625 grammes marked XXV. These are estimated theoretical weights based on very little evidence.68 The marks of value correspond to nothing on the imperial issues. The Frankish system, as displayed in the law codes of Merovingian Gaul which speak of a denarius worth i/qoth of a solidus, is also not Roman in form. The innumerable explanations of the Frankish system of coinage still leave the undeniable fact that, though it may be related to the imperial arrangement, it is not a true copy.

It is, of course, true that silver coinage was relatively unimportant in the Byzantine Empire, but it is impossible to speak of the unity of silver coinage in the Mediterranean or Europe after the Germanic invasions. The only coinage that was truly uniform in any sense was that in gold. The gold coinage served as the medium of international exchange and necessarily had to be relatively uniform. Even in the case of gold, however, it is evident that the normal gold coin of the West after the sixth century was the light tremissis while in the Byzantine Empire it was the solidus. The tremissis of the West is even lighter than the same de¬ nomination struck by the Byzantines, and this fact is reflected in the famous light gold coinage of Justinian and later emperors.69 The silver denominations after the middle years of the fourth century seem to have been for local use, and therefore the systems diverged. All used the same names for their denominations and all coins bore a relationship to the solidus, so that exchange was possible, but the monetary unity of the Mediterranean was simply non-existent in the coinage which did not enter into international trade.70

The fact that silver was not widely used in international trade is intimately connected with its disappearance as an important monetary medium. The ex¬ planation of this disappearance is somewhat exaggerated by the fact that it does not seem to have been noticed by the people of the period. In the early empire it was common usage to express value in terms of silver currency. In the fourth century the common method was in terms of a bronze unit of account, and in the sixth century gold itself was used to express value. These changes did not occur with equal rapidity in all parts of what had been the expanded Roman Empire at

Sabatier, Description generate des monnaies byzantines, I, p. 214. Cf. Engel and Serrure, Traite de nuimsmatique du moyen age, I, p. 18.

(,J See my book on Light Weight Solidi and Byzantine Trade.

,0 Cf. R. Blake, The Monetary Reform of Anastasius I and Its Economic Implications,” Studies in the History of Culture, The Disciplines of the Humanities, Dedicated to Waldo Gifford Leland (Menasha, Wisconsin, 1942), p. 8b.

20

HOWARD L. ADELSON

its height, but in the area that remained under imperial jurisdiction these shifts did occur in that sequence. The changes during the fourth century even in the lands under Roman dominion did not proceed with equal rapidity in all corners of the realm. The West retained its use of silver to a greater degree than did the East. The authors in the West did on occasion speak of coins such as the “denarius” which were evidently of sufficient value so that they would have been of silver.71 In the East, particularly in Egypt, where our evidence is clearest, there is no mention in contemporar}^ documents of either the new imperial silver or the new bronze by the names which were in use in the West.72 The papyri of the fourth century continue to use the terminology of the third century to describe the current coinage. The valuation applied to the units implied in these terms fell steadily throughout the century, but the same names continued to be used. Silver currency in the period from Diocletian to Heraclius was of so little impor¬ tance to the economic system and the daily life of the average Egyptian that only in the works of Palladius (fourth century) and Cosmas Indicopleustes (sixth century), of all the literary sources of certain Egyptian origin, are silver coins mentioned.73 This is a very impressive change from the third century with its dependence on “silver” tetradrachms.

The fact that the Alexandrian mint seems to have struck very little silver may be a partial explanation of this phenomenon because, as has been pointed out, silver coinage was primarily for local use. Diocletian had struck some silver at Alexandria, but the principal occupation of that mint was the striking of aes .74 Even this small amount of silver, however, which was struck locally does not seem to have been widely circulated or hoarded. Silver hoards in Egypt in the period after Diocletian are quite unusual.75 There was apparently some payment of taxes in terms of silver bullion during the fourth century but such payments might be commuted into gold at a fixed rate.76 The bronze of units account, the

71 Mattingly and Pearce, NC6, VI (1946), p. 172, point to the fact that Gildas knew of a silver coin current in P>ritain which he called a “denarius,” and which they identify as a Roman siliqua. The Germanic barbarians did strike silver coins. See Mommsen-Blacas, III, pp. 78-80.

72 West and Johnson, Currency in Roman and Byzantine Egypt, pp. 157-8.

73 Ibid., p. 107.

74 Ibid. , pp. 157-S; Mickwitz, Geld und Wirtschaft, p. 98.

75 Segre, Metrologia e circolazione monetaria degli antichi , p. 477, n. 3. This passage refers to miliarenses, hoards of siliquae arc unknown.

76 SB 6086, records just such a commutation of a silver levy. Cf. P.O. 1653 (306 A.D.) ; P. 1 head. 33 (312 A.D.) ; P.O. 2106 (300-16 A.D.) ; P. Bremer 83 (early fourth century) ; P. Flor. 95 (375 A.D.) ; P. Leipz. 61 (375 A.D.); P. Oslo 162 (fourth century).

CURRENCY IN EARLY BYZANTINE EMPIRE

21

drachma and the denarius, were still called silver just as they had been during the third century, and the taxes might be listed under the rubrics of apyupiKcov or XpuaiKcov titAcov, but this does not mean that the coinage in which these taxes were satisfied was either gold or silver.77 In the sixth century, and even in the seventh, taxes were generally reckoned in gold, but the use of the term apyupma to describe such payments is not at all uncommon-78 Even when the payment is specified as vopiapaTa apyupiou there is still some doubt as to whether the account was satisfied in actual silver or in the bronze equivalent.79 The expression Sokipos apyupios occurs only once in the Byzantine period. This instance involves the phrase (TaAavTa) p Soidpov apyuplou tcc cruvayopeva y- “Whether it applies to the forty talents of bronze or to an equation in real silver is not clear. The former is more probable.”80

Some silver coin was definitely in use during the Byzantine period as shown by a building account in which payments were made in silver coin as well as in solidi or myriads of denarii.81 Silver bullion, however, was in more common use, as has been shown, and it is most often described as “asemon” or uncoined.82

The great variation in the weights of the silver coins of the fourth century, however, can be removed from among the number of factors influencing the disappearance of silver as a monetary metal. Even if the weight variations in the first half of the fourth century did render the use of the balance necessary, as Mommsen proposed, still it cannot have been the deciding factor.83 These weight

77 West and Johnson, Currency in Roman and Byzantine Egypt, p. 114.

78 Ibid., p. 120. Cf. P.O. 1911; 1914; 2195; P. Cairo 67138-9.

79 This term occurs twice, SB 4898 and P.O. 2025. See West and Johnson, Currency in Roman and Byzantine Egypt, p. 119, concerning the uncertainty regarding the nature of the currency involved.

80 West and Johnson, loc. cit. These same authors suggest (p. 165) that in the adaeratio of clothing in A.D. 340 the assessment might have been made in silver, but that the payments could have been made in bronze. The expression apyupiou peyaAou (P. Warsaw 28; Kl. Form. 775; SB 6174) may also refer to either bronze or actual silver. West and Johnson, op. cit., p. 120.

81 SB 5300. West and Johnson, op. cit., p. 119.

8- Mickwitz, “New Papyri of Numismatic Interest,’’ NC'°, XVII (1937), p. 142, n. 2, suggests that the term “asemon’’ or “asemon katharon” meant “pure silver’’ in contrast to billon. He points out that “as a matter of fact the meaning ‘silver’ for ‘asemon’ is not known before the fourth century (Preisigke, Worterbuch, cf. C. A. Johnson, A J A, 1934, p. 51).” Therefore he holds that it is not improbable that since in the third century there was no pure silver save uncoined metal and the billon was called “argyrion,” the Egyptians borrowed the term “asemon” to mean “pure silver” whether in bullion or coin during the post-Diocletian period. There is no evidence to sup¬ port this hypothesis. Cf. P. Oslo 162 and P.O. 1223.

83 Mommsen-Blacas, III, pp. 74-5, 158-9.

22 HOWARD L. ADELSON

variations do indicate al-marco striking but nothing more. They cannot have greatly influenced the shift in the form of economic life, since this lessening in the accurarv of minting would be counter-balanced by the lowered value of silver as well as by its relative unimportance as compared with the earlier periods of the empire.84

Silver, according to Mommsen, lost its position as independent money with full bullion value when it was set into a definite relationship with gold by the introduction of the miliarense at i/i,oooth of a pound of gold. This mint ratio meant that the silver coins would rapidly become fiduciary and decline in weight, so that it would become difficult to exchange them for gold.85 Silver would then become unacceptable and would neither be hoarded nor passed easily. It has, however, been shown that during the entire fourth century the attempt was made to preserve and to operate the failing bimetallic system.86 The lightening of the coinage seems to coincide with the notices that are preserved concerning the mint ratio, and the hoards show that clipping of coinage after the reign of Honorius was quite open. The clipping in the British hoards is so open that it is impossible to believe that the clipper expected to perpetrate a fraud. Halving coins could not have been overlooked even if the coins passed by tale. Such clipping was in the nature of an attempt to lighten the coins to accord with the new mint ratio.

Mattingly has pointed out that silver fell in value in relation to gold at a period when the silver coins were generally struck below weight and when the silver hoards are restricted in the main to Britain. In 395 the government called in the “ pecunia Maior.’’87 Mattingly would suggest that the purpose of this move was the recovery of the silver in that denomination. He explains the entire complex of facts by saying that “everything, in fact, except the ratio of gold to silver, points rather to a scarcity than to a superabundance of silver. A possible ex¬ planation might be that silver was replacing gold on a large scale for articles of luxury in trade, and for this reason, silver though not actually scarce, was not readily available for the coinage.”88 According to Mattingly’s hypothesis the silver supplies had been squandered in the manufacture of masses of bullion

84 Mickwitz, Geld tend Wirtschaft, p. 80.

85 Mommsen-Blacas, III, pp. 159ft., see also ibid-., pp. 90-1.

86 Cf. Mickwitz, Die Systeme, p. 67. See also pp. 13-6.

87 See my article “The Bronze Alloys of the Later Roman Empire,” ANSMN, VI (1954), pp. m-129.

88 Mattingly, Roman Coins, p. 232.

CURRENCY IN EARLY BYZANTINE EMPIRE

23

during the third century, and the supplies available for coinage began to fail so that the Byzantine coinage rested on gold and aes.89

Actually the evidence for a scarcity of silver for monetary purposes is not at all incontrovertable. In 384 A.D., when salary schedules were put into law, they were expressed in terms of silver pounds.90 According to one source the expenses of the great Vandalic expedition of 467 were in large measure met by silver payments.91 In the seventh century, when the Empire was at the nadir of its power, we read tales of ships of the Alexandrian church transporting silver on the Adriatic Sea and elsewhere.92 That hoards of silver were restricted in the main to Britain does not indicate that the coins were not minted elsewhere, nor does the fact that silver coin was not a medium of exchange in trade preclude its use as an article of trade. In other words the silver coinage may well have been sold for

99 Ibid., p. 259. Cf. Gibbon, The Decline and Fall of the Roman Empire (New York, n.d.), I, p. 50, points to the falling value of silver as shown by the mint ratio and explains it on the basis of an increase in the silver supply. P. Strauss, “Remarques sur la monnaie de cuivre romaine au IVC siecle,” RN 5, VIII (1944), p. n, n. 12, feels on the basis of C. Th., XI, 21, 2, that the value of silver fell and that this was so because of a decline in the exploitation of mines during the troubled period. If one accepts Seeck’s view, as put forth in ZfN, XVII (1890), p. 152, that a system of parallel standards of gold, silver and bronze was in use both for private business transactions and for the government, then a solution must be found along other lines than those set forth in this paper, but there does not seem to be the slightest reason to assume such a system of parallel standards while there is a good deal of evidence indicating that such open bargaining was not permitted. P. Le Gentilhomme, “Le Monnayage et la circulation monetaire dans les royaumes barbares en Occident (Ve-VIIIc siecle),” RN5, VII (1943), pp. 5iff., does not accept the view that silver coinage was not fiduciary during the fourth century, and he points out that if it were overvalued in the Empire as a fiduciary coin it might migrate to Britain. This would not be so, unless other factors, such as an over-issue of silver, were present, so Le Gentilhomme’s view must be discarded. L. C. West, “The Coinage of Diocletian and the Edict of Prices,” in P. R. Coleman-Norton, Studies in Roman Economic and Social History in Honor of Allan Chester fohnson, p. 302, suggests two possible explanations for the failure of Diocletian’s monetary reform. The first is that Diocletian attempted to establish a bimetallic coinage as Augustus had done. The second is that Diocletian put too low a value on silver in relation to bronze, “and with the great mass of bronze in circulation, the silver disappeared.” Dr. Robert B. Warren in a letter to West {ibid., p. 302, n. 15) suggests that the ratio 1:12.5 which West gives as the highest possible gold-silver ratio during the reign of Diocletian is unusually high. He goes on to say that if one takes that ratio as indicating a real and growing scarcity of silver then that very scarcity of silver would make it impossible to establish a fixed gold-silver and silver-bronze ratio. He proposes that an investigation of what was transpiring at the silver mines in Spain might explain a good deal.

90 C. Th., VI, 4, 25.

91 John Lydus, Dc Magistratibus, III, 43 (ed. Bonn, p. 237).

92 Vita S. Johannes Elecm., c. 28. Cf. ibid., c. 10.

24

HOWARD L. ADELSON

bullion by the Romans and used as coin by the British and the Germans. It is now known that silver coinage was not as rigidly restricted to Britain as was previously thought to be the case.93 Two silver hoards from the Lower Danube region have been scientifically treated, and some small ones from France are known.94 In addition the finds of Ostrogothic and Ravennese silver coins indicate that the Germanic peoples in southern and western Germany continued to use coins of that metal. The coins found in these deposits as well as in those of Britain show that in the fourth century imperial mints in all parts of the Roman world struck silver coins. In the British hoards, as is to be expected, the western mints are more numerously represented, but in the two Danubian hoards the coins of the eastern mints are more common.95 This phenomenon Mickwitz pro¬ posed to explain on the basis of a specific British provincial customary preference for silver. He pointed to the fact that earlier investigators had noted such a customary preference in the second century. This poor province, according to Mickwitz, carried on most of its business transactions in silver, and, as a result, the quantity of silver in the province increased greatly. The absence of silver hoards elsewhere, of course, does not indicate that silver did not enter into trade there.96

The fact remains, however, that it is not only the lack of silver hoards in the interior of the Empire, but even the absence of direct literary evidence of its use that causes one to speak of the disappearance of silver. The solution proposed by Mickwitz amounts to little more than a statement that this is illusory, and that the concentration of silver in Britain is only relative. Another hypothesis, differ¬ ent from all those which have been offered, presents what seems like the most probable solution. It has already been pointed out that the mint ratio far from showing an increased value for silver, such as would be the case in the event of a scarcity, actually demonstrates just the opposite. For the period from the reign of Nero to the end of the second century and possibly even as late as the reform of Diocletian this mint ratio was approximately i:i2.9T Only in the reign of Julian do we have the slightest suggestion that such a ratio might have been accepted as the standard of the imperial government. At all other periods after

93 Cf. Mommsen-Blacas, III, pp. 132-3.

94 Mickwitz, Die Systeme, pp. 11-12.

95 hoc. cit. The figures for several hoards are given there.

9,i For a full statement of this hypothesis see op. cit., pp. 12-13.

97 See L. C. West, Gold and Silver Standards in the Roman Empire, NNM, No. 94 (1941), pp. iff.

CURRENCY IN EARLY BYZANTINE EMPIRE

25

the reign of Diocletian the ratio varied between 1:14.4 and 1:18. This may not demonstrate that silver became more plentiful, as Gibbon suggested, but it may indicate the increased demand for gold coinage during the fourth century.98

Mattingly has suggested that the Romans withdrew all gold from Britain when they abandoned the island, but of this there is no positive evidence.99 Such a move would have raised the value of gold in relation to silver in Britain to such an extent that silver would actually have left Britain in the form of gold purchases. Hoards of silver would have been more generally found elsewhere. Silver hoards are not restricted to Britain either because the Roman government removed the gold from the island or because the workers in the silver mines were paid in terms of silver, as Evans suggested.100 Silver was struck throughout the Empire, and it is illogical to assume that the government took silver from Britain to strike it elsewhere and then shipped it back to the island.

lliere is good reason, however, to believe that Roman coins were not always used merely as a medium of exchange. Indeed there is a law which specifically prohibits the export or trading in coinage.101 Such trading in coinage can only have been profitable because there was a differential in the value of the coin between one region and another. It is not in the least improbable that the Ger¬ manic peoples and outlying provinces in places such as Britain and the lower Danube valley might have continued to adhere to the older mint ratio, and that Roman merchants chose to sell currency in place of other articles. As the Germanic peoples became settled and gained closer contacts with the Romans, silver coinage lost its importance for them. Nevertheless the fact that the silver issues of Diocletian resembled those of Nero in weight and could be used by the barbarians must have created a demand for them. Only in the Lower Danube valley, where Germans must have been numerous in the fourth century, are hoards of imperial silver founds outside of Britain and adjacent Gaul. Britain was not firmly united to the Empire during the major portion of the fourth century, and in the fifth century it was definitely abandoned. The adherence to the old mint ratio there would cause a flow of silver coinage to the island. Silver was more valuable in Britain than elsewhere. Little silver was struck in any event by the mints of the

98 Cf. Mickwitz, Die Systemc, pp. 14-15.

99 Mattingly, Roman Coins, p. 257.

100 Evans, NC\ XV (1915), pp. 498ft.

101 C. Th., IX, 23, 1. This law is actually dated in 356, but Mommsen by comparing it with other edicts addressed to the Praetorian Prefect Rufinus has dated it in 352 A.D. See Mommsen's comments on C. 7 //., II, 9, i, in the Mommsen-Mayer edition.

26

HOWARD L. ADELSON

Empire, and because the mint ratio established by the government valued silver too low, coins of that metal were not generally acceptable save by merchants who intended to ship them to Britain or to the Germanic barbarians. Therefore when taxes were assessed in terms of gold and silver, the payments would, in most cases, require either commutations of at least some portion of the silver assessment into gold or the use of silver bullion in satisfaction of the exaction. This inter¬ pretation of the facts also serves to explain the mint ratio used by Julian. Perhaps that Emperor, who had great financial difficulties and was most aware of con¬ ditions along the western frontier, hoped to put an end to trading in silver coins by maintaining within the Empire the same ratio as prevailed among the barbari¬ ans. This would have ended the export of silver.

Admittedly this solution is not proven, but it does serve to explain all of the facts in what seems to be a more logical fashion than any of the preceding ones. The evidence is too slight to present any answer to the problem which is com¬ pletely verifiable, but there is an analogous modern counterpart. This counter¬ part is found in the history of the Maria Theresa taler in modern times. In our own days this coin was current only in the Near East, i.e. Arabia and Abyssinia. It was struck in Italy, Britain and Austria solely for the purpose of facilitating trade in the Near East, and its history is an important source for Anglo-Italian rivalry in that region.102 This taler could be sold at a premium in the Near East but was virtually worthless elsewhere. The result has been the accumulation of hoards of this coin there, and a virtual absence of the coin where it was actually struck. In the same sense it is known that, at least in the early period of the Roman Empire, the Germanic peoples showed a decided preference for silver as opposed to gold as a medium of exchange. It is hardly likely that the shift to gold could be ac¬ complished very quickly in the fourth century, and it is most probable that in the early fifth century and even earlier, when Britain was gradually losing strong contact with Rome, this preference for silver might have remained strong among the Germans who were settling there and caused it to have a premium value.

102 J. Hans, Zwei Jahrhundertc Maria-Theresien Taler, 7751-295/ (n.p., 1950), pp. iff., is the best modern work treating this matter.

THE PORTRAIT OF CAESAR ON THE DENARII OF 44B.C. AND THE SEQUENCE OF THE ISSUES

Andrew AlfOldi

I

Some years ago I tried to exploit the coins as a source of the history of the crucial early months of 44 B.C.1 Since then, several specialists have taken a position for or against my chronological reconstruction. K. Kraft2 accepts the sequence of the issues I proposed, whereas C. M. Kraay3 thinks that my numis¬ matic arguments, “if they were to prove correct, would be far-reaching in their results; but there is reason to think that they are, for the most part, unsound,

1 Alfoldi, Studien iiber Caesars Monarchic, I,” K. Humanistiska Vetenskapssamfundets i Lund Arsberdttelse, 1952-53, referred to hereafter as Studien. My sincere thanks are due to Mrs. E. Clain-Stefanelli for the photographs of the Vatican Collection, to A. Alfoldi, Jr. for those of the Italian and other collections, to H. A. Cahn for those of the collections of W. Niggeler and A. Voirol. Miss M. Thompson selected the coins of the American Numismatic Society for its splendid photographer, Mr. Raymond W. Johnson. T. V. Buttrey kindly polished my rough English.

2 K. Kraft, “Der goldene Kranz Caesars und der Kampf um die Entlarvung des Tyrannen,” JNG, I952/53 (I955)- We will return elsewhere to his important discovery of the exact nature of the wreath which Caesar wears on his portrait-coins; I do not accept the historical evaluation of his discovery and maintain against it my own views expressed in my Studien. O. Ulrich-Bansa discussed the issues of 44 B.C. in KIN, 56 (1954), pp. 52 sqq. He had not yet seen my Studien, so I need not comment on his article in this connection. Where he differs from my results, 1 prefer my own onclusions.

:1 C. M. Kraay, “ Caesar’s quattuorviri of 44 B.C. : the arrangement of their issues,” NC, 1054. pp. 18 sqq

27

28

ANDREW ALEOLD1

and therefore, do not illuminate Caesar’s policy in the way that Alfoldi claims.” In spite of this sweeping verdict - the reader of this note will be able to decide whether it is correct or not - I wish to welcome his contribution, both because he has made some sound observations on details hitherto unnoticed, and also for a more general reason.

A historian who uses the coins as well as many other groups of evidence for the reconstruction of the events - and I made my attempt as a historian - cannot go so far in detail as the specialist whose aim is to illuminate a more restricted field. Thus the intervention of a specialist is a healthy control and a supplement of every branch of historical research. Also Kraay’s attempt to upset my recon¬ struction and replace it by another offers a good starting-point for checking the facts anew; it helps to strengthen and correct my own results.4

I tried to show in my Studien that, as a consequence of the political recon¬ siderations and setbacks of the dictator, the coins of early 44 display first a short interval when the senior monetalis still had the freedom to glorify his ancestor Sulla, and then three successive phases of the titulature of Caesar, who calls himself first dictator quartum, then imperator and pontijex maximus, and finally dictator perpetuo. Further, I tried to show that Antony’s conception becomes apparent when after the Ides of March Caesar appears first as dictator perpetuo but veiled with the solemn ritual garb of the pontijex maximus , because a little later a concomitant obverse of Antony himself reveals the intentions of the pre¬ sumptive heir of the crown. I expounded also my conviction that, after the aboli¬ tion of the dictatorship, the title parens patriae and the temple of the divine dementia of Caesar were employed to foment hatred against his foes; and, lastly, that the reference to the games in honor of the dead Caesar as the new founder of the city at the Parilia closes the coinage of the year.

Kraay accepts my interpretation for the interval after the death of Caesar, but tries to supplant it with another reconstruction before that event. Though he objects to the use of the varying titles as a main criterion of the sequence of the issues, his own argument is based on a supposition of the same kind: he thinks (Kraay, p. 23 sq.) that he can separate a series with the title imperator as “military” issues from those with dictator ( quartum , resp. perpetuo ) as “civilian”. He believes that the latter (his group A) might be termed the normal issues of the state on

4 Kraay-l cite his paper henceforth by his name alone - states (p. 18) that Chilo struck (along with

P. Clodius, Regulus and Longus) also in the names of Octavian and Antony. This is a lapsus calami,

cf. Sydenham, Coins of the Roman Republic, no. 1088-9, referred to hereafter as Sydenham.

PORTRAIT OF CAESAR ON DENARII OF 44 B.C.

29

which Caesar defined the magistracy by which he governed; the issues of his group B (with imperator) would be the stockpile of coins for the Parthian war and were signed by Caesar as commander-in-chief.

I am glad to say that Kraay contributed significantly to the distinction of the varieties of the Venus-types on the reverses and the distribution of them between the workshops (Kraay, p. 28 sq.). I acknowledge gratefully that 1 am using his obser¬ vations in my synoptic table (pp. 43-44) of the Venus-types, to which all numbers in brackets without additional indication in the text of this paper refer. This table makes evident the comprehensive planning and careful distribution of the types. The central scheme seems to me not to be the work of the four moneyers, but to come from the intelligent Greek slaves, the pcculiares servi, whose interference was reckoned with the criminal ambitions of the dictator, which, overthrowing ancestral institutions, induced his murder. Disregarding the fact upon which we shall dwell once more below, that the complete plan was somewhat disturbed by the events, the picture that it offers is that of the differentiation of tasks and timing in one sole establishment. The twisting of the issues into two isolated parts does not work. The definition of those two supposed units of production is not consistent. Kraay speaks of two “workshops” (p. 28) of “the mint of the series B” (ibid., p. 30), “the mints A and B” (p. 30, n. 30), of two “geographically separate” establishments, and again of the exceptional and temporary establishment of the imperator- series (pp. 30 sq.). On the other hand, he admits that these expressions could give a deceptive impression of a wide separation ; he stresses that both series were struck in Rome under the same direction. Now, fortunately, we know that there was a second, geographically separated mint in Rome, under separate management. This was the mint for gold, working under praetors and city-pre¬ fects, which was separated from the main mint - surely to prevent political enemies from seizing it - and jealously guarded. I assume it was on the Janiculum.4'1 We know perfectly the uncouth style of this second mint5 which has nothing in common in style and so nothing to do with our denarii. We also know well Caesar’s numerous military issues. There are only two exceptional cases in which he calls himself imperator : the very scarce issue, Sydenham 1016, and the denarius struck in Sicily, Sydenham 1022. The latter case offers no support for the reasoning of Kraay: the proconsul A. Allienus had C- CAESAR IMP-COS-ITER inscribed on the

43 I his argument will he discussed soon in a paper on the Golden Chair of Caesar.

5 M. v. Bahrfeldt, “Die romische Goklmiinzenpragung wahrend der Republik und unter Augustus.’

Miinzsludicn, I (Halle, 1923), pi. IV, 1-22 and pp. 35 sqq.

3°

ANDREW ALFOLDI

silver only because Caesar did not possess at that time the dictatorship. I would like to suggest the same for Sydenham 1016 because when Caesar had the im- perium maximum of the dictatorship this superseded the title imperator. The aurei with DIC. QVAR. - COS. QVINC. produced undoubtedly for the army of the Parthian war in 44 in huge quantities6 in the second mint of Rome do not have this title, neither do the gold and silver of the moneta castrensis with LII, nor the elephant -denarii, etc. On the other hand the dictatorship was hated in Rome because it meant the overthrow of civilian freedom by the commander of the armed forces: to reserve mention of the dictatorship to the “civilian” issues would not fit in the categories of Roman thinking. But there are also other diffi¬ culties obstructing the possibility of distributing the denarii of 44 B.C. between two “establishments.” To assume that Mettius displays in one “mint” the glory of his own family with Juno Sospita and DICT. QVART., and simultaneously in the other “mint” begins with the special reverse displaying not his own ancestral divinity but that of Caesar, anticipating the commencement of the Venus-types of his colleagues who would use it only in a later phase, makes little sense. Kraay himself notes7 two specimens of the Venus of Macer coupled irregularly with IMPER - obverses of Mettius. But he overlooked that in this case the special reverse-variety (3) of Macer is not that which is always coupled with the IMP and star series, but that which goes always with the DICT. PERPETVO obverses. These hybrids clearly originated in the same mint. The dies of the IMP- and DICT.- series did not merge with the death of Caesar,8 but were always produced in the same mint.

A very good observation of Kraay is the fact I overlooked, that the IMP series of Mettius, in spite of the common reverse and the correlation clearly accentuated by the mint-marks, are split into two distinct groups: the series G-L with good, and the series A-E with coarse portraits. But it is impossible to generalize this very special case, as we shall see, and it is a simplification of a rather complicated situation to say that the imperator- group should have two parallel series, the one employing the fine style, the other the inferior. The inferior style is due to the cancellation of this half of the emission, of which the stockpile of reverses has been used later for an emergency issue.

Kraay’s argument is accepted without any reservation by R. A. G. Carson.9

6 M. v. Bahrfcldt, op. cit., pp. 39 sq., no. 23 and pi. IV, 14-16.

7 Kraay, p. 26, n. 23. Cf. also O. Ulrich-Bansa, RIN, 56 (1954), pi. 2, 22.

8 Kraay, p. 27. 9 R. A. G. Carson, Gnomon, 28 (1956), pp. 1S1 sqq.

PORTRAIT OF CAESAR ON DENARII OF 44 B.C.

31

We need to discuss only what he adds of his own. Some questions, such as the Hague denarius with Caesar’s diadem (Plate 1,8) will be discussed below, ‘so the following remarks will suffice here. There is a denarius of Buca10 with the head of Venus and the famous dream of Sulla on its reverse. Carson thinks - like Kraay11 - that Buca could not coin in two non-consecutive issues, and he would let the Sullan Venus run in pair with Mettius’ DICT. QVART.-type. But recent studies ha\ e elucidated somewhat the customary sequence of the supervisors in those years. I have shown12 that after Octavian’s marcia su Roma in 43, Flaminius Chilo, the senioi quattuoYviy struck first and alone, proudly underlining his right to priority with the statement pr(imus) fl(avit) ;13 after him, the pair Ti. Sempronius Gracchus and O. V'oconius Vitulus run the coin production. For the gold-issues of the following year, Th. V. Buttrey14 established the exact order: the senior officer was L. Livineius Regulus, after him P. Clodius signed the aurei, then L. Mussidius Longus, who was followed by C. Vibius Varus; but for a new' reverse- type Mussidius Longus w’as employed again. So contemporaneous practice demon¬ strates the sole rule and signature of the senior officer at the beginning, but admits the possibility of employing either one or two officials at the same time after him,

as well as the return of a monetalis as supervisor later, for a second period of activity.

Kraay legarded Mettius as the senior officer in 44* whereas Buttrey111 confirmed m\ chronology which put the Sudan Venus of Buca at the head of the emissions. In fact, on a small silver denomination17 Buca has the title Hll VIR used under Caesai only by the first ranking member of the Collegium of the moneyers; and Maridianus discreetly avoids express mention of it,19 demonstrating by a trans¬ parent hint that only Buca was allowed to use it.20

10 Studien, pi. I, 1-4.

11 Kraay, p. 27.

12 A- Alfoldi, “Portratkunst und Politik in 43 v. Chr.,” Netherlands Year-Book for History of Art

V. 1954. PP- 153 sqq.

13 Til- V- Futtrey, The Triumviral Portrait Gold of the Quattuorviri Monetales of 42 B.C., NNM,

No. 137 (New York, 1956), p. 39, n. 81, shares my views. 14 Buttrey, op. cit., passim.

I; kraay, p. 31. 16 Buttrey, op. cit., p. 39, n. 81. 17 Sydenham, 1065.

18 A. Alfoldi, “Studien zur Zeitfolge der Miinzpragung der romischen Republik,” Schweizerische A umismatische Rundschau, 36 (1954), pp. 28 sq.

18 He symbolizes with the saltire in which his name is written, Studien, pi. XIV, 4-6, the fourfold team of his office, but adds a.a.a.f.f., which is not against the Caesarian rule, the title of Mil VIR being reserved for the senior.

What Kraay, p. 27, n. 47, says against this is not persuasive, at least for me.

ANDREW ALFOLDI

32

If Carson21 doubts that the portraits of Caesar began to be struck only after an interval, rather than beginning immediately after the New Year, we have now decisive evidence for it through the discovery of Kraft that the wreath on Caesar’s head is a golden one, made in archaic Etruscan style. We know, too, that the high boots and the purple cloak of Caesar were anachronistic features of an archaizing royal attire. Caesar appeared in the purple robes with the golden wreath and sat on the golden chair awarded to him first on February 14th in 44. 22 This means that on the solemn days just after the New Year, he did not yet have those distinctions. If they were voted only shortly before February 14th, the portrait with the golden wreath could not have appeared much earlier on the denarii. Before this first issue of Mettius with DICT. QVART. we have a gap of some weeks from the New Year on.23 The type of Buca with Sulla’s dream fits ex¬ cellently in this frame; he could display Sulla’s dream in the first and foremost silver issue of the year, on account of the leniency and magnanimity, the dementia, of Caesar. In my mind, it would go too far to suppose, as Kraay does, that this remarkable witness of the free decision of the monetalis, so abhorrent from the uniformly planned quadrupartite complex of the later Venus-reverses, and so inconsistent with the portrait of the ruler obligatory since the first Mettius- emission, could go along with these monarchic features and could have been issued with them. The time was over for such individual and heretical types after the total concentration of the coinage on the person of the ruler.

Thus it is evident - even without the Hague-denarius with the diadem - that the DICT. QVART. issue of Mettius must have been continued until a date very close to the Lupercalia on February 15th.

In opposition to Kraay,24 Carson20 does not believe that the veiled head is post¬ humous. But this is not a mere guess. The veiled head of the later Pariliaissue struck together with the veiled head of Antony, bearded in sign of mourning, is coupled with a reverse die26 so near the other that the existence of common dies

21 Gnomon, 28 (1956), p. 1S3.

22 Cf. Studien, I, pp. 21 sq.

23 How large an amount of the denarii of Mettius with DICT. QVART. could have been produced, if begun with the New Year, is shown by the enormous quantity of gold with DIC. QVAR.-COS QVINC.

24 Kraay, p. 18.

25 Gnomon, 28 (1956), p. 1S3.

26 Studien, I, pi. XV, 1 where the reverse was unfortunately exchanged with that of XV, 3 by the printer. Cp. also RIN, 56 (1954), pi. 2, 34 with my Studien, pi. XV, 2.

PORTRAIT OF CAESAR ON DENARII OF 44 B.C.

33

must be assumed. The carefully harmonized heads of Antony and Caesar, both with veil, both with implements of state-priests, establish contemporaneity and exclude the supposition of subsequent issues. Therefore, the previous emission with veiled head must also indicate the intention of Antony to exhibit the treacher¬ ously assassinated high priest of the Roman people, to enhance the sacrilegous act of the murderers. But, quite apart from these considerations, the veiled head appears only after the wreathed head with DICT. PERPETVO under the auspices of Macer, and only after Buca - who struck a considerable quantity of DICT. PERPETVO types without veil - was eliminated. There is no possibility of placing the veiled head before March 15 : there is no room for it before that date as we shall see. T he association of the posthumous parens patriae issues with the abolition of the dictatorship and the attribution of the desultores to the Parilia (accepted by Kraay and denied by Carson) are reaffirmed again by the chronolog¬ ical results expounded below. It is plain that Octavian was not present at the games as Dio thought; the young Ouintus Cicero celebrated them in the hope of gaining favor with Antony.

II

One of the ways which allow us to understand the chronology more fully is the reconstruction of the development of Caesar’s portrait on the coins, which will reveal itself as an evolution from a single root, but with many successive ramifi¬ cations. In the same way as prehistoric archaeology establishes a sequence for certain categories of objects in order to base the chronological sequence on a typological chain, so the successive stages of the evolution of Caesar’s portrait- types offer a similar possibility. My plan was and still is to collect the denarii struck in 44 B.C. from the collections all over the world; but a survey on the ground of a rather restricted amount of material should suffice for a first draft. This shows the following branches and stages of the typological aspects of Caesar’s portrait ;

1. The head of the denarii of M. Mettius (Plate 1, 1-8; II, 1,3; IV, 1) has un¬ questionably the high priority over all the other types. Since there is no second starting point for the evolution, Kraay’s assumption of a second “establishment” within the frame of the mint of the denarii in 44 B.C. has no typological evidence. It is also to be underlined that, though the imitations and retouches of the chang¬ ing dies break the way for a gentle and unobtrusive transformation, the general

3 Centennial

34

ANDREW ALFOLDI

characteristics of this portrait-type remain stable. Thus, we shall call it henceforth the “prototype.”

2. Next to the prototype stand the portraits with the Mettius-reverse (ioa) (Plate 1,9-24; II, 10; V,5), representing a special subtype. The progress of the stylistic development can be checked by the succession of the mint letters G-L. The starting point of this is a somewhat enlarged imitation of the prototype (Plate 1,9), carefully performed, though the fine curve of the nose is straightened, evidently against Caesar’s natural appearance. The size of the head is already enlarged on the earliest dies (with G), but its execution remains constantly on the same level. Whereas the hair is always combed downwards on the neck of the prototype, here it is with rare exceptions (Plate 1, 10) more or less sharply brushed backwards and up. There is a noticeable endeavor by the die-cutter to make the face smooth and beautiful, younger than it was in reality (Plate 1,10,14,15) which points towards a marked encroachment on the likeness of the portrait in the course of the subsequent issues. Only the transverse striae on the abnormally elongated neck represent a dissonant “cacophonic” element, in terms of the con¬ temporary taste. But with the letters Iv and L on the reverse, i.e., with the last ones of the sequence applied as special marks of control (Plate 1,20-24), the features of the head begin to be distorted through technical deficiencies. The metal of the dies was not hard enough to bear the brunt of the wear and tear, and qualified engravers were not there in sufficient number to supply first-class sub¬ stitutes for these. It became necessary to mend the worn-out dies which were damaged, above all, on the nose. Thus the nose is very frequently retouched by a pointed tool and the mechanical reproduction of previous dies perpetuates its deformation. This is the main reason for the ever-growing transformation of Caesar’s profile (Plate 1,23), often reaching the grade of a caricature (Plate 1,24; for other types especially Plate V, 21-24). The high pressure of the ceaselessly changing types of 44 B.C. on the die-production enhanced the speed of this process considerably.

The place of this head-type of Mettius with CAESAR IMP in the evolution can be determined by the following indications: (a) the dependence on, and the close connection with, the prototype defines our head-type as one immediately following the DICT. QVART obverses; (b) several varieties are almost identical with those of the IM.-P.M. heads of Buca (cf. Plate 1,22 with II, 20 and II, 10 with 11,9), which implies an approximate contemporaneity. 1 he profile with the aquiline nose which emerges at the end of the emission of this group (Plate 1,23)

PORTRAIT OF CAESAR ON DENARII OF 44 B.C.

35

is the model of the head struck under the supervision of Macer with IMP and star (Plate 1 1 ,20-21) ; the latter then seems to have followed immediately the IM.- P.M. series of Buca.

3. The issue of P. Aemilius Buca with the legend IM.-P.M. and the crescent of the moon (Plate 1 1, 2, 4-9, 11-13) is directly based, like the previous one of M. Mettius, on the prototype (cf. Plate II, i with 2, and 11,3 with 4). The natu¬ ralistic modelling of the flesh is vanishing, the deep wrinkle which splits the face vertically behind the profile (Plate 11,3) is exaggerated (Plate 11,5); the chin is bigger, so that the proportions of the features and the character of the face are greatly altered. I he whole head is enlarged, exactly as on the previous head-type. But nevertheless, the style of this portrait of the IM.-P.M. issue of Buca is stable and consistent. So it, too, forms an individual phase within the frame of the multifarious evolution. Its good quality is illustrated by the deterioration of the workmanship in the successive head-types based on it (Plate 11,14-16). We have already seen that this issue was produced along with the IMP-issue (10 a) of Mettius.

4. We have seen that the IMP-and-star obverse of Macer (1-2) closely followed the IMP obverse (10a) of Mettius. I he head with the star and IMP legend (Plate IT17-24; III,i-io; IV, 21, 24) grows very much in the course of its production, from the smallest variation (Plate 11,17) to the largest (Plate 111,9-10). It starts with a portrait-type (Plate II, 17) based not on the prototype, but on the head of the previous issue (Plate 11,5)- The deformation of features on the suc¬ cessive dies is much stronger than it was before; many varieties show very different faces indeed. Nevertheless, the issue represents in its style a special coherent unit and an intermediary stage between the head-types of the earlier IMP . and of the subsequent DICT. PERPETVO issues. For the connection with the latter, Plate IV, 21 is to be compared with IV, 22 and IV, 23 with IV, 24.

A very essential new trait of this stage is that the power to maintain a certain artistic level vanished, and in spite of the excellent dies introduced at the beginning of some following issues, the deterioration is abrupt. The die-engravers show a complete failure to overcome the strain of ever-changing devices.

5- 1 he portraits of the next short but crucial and intricate emissions with the new legend CAESAR DICT. PERPETVO were produced under the supervision of three moneyers. Buca has no veiled head, Maridianus has only that; Macer strikes both with wreath only, and with veil. The study of the portraits reaffirms that the wreath without veil belongs to the earlier section of this phase of pro-

)•

ANDREW ALFOLDI

36

duction. The complicated characteristics of this earlier stage of the DICT. PERPETVO issues will be illustrated here only by some selected samples suffi¬ cient for our present purpose.

Hitherto every portrait-group corresponded to a single reverse, as was the original intent. Macer marks the innovation by changing the direction in which his cognomen is written (compare (2) with (3)), and by introducing another variety of his Venus-reverse with shield, with the cognomen written counterclockwise at first (4) and, at the very end, clockwise (5). These varieties correspond to three of Buca, who first only distinguishes the new start with the abbreviated form of his name (8) ; but he adds new reverses to this, one with the sitting Venus (9) and one with a group of symbols of the new Golden Age (6).

The first dies of Buca are excellent, made presumably by the best man who created the first portrait with DICT. QVART., but the level drops instantly on the subsequent dies. The only reverse, if my incomplete collection of materials does not deceive me, which is still connected with an individual development of the portrait, is the group of symbols of world-happiness (6). The head (Plate V,9), which may be the primary die of the series, is finely engraved, though it is a derivate which lost much of Caesar’s resemblance. Then the nose becomes upturned (Plate V,io,iq), the jaw too much stressed, the chin drawn backwards, the neck too much elongated (Plate V,ii-i3), the face more rugged. The outline of the nose varies (Plate V, 11,12, 13) and becomes bulbous on the coarse dies at the end (Plate V,i5).

Apart from this case, the obverse dies are used now not only by one monetalis, but by two members of their quattuorvirate; and they are no longer restricted to one reverse, but are coupled with all of them. The combinations of the same obverse and reverse dies between the different varieties struck simultaneously will help greatly in the complete reconstruction in the future. But this opportunity arises only from this point on. The die-engravers are no longer able to solve the technical problems of die-production owing to the multiplication of their task through the ever-changing new devices. This is mirrored by the quickly vanishing artistic level of the new portrait varieties, in spite of the existence of excellent dies with which the series of Buca started. The head on Plate IV, 2 is, e.g., the best imitation of the prototype (cf. Plate IV, i) we have. But the subsequent dies first reproduce the same type of face in a very summary way (Plate IV, 3-4), then the following are done more roughly (Plate IV, 5-6), and they deteriorate into a clumsy face which has nothing to do with Caesar. The variety Plate IV, ii is re-engraved

PORTRAIT OF CAESAR ON DENARII OF 44 B.C.

37

with a veil (Plate VI, 15) for Macer (4). Now, this chain of obverse-dies is used with the standing Venus (8) of Buca (Plate IV, 2, 6, 8), with the seated Venus (9) of the same (Plate IV,io,ii) as well as with the Venus-reverse (4) of Macer with shield (Plate IV, 3, 5, 7; VI, 15) and with the latest variety of the same (5) with cognomen written counterclockwise (Plate IV, 4). A ramification of this chain of portrait-punches starts with Plate IV, 9 (9), after the nose has been patched up through piercing its nostril (Plate IV, 12) ; thus by the same hand later the head- type with veil (Plate VI, 23) (5), employed also by Maridianus (Plate VI, 24).

Other smaller varieties of the same portrait-pattern are a head-die of Macer (4) rather well done (Plate V4), a rather distorted version (6) with the name of Buca (Plate V,3) and another clumsy head (8) of Buca (Plate V,i8). The variety with a big hooked nose can be found as well (3) for Macer (Plate V,i7) as for Buca (Plate V, 18-19; VI, 1 (8) and 2 (9)). The caricature Plate V,23 with the standing Venus (8) of Buca (Plate V, 21-22) is yet worse and is even re-engraved for Maridianus with a veil (Plate V,24).

The insufficient quantity of obverse-dies compelled the personnel to use the outworn and distorted head-dies. The same is to be seen in a head-punch with straight nose produced as well for Buca (8), Plate IV, 17, as for Macer (3), Plate

IV, 18. 19. A very successful enlargement of the head-punch Plate IV, 2 is Plate

V, i (9). The horizontal slit on the nostril is due to the engraver who made the obverse dies VI, 12 and 14. The abrupt deterioration of that excellent pattern (Plate V,i) is illustrated by the head of Plate V,2.

The transition to the veiled head is also clearly to follow on the more round- headed portrait variety of Macer Plate VI, 4-6 (3). The coarse profile of Plate

VI, 10 (3) could have been brought about by the deformation of the nose of the Macer dies represented by Plate VI, 8 (3) and VI, 9 (4).

6. The earlier DICT. PERPETVO emissions are the key to the chronological determination of the postponed issue of M. Mettius (10b). Though the reverses signed by A-E (10b) were prepared without doubt together with the others (10 a) signed by G-L, the heads (Plate III, 12-15, 17-21, 23-24) are a rough continuation of that on the previous DICT. PERPETVO issues just described, without any good new dies manufactured especially for them. Their dependence on the previous DICT. PERPETVO obverses can be established by the following examples:

a) Plate II 1,12 (with letter A on the reverse) copies Plate III.ii (3).

b) Plate II 1,15 (with letter C) is an imitation of Plate 1 11, 16 (3).

3§

ANDREW ALFOLDI

c) Plate III, 21 imitates Plate 111,22 (6).

d) Plate IV, 13 - a hybrid with a reverse of Macer (3) - and Plate IV, 14 (with letter B) are closely akin to Plate IV, 12 (8).

e) Plate IV, 16 with the deformed nose relies on the head, Plate IV, 15 (8).

f) The completely worn-out, but re-engraved head Plate IV, 20 (with letter C, altered from B) is based on Plate IV, 19 (8).

g) The coarse head Plate V,i6 (with letter D) roughly copies Plate V,i5 (6).

h) The head Plate V,2o copies the Plate V,i9 (8).

i) The face of Plate VI, 7 (with letter B) copies the profile Plate VI, 4-5 (3) and is very close to its veiled variety Plate VI, 6 (3).

If anyone unfamiliar with the style should be unable to accept this proof, he would be obliged to do so in view of Macer’s reverse (3) struck exclusively with the DICT. PERPETVO obverses, but also coupled with the IMPER legend.

It can be no surprise that heads of the first half of the Mettius-issue (10 a) were also imitated, as a comparison of Plate V,5 (letter I) with V,6 (letter B) makes cer¬ tain. But this is only a terminus post quern. The dependence on the DICT. PERPETVO issues decides the chronology. What Kraay supposed (p. 30), namely, that Macer uses some obverse dies of Mettius A-E (10b), is erroneous. The con¬ trary is right.

7. The first issues with veiled head belong to Macer. There was no attempt to introduce new and better head-dies. Most of the heads (Plate VT11-17, 19-20) are coupled with one and the same reverse (4), whereas the other reverses of the previous emissions were used only in a very short interval, as (3), Plate VI, 18 and (5), Plate VI, 23. Several examples illustrate how the nose was damaged and deformed. Plate VI, 20 shows a deformation by the double or treble outline, caused in sinking the head-punch. Plate VI, 16 has the nostrils filled with melted metal; Plate VI, 12 and 14 demonstrate how this was cut away by a horizontal slit. Plate VI, 17 has the nose entirely replaced.

With C. Cossutius Maridianus a last finely engraved head of Caesar appears, Plate VI, 21 (11), modelled on the prototype, but with a somewhat protruding and elongated nose. Degeneration follows instantly on the reproduction of the die, as Plate VI, 22 demonstrates. The head-type, Plate VI, 23, emerging at the end of the supervising activity of Macer (5), passes over to Maridianus, Plate VI, 24 (with name in saltire on the reverse). But I do not intend to deal fully with the veiled head. It has no value for the portrait of Caesar and the chronology is so

PORTRAIT OF CAESAR ON DENARII OF 44 13. C.

39

firmly established (after the supposed doubling with the imperator- issues has been proved to be non-existent) that there is no need to pursue the typological develop¬ ment any further.

Our conclusion for the problem of Caesar’s portrait is as follows: there is only one of his portraits on the coins which has the claim to be authentic: our proto¬ type. The IMP and IM.-P.M. issues of Mettius and Buca do not reproduce Caesar’s physiognomy, but only their monetary prototype. But they still have a touch of the contemporary general portrait-style and may have influenced the Caesar- portrait in the minor arts.

Ill

After the chronological issues are stabilized, we can return to the question of the Hague denarius Plate 1,8. This has behind the neck of dictator quartum a ribbon hanging on a nail which I identified with the diadem offered to Caesar, but declined by him and dedicated to the “Best and Greatest divine King of the Romans,” Jupiter. I have already taken a position27 against the objection that the hanging ribbon was recut from the titans. My argumentation did not convince Kraay who tried with new observations to support the interpretation of this symbol as a lituus.

He admits28 that the coin itself is undoubtedly genuine and the “diadem” which stands out sharply from the field is not a product of tooling after striking. He does not exclude completely the possibility that the diadem is due to the deliberate alteration of the die on or after the 15th of February.29 He also does not conceal the fact I stressed that behind the heads of the DICT. QVART coins (Plate 1,1-7) the lituus is curved in the other way and lends no opportunity to be reshaped. But he relies on the second Mettius issue (here Plate 1,9 sq.) on which the curved head of that ritual implement is without exception turned to the right, and he thinks that the workshops producing the types in question were sufficiently near in both time and space to influence each other. But we know now, that it was the one and the same establishment which made both series, and that the type with lituus curving clockwise was not yet in existence when DICT. QVART. was in¬ scribed on the silver coins. Thus the influence of the later tvpe could not have

27 Studien, I, p. 6.

28 Kraay, p. 20.

29 Ibid., p. 24.

40

ANDREW ALFOLDI

been exerted; this would be a films ante patrem. Even if we would forget the accurate planning and the rigid control of the slaves and freedmen who cut the dies, and would admit the irregularity exempli gratia, the diameter of the curve of the ribbon is decidedly smaller than that of the litui, Plate 1,1-7, and I do not see any trace of narrowing by later chiselling.

Kraay believes that his “lituus” (i.e., the right streamer of the ribbon) was made by one die-engraver and the straight line which became the other streamer was cut by another: since he assumes, as I do, that all these operations must have been made before the striking of the coin, this would make no difference, for the meaning and significance of the ribbon. If he once stresses the “deepcut straight line”30 and then considers again the possibility of chance damage,31 one of the arguments must be dropped - they mutually exclude each other.

The corrosion of outworn dies can, indeed, occasionally cause a die flaw, as Carson too thinks.32 The reverse of our Plate V,20 is a perfect example of this, with ramified fissures, the opposite of the violently cut straight-forward line of the left streamer of our ribbon. Such deep cuts could be originated by cracks, too: but they never stop just before the rim as in our case.33 The supposed accident must certainly be abandoned as the explanation for a reason my opponents have neg¬ lected. I mentioned34 the row of small dots visible on both streamers as well as on the loop:35 they have been driven in the obverse to mark the outline of the symbol to enable the die-cutter to carve it out properly afterwards. No corrosion, no die-flaw could engender this, only the intentional intervention of the die-cutter -an impro¬ visation on or after the 15th of February under the impact of the events known to the general reader from Shakespeare. The hanging ribbon remains the key to the timetable of the emissions. As a written document36 gives the title dictator quartum to Caesar on the 9th February, the leap to the Lupercalia a week later is not very large. Also the interlude of the imperator pontifex maximus titles sup¬ planting the mention of the abominated dictatorship remains valid. They appear on comprehensively planned reverse and obverse types which were presumably not intended originally for them: perhaps their emission was postponed until the

30 Ibid., p. 20.

31 Ibid., pp. 2i, 24.

32 Gnomon, 28 (1956), p. 182.

33 Cf. also Studien, I, p. 83, fig. 1 and pi. II, 5-6.

34 Studien, I, p. 7.

35 Ibid., fig. 1 and pi. II, 5-6.

36 Cf. Studien, I, p. 17.

PORTRAIT OF CAESAR ON DENARII OF 44 B.C.

41

Lupercalia, until it was known whether Caesar would become a rex or not. Owing to the perplexity of the political situation created by the emotional rousing of the people on the 15th of February, the timetable of the Venus-reverses of the quattuorvirate was upset. Originally, Mettius A-E (10b) was to be struck along with G-E (10 a). And some attribute corresponding to the lituus and the ladle of the latter was certainly foreseen - just as in the next planned complementary series of Buca and Macer the star and crescent should be put out together, as com¬ plementary symbols of Hellenistic royal eternity. But the attribute on the obverse of the cancelled Mettius issue (10b) was omitted when it later came to be issued. I think it may have had some connection with royalty.

The title dictator perpetuo is used only after the imperator series. It is not an honor granted then, but a new way to use an older one. We know that Caesar was already dictator perpetuo designatus for some time, but he still continued to count his power as a chain of one-year offices: just as Augustus did not use the tribunicia potcstas as a life-long mandate (i.e., in perpetuo), but only cautiously as a one-year privilege, renewed legally. Caesar lost his patience around the first of March and stressed the over-all nature of his power in the opposite way.

Around the time of the death of Caesar, apparently his veiled head appeared, the suppressed part of the Mettius issue (10 b) was brought out. Instead of IMP and an attiibute, the title is prolongated to IMPER to cover the empty space. Concerning the chronology of the subsequent issues, I hope I may confidently refer to my Studien.

binally, a word on the method. A. Momigliano,37 whose sharp mind and scholar¬ ly erudition I admire, tried to invalidate my interpretation of numismatic facts in quoting Kraay and Carson38 and in stressing that the divergences of opinion of the experts are diametral and reduce each other to nil.39 I do not believe that the obligation of the historians to exploit coin evidence can be dismissed by stressing the differences between specialists. “Mag der kritisch geschulte Lehrer der Geschichte, my predecessor in the chair of Ancient History at the University of Basel10 wrote long ago, auch nicht auf alien Gebieten selber produktiv mit- arbeiten, so kann man doch von ihm verlangen, daD er die Methode, mit der andere auf ihrem Gebiet arbeiten, zu beurteilen imstande sei.”

37 Athenaeum, 34 (1956), pp. 2 37 sq.

38 Op. cit., p. 241, n. 2.

Loc. cit. . le probability di accordo tra compctenti si riducono in pratica alio zero.” By the way: if the specialists do agree, must the historian accept their concord as obligatory proof?

40 Felix Staehelin, quoted by M. Gelzer in Gnomon, 1956, p. 624.

42

ANDREW ALFOLDI

Key to the Plates

Reverse-types :

(The group-numbers refer to the table of types on pp. 43-44)

1. Plate II, 17, 20-21, 22(?)

2. Plates II, 18-19, 23-24; III, 1-10; IV, 21, 24

3. Plates III, 11, 16; IV, 13, 16, 18-19, 23; V 17; VI, 3-6, 8, 10, 18

4. Plates II, 14; IV, 3, 5, 7, 22; V, 4; VI, 9, n-17, 19-20

5. Plates IV, 4; VI, 23

6. Plates III, 22; V, 3, 9-15

7. Plates II, 2, 4-9, n-13

8. Plates IV, 2, 6, 8, 12, 15, 18-19, 21-23; VI, 1

9. Plates IV, 9-1 1, 17; V, 1-2, 8; VI, 2

10a. Plates I, 9(G), 10(G), n(G), 12(H), 13(H), 14(H), 15(H), 16(H), 17(1), i8(I), 19(1), 2o(Iv), 21 (K), 22(E), 23 (L), 24 (L) ; II, io(K) ; V, 5(1)

10b. Plates III, 12(A), 13(B), i4(C? altered into O?), 15(C), 17(B), i8(B altered into C), 19(C), 20(D), 21(D), 23(D), 24(E); IV, 14(B), 20 (B altered into C) ; V, 6(B), 7(B), 16(D), 20 (=Plate

III, 14); VI, 7(B)

11. Plates II, 15-16; V, 24; VI, 21-22

Not included in the table are: (a) M. Mettius, Sydenham 1057: Plates I, 1-8; II, 1, 3;

IV, 1; (b) C. Cossutius Maridianus, Sydenham 1069: Plate VI, 24.

Provenance of the coins illustrated:

E. J. Haeberlin, Auct.-catal. : Plates IV, 16; VI, 18 Hague, R. Cabinet of Coins: Plates I, 8; II, 18; V, 21; VI, 23

Italy, Private Coll.: Plates I, 1; II, 15, 21-23; III, 9, 11, 14; IV, 3, 5, 6, 17-20; V, 3, 17. 20, 22-24; VI, 14, 16

London, British Museum: Plates IV, 4, 9, 12, 14; V, 4 Miinchen, Staatl. Munzsammlung: Plates I, 12; II, 20 Miinzen und Medaillen A.G., Basel: Plates II, 9, 19; III, 18

New York, Am. Num. Soc.: Plates I, 2, 4, 14-15; H, 2, 4, 5, 16; III, 6-7, 15; IV, 1-2, 7, 21, 24;

V, 2, 11, 15, 18; VI, 4, 13, 19-20

W. Niggeler, Baden bei Zurich: Plates I, 5, 16; II, 1, 13-14; III, 1, 4-5. 8, I3> l(h 24> IV, 22-23; V, 1, 9, 10, 13

Paris, Cabinet des Medailles: Plates I, 6, 11, 13, 17, 20, 24; II, 3; III, 12, 19; V, 5-6 Princeton Univ. Coll.: Plates I, 10; VI, 12 Roma, Museo Nazionale: Plate II, 7

Vatican, Medagliere: Plates I, 7, 19; II, 6, 24; III, 3, 22-23; V, 8, 16, 19; VI, 6-9, 11, 22, 24 Vienna, Bundessammlung von Medaillen und Miinzen: Plates. I, 3, 9, 22, 23; II, 8, 11; III, 10, 17, 21; IV, 8, 10— 1 1 , 15; V, 7, 12, 14; VI, 3, 5, 15 Dr. A. Voirol, Basel: Plates II, 12; III, 2; VI, 21 Yale Univ. Coll.: Plate VI, 17

Ap. Zeno (Auct., Vienna, 1956.pt. 2): Plates I, 18, 21; II, 10, 17; III, 20; VI, 1, 2, 10

Plate

i I

DENARII OE 4

IHE PORT RAI1 01' CAESAR ON III!

Plate 11

III! PORTRAIT OF CAESAR ON THE DENARII OF 44 B. C.

THE PORTRAIT 01* CAESAR ON THE DENARII 01 44 B <

Plate IN’

I HE PORT RAI'J OF CAESAR ON THE DENARII OF 44 B. C.

Plate V

THE PORTRAIT OF CAESAR ON THE DENARII OF 44 B.C

I

Plate VI

niE PORTRAIT OF CAESAR ON THE DENARII OK 44 B. C.

PORTRAIT OF CAESAR ON DENARII OF 44 B.C.

43

CAESAJU/*^

Caesar 5MP ^

with wreath only

<SM3AR/bICT-

a) with wreath only

b) veiled head

CAESAK{.kP

with wreath only

CAESAK/4kT

PIMUPKTV®

a) with wreath only

b) veiled head

SAB. /BIST

f’SRPETVO

a) with wreath only

b) veiled head

CAESA^-Bicy/

with wreath only

The distribution of the Venus-types

44

ANDREW ALFOLDI

7

CMSA&’IM.

IP

/CAESAR/WCT

®I®JPET

with wreath only

with wreath only

<5AESAB»/BieT

SERtETv©

SAESAR.&ICT/

SES.I’ETv©

with wreath only

with wreath only

10 b

CAisAhMxnf,

with wreath only

1 1

CAESAR./ MCT •

S’IRFEW©

veiled head

The distribution of the Venus-types

A KNOTTY PROBLEM CONNECTED WITH THE

COINAGE OF AZES

A. S. Altekar

a

Xxmong t lie Scythian and Parthian invaders of the Punjab who ruled in India during the ist century B.C. and A. I)., Azes occupies an important place. Sir John Marshall identified him with King Aya mentioned in the Taxila Silver Scroll inscription of the year 136, and postulated the theory, that he started the Vikrama era of 56 B.C.1 Though this view is not now usually accepted, there cannot be any doubt that Azes was a powerful ruler whose sway extended over the Western Punjab and some portion of Afghanistan. Our information about Azes is mostly based upon his coins, which were issued in large numbers and in as many as 39 different types.

Azes was followed by another king named Azilises, who is usually taken to be his son. Azilises is known to us only from his coinage, issued in 22 different types. These two rulers usually borrow the main types of their predecessors, the Indo- Bactrian kings. Ihey also follow them in giving the legend on the obverse in Greek script and language and the legend on the reverse in Kharoshtri script

and I lakrit language, then well established among the Indian population of the Punjab.

On most of the types of Azes, the obverse legend is BAIIAEQI BAIIAEflN METAAOY AZOY and the Kharoshtri legend is Maharajasa rajarajasa mahatasa 1 Cambridge History of India, I (Cambridge, 1922), p. 571.

45

46

A. S. ALTEKAR

Ayasa. The Prakrit legend in Kharoshtrl script is a translation of the Greek legend, both meaning, “Of the great king, king of kings, mighty Azes.’’ Azes in Greek is Indianised into Aya.

On most of the coin types of Azilises, the obverse legend in Greek is BAZIAE^I BAZIAEQN METAAOY AZIAIZOY and that on the reverse in Kharoshtrl script is Maharajasa rajarajasa mahatasa Ayilishasa. The meaning of this legend is, Of the great king, king of kings, mighty Azilises.’ Here the name Azilises is Indianised into Ayilisha.

There are two coin types, however, which are extremely rare, on which the legends are:

A.

Obv. BAZIAEQZ BAZIAEftN MErAAOY AZOY

Rev. Maharajasa rajarajasa mahatasa Ayilishasa

B.

Obv. BAZIAEQX BAZIAEfiN MErAAOY AZIAIZOY

Rev. Maharajasa rajarajasa mahatasa Ayasa.

Specimens of these coins are described in detail and illustiated at the end of this paper.

These two types with their characteristically divergent legends have given rise to a controversy among scholars. Joint rule was a peculiar charat eristic of the Indo-Scythian and Indo-Parthian rulers. Gondopharnes issued coins jointly with his nephew Abdagases. In the dynasty of the Saka Kshatrapas of Western India we have several instances of the king issuing coins with the title Maha- k shat rap a and his heir apparent simultaneously striking coins with the lower title Kshatrapa. In the case of the joint coinage of the Parthian ruler Vonones also, while he is given the imperial titles on the obverse in the Greek legend, his brother Spalahores or nephew Spalagadama bear no regal or imperial titles but are simply described as the brother or the nephew of the emperor.

It is usually assumed that the legend groups A and B described above demon¬ strate either that there were two rulers bearing the name Azes or that theie weie two with the name Azilises. Rapson,2 Smith,3 Raychoudhari4 and others hold

2 Cambridge History of India, I, p. 572-

3 Vincent A. Smith, Catalogue of the Coins in the Indian Museum, Calcutta, I (Oxford, 190G),

pp. 45-50.

4 H. C. Raychoudhari, Political History of Ancient India, 4th ed., pp. 365-67-

COINAGE OF AZES

47

that Azes came first; during his old age he was ruling jointly with his son and heir apparent Azilises, who associated with the emperor also on the coinage, being given the place on the reverse. When in the course of time Azes I died, Azilises started issuing coins alone. But he also had a son, named after the grand¬ father Azes, who later became Azes II. During his old age Azilises associated his crown prince Azes with himself in the coinage and issued coins bearing his own name in Greek script on the obverse as AZOY and that of his crown prince Azilises as Ayilishasa in Kharoshtrl on the reverse. This situation resulted in legend group B. In the course of time Azilises died and Azes II succeeded him, issuing coins with his own name on both obverse and reverse. This school of scholars maintains that the coins of Azes II are at present confused with those of Azes I. Of the 39 types bearing the name of Azes, some are of Azes I and some are of Azes II. It is difficult to believe that a king could have issued coins of 39 differ- ent types, however long might have been his reign or extensive his kingdom. Smith thinks that the artistically inferior types of Azes should be attributed to Azes II, and he has separated them in the Catalogue of the Coins in the Indian Museum Calcutta, I, pp. 47-50.

Dr. Whitehead, however, disagrees. In editing the catalogue of the coins in the Panjab Museum,6 he observes that the coins of Azilises are on the whole artistically superior to those of Azes. Art deterioration would thus indicate that Azilises preceded Azes and issued the usual type with the legends giving his name as AZIAIZOY on the obverse and as Ayilishasa on the reverse. During his old age he associated his son Azes with himself, and during this transition period coins with the legend group B were issued, giving the name of Azilises in Greek on the obverse and that of his son Azes as Aya in Kharoshtrl on the reverse. When Azilises I died, Azes ascended the throne and issued the 39 types of coins referred to earlier. Azes, however, had a son, named Azilises after his grand-father, who was associated with the government as crown prince during the old age of Azes. This period is represented by legend group A, where the name of Azilises appears on the obverse in Greek script, that of Azes as Aya on the reverse in Kharoshtrl script. Azes died in due course, and Azilises ascended the throne as Azilises II, issuing coins bearing his own name on the obverse in Greek as Azilises and on the reverse in Kharoshtrl as Ayalisha. According to Dr. Whitehead all the 39 types of Azes belong to one single king, while of the 22 types of Azilises some belong to Azilises 1 and some to Azilises II.

0 Catalogue of Coins in the Punjab Museum, Lahore, I (Oxford, IQ14), p. 93.

48

A. S. ALTER AR

For various reasons this view of Whitehead has not found general acceptance:

1. The argument of art deterioration is highly subjective and it cuts both ways. If we accept this art deterioration theory and the view that there were two kings of the nameAzilises separated byAzes, then, there would be some coins of Azes artistically superior to those of Azilises (who is Azilises II of Whitehead), and some coins of Azes also inferior to those of Azilises (who would be Azilises I of Whitehead). We can as well argue that those coins of Azes which are artistically superior to those of Azilises even according to Whitehead belong to Azes I, and those of Azes which are artistically inferior to those of Azilises belong to Azes II.

2. The second objection to Whitehead’s theory is the fact that there are 39 types of Azes and only 22 of Azilises. If we have to assume two kings of the same name, it is more reasonable to postulate two kings named Azes, rather than two named Azilises.

3. A third argument is also adduced in favor of the view that these were two kings with the name of Azes. The dynasty of Azes was succeeded by that of Gondaphar- nes, so well known from the legend of St. Thomas. The coinage of Gondopharnes shows that he was associated in the government with a strategos or general named Aspavarman, whose name appears on the reverse in the Kharoshtrl legend : Indravarmaputrasa Aspavarmasa strategasa jayatasa tratarasa. We, however, find this general Aspavarman associated with Azes in his coinage, where also he appears on the reverse with above coin legend, only with the last word tratarasa omitted0. If we assume that there were two Azilises and only one Azes, we shall have to assume that the career of Aspavarman extended over part of the reign of Azes, the whole reign of Azilises II and part of the reign of Gondopharnes. this is very improbable, if not impossible. We also cannot easily explain why we should not have some coins where Aspavarman is associated with Azilises II on his coinage. If, on the other hand, we hold that there were two kings with the name of Azes, we can assume that the king Azes associated with Aspavarman is Azes II. When his dynasty was overthrown by Gondopharnes, Aspavarman transferred his allegiance to the new conqueror and was permitted to be asso¬ ciated with him on his coinage.

The view now generally accepted by scholars is that Azes I was succeeded by Azilises, who in his turn was followed by Azes II, Aspavarman being associated with Azes II and his conqueror, Gondopharnes. Among semi-barbarous tribes subordinates transfer their allegiance to a conqueror without much compunction. 6 R. B. Whitehead, op. cit., I, p. 150; p. 130.

COINAGE OF AZES

49

This view, though now almost universally accepted, is not without its own drawbacks. Proper weight has not been attached to the very important circum¬ stance that in the case of the joint types both Azes and Azilises bear titles of identical imperial significance, BAZIAEQ2 BA2IAEQN METAAOY on the obverse, and Maharajasa rajarajasa mahatasa on the reverse. If one was the crown prince or viceroy or general of the other, then his titles should have been those of a subordinate ruler or officer and not identical with those of the emperor. Azes is associated with his general, ASpavarman, on his coinage; the obverse legend gives the former the imperial title BAZIAEQZ BAIIAEQN while the reverse legend gives the latter the simple title 'jayatasa or victorious. Vonones is associated with his brother Spalahores and nephew Spalagadames; the Greek legend gives him the imperial titles BASIAEQS BA2IAEQN MErAAOY while the Kharoshtrl legend simply describes the subordinate either as maharajabhrata, the brother of the emperor, or as Spalahoraputra, son of Spalahora. The usual practice in the Punjab at this time no doubt favoured joint coinage, but it was particular in giving distinctly inferior titles to the crown prince or viceroy associated with the emperor in the administration. Why then should the Azes group of kings depart from this custom and permit the crown prince or the viceroy to have identical imperial titles? That the crown prince or the viceroy should have imperial titles, identical with those of the emperor, is almost unthinkable.

How then are we to explain this riddle of the so-called joint coinage of Azes and Azilises ? I would like to suggest a new way out of the difficulty.

We must not forget in this connection that the so-called joint coinage is repre¬ sented by extremely rare coins; there are only two coins having the name of Azes on the obverse and Ayilisha i. e Azilises on the reverse, and only one having the names of Azilises on the obverse and Aya i. e Azes on the reverse7. This fact has its own significance when wre remember that we have scores of specimens of most of the other types of Azes and Azilises. The joint rule must have extended over at least a few months, if not over a few years, and we should have fairly numerous specimens of both the so-called joint types. The joint type must have been intend¬ ed to proclaim to the subjects the name of the crown prince, and the best way to do so was to issue it in abundant numbers. This, however, was not done. The joint types are, as observed above, extremely rare.

1 he best and perhaps the only way to explain these extremely rare types is to assume that they were the result of a mistake in the mint. At the time of the death 7 R. 13. Whitehead, op. cit., I, p. 132.

4 Centennial

50

A. S. ALTER AR

of Azes I the mint must have possessed many dies of his 39 different types. They were probably not all of them destroyed immediately after his reign but were allowed to remain in some of the drawers or cabinets of the mint. When Azilises ascended the throne, new dies were cut for him, both for the obverse and reverse. The dies of the old reign of Azes were perhaps lying in lower drawers and those of Azilises in upper drawers. It is not impossible to assume that a mistake might have been committed by a mint official in handing over to an artisan the obverse die of Azes bearing the Greek legend BAZIAEftI BAZIAEQN MEFAAOY AZOY along with a reverse die of the new king Azilises bearing the Kharoshtrl legend Mahay ajasa rajarajasa mahatasa Ayilishasa. The muling of these two dies would result in legends of group A, giving rise to the false impression that Azes was ruling jointly with his son Azilises. Similarly, another mistake of the same nature might have been responsible for the combination of the obverse die of the new reign of Azilises with the reverse die of the old reign of Azes, resulting in the legends of group B and rendering the theory plausible that Azilises ruled for some time as emperor along with his son Azes as his crown prince.

These mistakes, however, would have been detected soon and the output of coinage with these wrong combinations of dies discontinued. This may be the reason why the so-called joint types are so far known in extremely small numbers.

If we accept this theory of a mint error, we can well understand why there is no difference in the status of Azes and Azilises in the legends occurring on their so-called joint coinage. This circumstance can hardly be explained by any other hypothesis.

The joint coinage of Aspavarman with Azes and Gondopharnes need not compel us to assume an Azes II. Aspavarman might have become persona non grata with Azilises; he might have been expelled from the administration and so could not figure on the coinage of Azilises. Being disgraced by Azilises, Aspavar¬ man might have bided his time and joined Gondopharnes when he attacked Azilises and conquered his dominion. He might have been rewarded for his service by the new conqueror by being reappointed a strategos and by being permitted to become associated with the emperor on his coinage in a subordinate capacity.

To conclude, we have insufficient evidence to warrant the theories that there were two I ndo- Scythian kings named Azes or Azilises. 1 he so-called joint coinage is so far known from two specimens of one variety and only one specimen of the other. Since they give identical imperial titles to the emperor on the obverse and to the subordinate crown-prince or viceroy on the reverse, it is best to assume

COINAGE OF AZES

51

that these rare coins were no real independent types, but owe their existence to a mistake in the mint, resulting in the use of the obverse or reverse of the previous reign of Azes with the reverse or obverse of the subsequent reign of Azilises.

We give below a description of the coins under discussion. The figures 1 and 2 are reproduced from NC, 1890, PI. X,i-2.

Fig. 1

Fig. 2

Type A: (Fig. 1)

Azes on the obverse and Azilises on the reverse.

Silver, circular, 1.1" diameter, 140 grains; stated by Cunningham to be in “the French collection.”

Obv. King on the horseback to r., carrying a spear; circular Greek legend: BAIIAEQZ BAIIAEQN MErAAOY AZOY.

Rev. A deity standing to 1., right hand extended, left hand carrying palm and banner. Legend: Makar ajasa rajarajasa mahatasa Ayilisliasa.

1 he illustration of Cunningham is obviously from a drawing. The second coin of this type is in the Punjab Museum, Lahore, no. 319.

Type B: (Fig. 2)

Azilises on the obverse and Azes on the reverse.

Silver, circular, 1" diameter, 140 grains; Cunningham’s collection.

Obv. As above; circular legend: BAZIAEQZ BAIIAEQN MErAAOY AZIAIIOY.

Rev. Pallas armed to 1., shield on left arm, thunderbolt in upraised right hand, ready to hurl. Kharoshtri legend: Maharajasa rajarajasa mahatasa Ayilisliasa .

LE ROI PYRRHOS

Jean Babelon

jLa personnalite de Pyrrhos a ete magnifiee jusqu’a devcnir legendaire, par les historiographes latins jaloux d’eriger en puissante effigie l’image d’un heros qui avait fait trembler Rome, mais que Rome avait vaincu. Et Plutarque cn a 1 aconte la vie exemplaire. II appartenait a la critique moderne d’examiner de plus pres, dans le concret, ce que tut ce chef d’armees, auteur de victoires eclatantes, mais fallacieuses, de determiner sa situation au debut de l epoque hellenistique, au moment oil s eveillaient les convoitises des diadoques, et dans une incessante emulation, leur volonte de puissance, leur virtu, a l’italienne. C’est dans des con¬ ditions assez bien definies, qui determinent une attitude politique et une ethique que Pyrrhos prend le titre de basileus en se tournant vers l’Occident ouvert a son ambition par ses rivaux qui luttent en Orient.1 Or, Pyrrhos est l’heritier dune

1 Alexandre prend le titre de basileus apres 330, d’abord a Alexandria ad Issum, puis a Citium, k Arados, a Amathous, a Side, a Tarse un an seulement avant sa mort, a Babylone. A Amphipolis le titre royal apparait pour la premiere fois sur une monnaie frappee en Macedoine. A Pella, les monnaies portent seulement le nom d’Alexandre.

L'etymologie du mot PoktiXeus demeure incertaine. Quant a sa portee, d’apres Aristote ( Polit . Ill, 12H5, E 9) h confere a celui qui en est revetu une autorite qui s’cxerce en mati6re de guerre, de religion et de justice. Cette autorite est legale, repose sur la faveur de Zeus, et peut etre hereditaire. Le terme apparait des l’epoque homerique. II s’applique aussi a des rois barbares, en Thrace, en Illyrie, en 1 hessalie, comme en Epire. Voy. l’articlc de Schoeffer, dans RE.

A paaiXtOs s’oppose dvAacxov, suivi d’un genitif ( Od . IV, 602 - II. X, 33; Pindare, Olytvp. VI, 34 XIII, 24, bur. Andromaque, 22) voy. les observations de F. Chamoux, Fouilles de Delphes. L’Aurige, I955> P- 2&- Tupawos designe le maitre absolu, l’usurpateur du pouvoir abusif.

53

54

JEAN BABELON

autorite qui lui a ete transmise par une dynastie. Cette tradition amalgamee a la conception monarchique suscitee par Alexandre le grand, compose en sa personne un type de souverain dont nous pouvons dessiner le caractere dans le moment historique ou se produit son ephemere epiphanie.2

La numismatique, les emissions monetaires, nous instruisent autant que les textes des historiens ou les inscriptions, sur le destin du roi, sur la structure politique ou sociale, qui sent les conditions de son avenement. Dans les limites etroites du present article, nous concentrerons notre attention sur les tetradrach- mes qui ont pour types Zeus dodoneen et Dione, et pour legende BAEIAEftS nyppoY.

Mais tout d’abord il importe de considerer le milieu d’ou jaillit l’impetueuse activite de celui que les Romains appelaient l’Epirote.

Dans le Nord de la Grece, occupe par des populations montagnardes, par des races dont l’energie singuliere devait se manifester au cours des temps par Taction non seulement d’hommes mais de femmes restes celebres dans Thistoire, la nature meme du pays, son morcellement, et jusqu’au climat rudement con¬ trasts, Tindividualisme forcene qui en est la resultante, devaient entretenir un etat durable de demi-anarchie. Clans ou bandes, Molosses, Thesprotes, Athamanes, Chaones, Atintanes, Paraueens, Orestes - la liste abondante en est donnee par Hellanikos - facilement excites au pillage et aux raids profitables, mais attaches au terroir natal par un amour qui ne s’eleve pas jusqu’a la notion de patriotisme,

On note aussi que la basileia, a l’orientale, n’a pas ete adoptee sans nuances par les diadoques, les uns, comme les Seleucides s’etant considered comme les successeurs du Grand Roi, les autres, a Pergame notamment, ayant edulcore la manifestation du pouvoir supreme, en presence des peuples helleniques qu’ils avaient a gouverner.

Quant au terme de Sectttottis il n’a pas de portee proprement politique, et se refere souvent au maitre des dieux. Awdorns au contraire, contient l’idee de souverainete, mais le dynaste est con¬ siders comme un prince de peu d’importance, un principicule ou un roitelet, parfois subordonne au Grand Roi. Auvotcrrai, pacn^eis, e’est ainsi, par exemple, que les auteurs grecs designent, au milieu du IIP siecle, les chefs libyensque les Romains appellent reguli, au moment de la guerre des mercenaircs contre Carthage.

Polybe (VII, 4, 5) qualifie Pyrrhos d’ hyepebu Kal pacjiAeus. A. Aymard, “Le protocole royal grec et son evolution”, REA, 1948, pp. 232, 268. Le tetradrachme publie par A. Amoros, Arch, espahol de arqueologia, 1950, nous presente avec la legende BA2IAE03I FIYPPOY, le portrait du roi non pas diademe, mais couronne de chene comme le Zeus de Dodone 2 Martin P. Nilsson, Studien zur Geschichte des alien Epciros, Lunds Universitets Arsskrift, N.F. 1, VI, 4, 1909. Idem, Cidts, Myths, Oracles and Politics in Ancient Greece (Lund, 1951) : The Epirote Royal House. Epeiros, dans RE. Guiseppe Nenci, Pino, aspirazioni egemoniche ed equilibria mediterranco (Torino, 1953), p. 71 et ss., pp. 84, 85. Cross, Epirus (Cambridge, 1932).

LE ROI PYRRHOS

55

Macedoniens ou Epirotes “les habitants de la terre ferme”, devaient suivre leurs rois avec une fidelite fondee sur 1’esprit d’aventure comme sur la cupidite, mais avec la fierte arrogante de guerriers qui ne voient dans leur chef qu’un egal a la merci de leur consentement.3

Entre toutes ces peuplades, les Molosses apparaissent comme la plus fortement organisee. On discerne chez eux les fondements d’un pouvoir que Ton peut deja qualifier de royal, mais limite par les droits que revendique le peuple en armes. Ce pouvoir est done constitutionnel, si Ton consent a appliquer ce terme ana- chronique a des populations peu evoluees, mais foncierement democratiques. De meme qu’en Macedoine, oil pourtant la descendance de la famille royale ne souffre aucune contestation - elle est d’origine divine, et la dynastie est heracleenne -, 1’ autorite du roi n’est pas reconnue sans examen chez les Molosses. A sa mort, e’est l’armee qui procede par acclamation au choix du successeur. Celui-ci n’est pas obligatoirement le fils du defunt. Toutefois l’espoir d’une transmission familiale reste vivace: citons le cas d’Alketas I, qui partage le pouvoir avec son fils, le futur Neoptoleme I. Mais le nouveau chef est normalement l’objet d’une designation deliberee. S’il est trop jeune pour exercer le gouvernement et conduire les troupes a la guerre, on place a ses cotes un regent. C’est le role joue par la fameuse et cruelle Olympias, aupres de Neoptoleme. Si le roi se montre inferieur a sa tache, il est toujours loisible aux hommes en armes dele deposer, et Ton n’oserait appli¬ quer le nom de mutinerie a l’explosion de la volonte plurale qui entraine la decheance de l’incapable, et d’aventure sa mise a mort.

Ajoutons que le roi en exercice peut adjoindre a son autorite un representant de la famille adverse qui la lui dispute. En etudiant revolution de la monarchic en Epire, sous les rois qui se succedent: Tharyps, Arybbas, Alketas, Aeacides, Nilsson etablit une comparaison avec la dyarchie de regie a Lacedemone,3 a quoi Ton peut opposer que la double magistrature, immuable comme le couple des Dioscures, est introduite officiellement dans la legislation de Lycurgue, alors qu’elle n’est chez les Molosses qu’un remede d’occasion. En outre, on constate aux cotes du roi la presence d’un prostates, ministre ou delegue du peuple, dont le role important est atteste par son nom inscrit sur les documents officiels immediate- ment apres celui du basileus.4

L’ascendant d’une personnalite douee d’une vigucur exceptionnelle, telle que colie de Philippe de Macedoine, le genie d’Alexandre et l’eclat fulgurant de ses

3 S. Accame, “La diarchia dei Molossi”, Riv. filol. class., 1934, pp. 52 2 ss.

4 Voir les inscriptions commentees par Nilsson, Sammlung der griechischen Dialeklinschriften.

56

JEAN BABELON

conquetes devaient modifier cette conception de la royaute, dont l’absolutisme s’inspire desormais de l’exemple offert par la monarchic iranienne subjuguee. Toutefois, si Ton peut dater de cette epoque la naissance d’une nation macedo- nienne, on a pu dire que sous Antigone Doson encore, le nom de roi etait un titre vide. Neanmoins une evolution sociale se produisait, dont nul ne mesurait la portee, la monarchic absolue se revelait peu a pen comme le seul principe capable de cristalliser d’abord chacune des parties de l’univers, puis de les rassembler en un immense organisme. Cette gestation fut longue et penible, elle se heurtait a des institutions confuses, a des traditions inveterees, plus fortes encore que des lois, et merne a une philosophic propagee par les penseurs, dont les principes s’etaient repandus chez les “philhellenes”.

Sans nous attarder dans ces anticipations, nous constatons en Epire, plus encore qu’en Macedoine, et jusqu’a une epoque plus tardive, celle oil nous nous plagons, au moment oil Pyrrhos accede au pouvoir, un conflit entre ce que nous appellerions, en des termes trop systematiques, le federalisme et la monarchic.

Avant le debut du IIIe siecle, trois rameaux, les Molosses, les Chaones et les Thesprotes composent, en fait sinon en droit, une alliance federale, sous L autorite du roi des Molosses, autorite sans cesse chancelante, et qui commit des eclipses comme des retours de lumiere. La encore, les qualites personnelles d’un jeune chef devaient lui permettre d’imposer son autocratie, dans le moment oil les circon- stances politiques etaient propices a l’essor de son ambition.5

Le nom meme d’Epire, avec toute sa signification, non seulement son acception geographique, mais son unite sociale et politique, ne se manifeste pas avant le milieu du IVe siecle. La symmachie des Epirotes, AFTEIPftTAN, est constitute des 330, apres la mort d’Alexandre le Molosse, et des lors la frappe de monnaies de bronze en leur nom en atteste l’existence reelle. Dans un ouvrage recent, le Dr. P. R. Franke6 a accumule des arguments qui lui permettent d’affirmer qu’il

5 Julius Beloch, Griechische Geschichte, 1893-1897. Fr. Klotzsch, Epirotische Geschichte bis zmn Jahr 280 (Berlin, 1910). Niese, Geschichte der griechischen und makedonischen Staaten seit der Schlacht bei Chaeronea. M. Holleaux, Rome, la Grece et les monarchies hellenistiques (Paris, 1921). Peter Robert Franke, All-Epirus und das Konigtum der Molosser, 1955.

6 P. R. Franke, “Geschichte, Politik und Miinzpragung im friihen Makedonien”, JNG, 1952-53, pp. 99 ss. Doris Raymond, Macedonian Regal Coinage, NNM, No. 126 (New York, 1953). Konrad Kraft, “Der goldene Ivranz Caesars”, JNG, 1952-53, p. 26 (sur Zeus couronne dechene). Rappelons que de 513 a 480 la Macedoine est une satrapie perse. D’apres Herodote le premier “roi” est Perdikkas, vers 700. L. Breglia, “Nuovi elementi per la circolazione monetaria e la storia dell’ Epiro”, R. C. Accad. di Napoli, 1941, pp. 145, 260.

LE ROI PYRRHOS

57

ne peut etre question jusqu a cette date, dune confederation des Epirotes sous r autorite d’un roi d’Epire. C’est a titre d’heritier des rois des Molosses que Pyrrhos aurait acquis l’ascendant qui lui permit d’entreprendre son expedition en Italie puis en Sicile. II convient done d’examiner la portee du titre de basileus dont il fit preceder son nom. Ouelle fut en son essence l’institution de la royaute en Epire,

pai quels liens se rattache-t-elle aux monarchies hellenistiques, e'est la question que posent les monnaies que nous avons sous les yeux.

L’evenement de 306, la victoire fameuse d’Antigone et de Demetrios Poliorcete sur Ptolemee, a Salamine de Chypre, ouvrit la voie aux ambitions personnclles des diadoques, que Ton voit tour a tour se proclamer basileis. Par emulation, le roi d’Epire se voyait autorise a se parer du meme titre, tandis que ses emules en competition le poussaient habilement vers 1’Occident et l’incitaient a s’y tailler un domainc egal a ceux qu 1 Is se disputaient en Orient. L’orgueil des Epirotes y trouvait son compte tout autant que leur cupidite. C’est ainsi que l’on peut se repi esenter le climat moral et politique qui regnait en Epire, quand Pyrrhos congut son grand dessein. Mais son ascension meme ne devait pas etre durable: il n’en subsista rien apres son echec. En235, les Epirotes se debarrassaient de ses derniers descendants, et revenaient a leur independance, a laquelle ils donnaient la forme dune republique federate.

Quelques textes sont a relever a cet egard. Sur la plaque de bronze portant une inscription dedicatoire consecutive a la victoire d’Heraclee, on lit ceci:7 BaaiAeus lluppos Kai ’AimpcoTou Kai TapavTlvoi caro ‘Pcopaicov Kai tcov auiiiadycov A 11 Nai'coi. Il n est done pas question ici, apparemment de “roi des Epirotes,” mais du basileus Pyrrhos, dans l’absolu, qui rend hommage au Zeus de Dodone avec les Epirotes et les Tarentins. Mais peut-etre ne faut-il pas pousser trop loin, comme le fait P. R. Franke, cette exegese d’un formulaire officiel, ou l’on trouverait la preuve d une sorte de dissociation entre le roi et le peuple qu’il avait entraine a la conquete, et dont le nom ne figurerait ici que pour attester une maniere d’alli- ance. Toujours est-il que le titre de basileus, tel qu’il est assume par Pyrrhos, comme pm Antigone, affiime une qualite eminente prise d’emblee par la volonte person- n( lie du 1 oi, sans qu apparaisse 1 opinion populaire dont elle suppose ou escompte 1’acquiescement.

7 SIG, n° 392. Cf. P.R. Franke, sur les restitutions et les corrections. Cf. l’inscription d’Antigone basileus et des Mac^doniens, et Hampl, Der Konig dcr Makedoner (Leipzig, 1934). II est k noter que sur les monnaies que nous aurons k examiner, le titre de BAIIAEYZ apparait seul, sans etre accompagne d un ethnique tel qu ’ATTEipcoToov.

JEAN BABELON

58

3

On pourrait ici evoquer, par contraste, la “royaute philosophique” d’ Antigone Gonatas, conseille par Menedeme d’Eretrie, et dont le maitre, Euphantos d’Olynthe, avait ecrit, a l’usage de son disciple, un traite sur Part de gouverner un royaume. Antigone se declarait l’eleve de Zenon, et le serviteur de son peuple. Aussi bien se refusait-il a reconnaitre le titre de basileus a Seleucos oil a Ptolemee, avant, il est vrai, de s’en parer lui-meme8.

Les legendes monetaires se presentent a l’appui de ces considerations. Des le VI° siecle, nous voyons un dynaste thrace prendre le titre de basileus, mais sous line forme particuliere; l~ETAE BAZIAES3E HAONAN, ou TETA2 HAONEON BA2IAEYX ou TETA BAEIAEY HAONEON : Getas, roi des Edoniens,9 et il est a supposer que la monarchic prenait cette forme democratique sous les roitelets dont nous parle Thucydide: Teres, Sitalces, plus tard Seuthes, l’adversaire redoute deLysimaque, dont on a des monnaies a son effigie, en 323-31 1. 10

Au contraire, sous les monarchies hellenistiques, le nom de basileus , sans evocation du peuple gouverne atteste un absolutisme tout personnel. On en obverse l’introduction progressive dans la titulature des diadoques.

Les monnaies de Ptolemee Soter se repartissent ainsi:

Premiere periode, 323-305, monnaies au nom des suzerains successifs, aux types d’Alexandre le grand. Vers 316, la tete d’Heracles se trouve remplacee par celle d’Alexandre coiffe de la peau d elephant, avec la legende TTTOAEMAIOY AAE1ANAPEION.

Une deuxieme periode s’etend de 305 a 285. Ptolemee prend le titre de basileus sur des stateres qui portent son effigie, sur des tetradrachmes, aux types de la tete d’Alexandre coiffe de la peau d’elephant, et d’ Athena Promachos, avec la legende AAE2ANAPOY - sur des monnaies de bronze au type d’Alexandre cornu et l’aigle, avec la legende PITOAEMAIOY BA2IAESi!2 - enfin sur des tetradrachmes a son effigie et l’aigle au revers, avec la legende BAIIAEQI fTTOAEMAlOY.11

Les monnaies de Philippe Arrhidee, puis de Cassandre ne sont pas moins elo- quentes: Philippe Arrhidee, de 323 a 316, frappe des tetradrachmes aux types alexandrins, avec BASIAEfiS OlAirTTJOY. Cassandre, de 3*6 ^ 29 7> ne iraPPe Pas

8 Tarn, Antigonus Gonatas, p. 21.

9 Svoronos, L’Hellenisme primitif de la Macedoine, 1919, p. 49-

10 Gerassimov, “Effigie de Seuthes III 323- 311 sur les monnaies’ , Bull, dela Soc. arch, bulgan, XIX (1955, en bulgare). Apres 323, quand Seuthes a ete proclame roi independant de la Thrace. Cf. Spokes qui prend le titre de Basileus a Abdere au IVe s.: HN\ p. 283. Remarquer que le titre de basileus est seul usite sur les inscriptions: SIG, nos. 369, 392, 393, 453.

11 Svoronos, Ta vopiapaTa toO KpdTous tcov rtToAspaicov, I, p. p(3 .

LE KOI PYRRHOS

59

d’abord de monnaies d’or ou d’argent a son nom, mais les bronzes portent la legende KAZIANAPOY, et plus tard BA2IAE QY KAI2ANAPOY.

Quels sont les ressorts que pouvait faire jouer Pyrrhos pour imposer a ses tumultueux compatriotes un titre qui heurtait de front leurs tendances egalitaires et l’arrogance de leurs traditions? La religion d’abord, la reference au dieu national, le Zeus de Dodone, et a sa paredre Dione. II est notable qu’apres lui, les Epirotes libercs de la royaute, continuerent a invoquer sur leurs monnaies les divinites dont il avait accapare naguere pour lui seul la protection.12

D’autre part, il ne parait point que le souvenir d’Alexandre ait garde en Epire la meme vigueur que dans les districts occupes par les Macedoniens. En fait, nous n’avons recueilli aucun temoignage sur la diffusion dans ces contrees des tetradrachmes aux types traditionnels d’Heracles et de Zeus olympien. Il est vraisemblable pourtant que les secours envoyes a Pyrrhos par Ptolemee Keraunos ou par Antigone, se soldaient par des sommes d’argent comptees en "alexandres.” Mais l’orgueil des Epirotes se fondait sur une descendance non moins illustre, et, a leurs yeux, plus prestigieuse encore, pour etre plus lointaine et plus mythique, celle d’Achille, le heros de la guerre deTroie. Tout concourt done a nous faire sup- poser que Pyrrhos, affirmant sa qualite de basileus, proceda de bonne heure a remission d’un monnayage personnel, et que d’autre part le repertoire de la propagande destinee a eveiller l’imagination de ses soldats, etait emprunte au vieux fond national ou se rencontraient le Zeus couronne de chene, et le heros casque dont la legende attisait tous les courages. Par un dernier subterfuge, le portrait meme de Pyrrhos devait s’insinuer dans la suite iconographique du fils de Pelee et du vainqueur d’Hector.13

12 Jean Gage, “Pyrrhus et l’influence religieuse dc Dodone dans l’ltalie primitive”, RIIR, 1954, 1 (Avril-Juin) et 3 (Octobre-Decembre). P. Vuilleumier, Tarente des origines a la conquete romainc, x937- J- Pcrret, Les Origines de la legende troyenne de Rome. Cf. la critique d’A. Momigliano, JRS, 35. 1945. PP- 99 ss.

Giannelli, Culti e miti della Magna Grecia, pp. 167-169. Dione, paredre de Zeus Naios a Dodone, variante ou prototype de Hera Pelasgis, dont Diomdde etait devot. Diomede etait cense le fils d*1 la Dione de Dodone, dont il aurait emporte ou retrouve l’image au cours de ses migrations en Italie. P. Fraser, “A bronze from Dodona”, JHS, 1954, p. 56: dedicace a Zeus Naios.

Carapanos, Dodone ct ses mines. Bursian, “Die wissenschaftlichen Ergebnisse dcr Ausgrabungen in Dodona”, Sitzber. d. Akad. zu Munchcn, 1878. Cf. Cook, Zeus, I, p. 369. II, I, p. 753. La tete de Dione parait sur des monnaies de bronze des Athamanes, vers 220-190; Zeus couronne de chene sur des monnaies d argent de Cassope, vers 238-168. Enfin, sur une monnaie de bronze dc Do¬ done, du milieu du IV* s., on voit l’image de Zeus, la tete ceinte d une simple taenia, avec la legende AIA NAON: E. Babelon, Inv. Waddington, p. 431. Tli. Reinach, Rev. Arch., 1905, II, p. 97.

13 Pausanias, 11,11,7: landis que les envoyes des Tarentins parlaient ainsi, le souvenir de la prise

6o

JEAN BABELON

Les premieres pieces qui se presentent a notre examen dans ce monnayage ostentatoire par lequel le roi d’Epire manifeste sa puissance et revendique sa qualite de basileus, sont de beaux tetradrachmes, d’execution tres soignee, frappes sur des flans larges et minces, d’etalon ponderal attique. Leur style est si particu- lier qu’il les a fait juger avec quelque severite par les partisans intransigeants du classicisme, mais par la meme ils se signalent a notre attention, selon le penchant plus marque que nous avons de nos jours pour le gout baroque. D’autre part, on pent relever une certaine opposition entre le developpement assez tumultueux de beffigie de Zeus dodoneen qui parait au droit, et la rigueur archaisante de la figure de Dione qui en forme le revers. Ces monnaies si raffinees marquent done un jalon dans l’histoire de l’art. II suffit pour s’en convaincre de les comparer avec celles des diadoques a la meme epoque.

Les differents exemplaires que nous avons eus entre les mains autorisent un classement dont on trouvera ici l’esquisse, mais les nuances que Ton peut relever de bun a b autre ne peuvent etre dument interpretees qu’en se referant a la maniere particuliere de graveurs employes en un meme atelier, et dont bactivite se concentre en un bref laps de temps. Les besoins de la tresorerie du roi, dans les circonstances que nous pouvons imaginer, au debut de sa campagne d’ltalie, exigeaient l’emission cl’un numeraire tres abondant, et la production en fut hatee par tous les moyens.

Dans son Memoire sur les medailles siciliennes de Pyrrhus, en date du 16 De- cembre 1831, Raoul-Rochette enumere deux tetradrachmes provenant de Sidle, deux autres au Cabinet des medailles, de provenance inconnue, un dans la collec¬ tion du due de Luynes, provenant de Carelli, qui l’avait regu de Sicile, un au British Museum ( Recueil de Combe, 1814, p. 117, n. 2), un autre au British Museum, provenant de hancienne collection Payne-Knight, trouve en Sicile, publie par Sestini (Musee Venuti, Descriptio numorum veterum, pi. IV, n. 1, p. 163), un autre, inedit, pi. 1, n. 7.

Les tetradrachmes en question portent au droit la tete de Zeus dodoneen, et au revers la figure de Dione assise. Nous decrivons les deux faces successivement, en en analysant les variantes.

de Troie s’imposa soudain a Pyrrhus, avec l’assurance de vaincre, lui aussi, et pour les memes raisons: n’allait-il pas combattre contre une colonie de Troyens, lui, le descendant d’Achille?” On me permettra de renvoyer a ce sujet a ma communication au Congres numismatique de 1953. relative aux monnaies de Pyrrhos qui ont pour types la tete d’Achille et au revers Thetis sur l’hippocampe. Cf. E. Ciaceri, Storia della Magna Graecia, III (Rome, 1916), p. 78, n. 4. Un vase trouve a Tarente et actuellement an musee de Benevent porte l’image de Pyrrhos, avec la le- gende paaiAeuc; fluppo.

LE ROI PYRRHOS

6l

DROIT

A. Tete de Zeus dodoneen, a g., couronne de chene, Sous la tranche du cou, oE Le nez droit, la pupille marquee par un point pres de la paupiere superieure. L’oeil est enfonce sous l’arcade sourciliere, indiquee par un bourrelet qui va de la tempe a la racine du nez, laissant un espace sous les cheveux. Le rameau de chene qui forme la couronne, s etend du sommet du crane jusqu a la tempe. Boucles de cheveux en S, etagees jusqu a la nuque. La tranche du cou est ondulee.

Luynes, n°. 1898 (Cabinet des Medailles, 294). i6gr. 61 Peanche VII, i

Am. Num. Soc.

Locker Lampson, n°. 187. i6gr. 82.

Naples, Fiorelli, n°. 6831 (OAT ?).

Jameson, n°. 1127. i6gr. 95.

Pozzi, n°. 1287. i6gr. 52

Fitzwilliam Museum, n°. 5164. I7gr. 05.

E. Seltman, RBN, 1912, p. 1. Trouve a Gerace.

Coll. Collignon, n°8. 228, 229.

Naville, X, 1925. Hirsch, n°. 533. Trouve a Cacace. i6gr. 71.

Weber, n°. 3014. i6gr.82.

Fiorelli, Santangelo, n°. 10340.

Mionnet, nos. 23, 24 (avec le foudre). Suppl. n°. 10. Raoul-Roehette, op. cit., pi. 1, n°. 7.

BM. A Guide to the Principal Coins of the Greeks, pi. 37, n°. 17. BMC, Thessaly, p. in, n°. 6. ibgr. 70.

Ilya un coin de Becker, avec le foudre au droit, et au revers A. G.F. Hill, Becker the Counter¬ feiter, 1924, pi. IV, n°. 57.

B. Tete de Zeus dodoneen, a g., couronne de chene. Sous la tranche du cou, A. Le nez legerement ondule. La pupille marquee au centre du globe oculaire. L’oeil enfonce dans un bourrelet de chair assez accentue. La couronne de chene enlacee dans les boucles de la chevelure, en un simple rameau qui va de la nuque au-dessus du crane. La tranche du cou en arc de cercle a peine ondule.

BM. Lloyd Bequest

Coll. Walcher de Moltheim, no. 1299. i6gr.

Cabinet des Medailles, no. 295. i7gr. Planche VII, 3

Coll, de Nanteuil, trouve a l’emplacement de Locres. ibgr. 28.

(.at. Ad. Hess, Lucerne, 1956, no. 288: Bement, no. 972.

Coll. Warren, no. 735. i6gr. 98.

A. Baldwin Brett, Museum of Fine Arts. Boston, Catalogue of Greek Coins (Boston, 1955), no. 944. ibgr. 98.

62

JEAN BABELON

The Pierpont Morgan Collection. Catalogue of the Greek and Roman Coins, Abukir Medallions, Roman Gold Bar , arranged by Wayte Raymond (New York, 1953), n°. 154- Ex. Sambon Coll. 415, du Chastel, 104.

C. Tete de Zeus dodoneen, a g. couronne de chene.

Le nez presque droit. La pupille marquee d’un point au centre du globe de l’oeil. Le bourrelet de chair au-dessus de l’oeil est legerement accentue. La couronne de chene se divise en deux rameaux, l’un s’etend jusqu’au sommet du crane, l’autre jusqu’a la tempe. La tranche du cou presque droite.

BMC, Thessaly, p. hi, no. 7. i6gr. 13.

D. Tete de Zeus dodoneen, a g.

Le relief est relativement peu accentue. Le nez ondule. Pli marque entre le nez

et la joue. Les boucles de cheveux et de la barbe sont serrees et sommairement indiquees. L’oeil est relativement petit, l’iris et la pupille indiques. La couronne de chene est etalee sur le crane et n’atteint pas la tempe sechement modelee. La tranche du cou est rectiligne.

Cabinet des Medailles, no. 296. i6gr. 37. Planche VII, 2

REVERS

a. BA2IAEQ2 nYPPOY. Dione drapee, coiffee du calathos, assise a g. sur un trone a dossier, tenant un long sceptre, et soulevant son peplos de sa main g. A l’ex. A.

Le dossier du trone est carre. Aux deux angles, un ornement quadrangulaire. Les pieds sont tournes et forme de balustres, avec deux cercles, les pointes en bas. Le sceptre, pose transversalement et tenu par la main dr. va de la ligne de terre jusqu’au dessus du calathos. La main g. est ouverte, les doigts ecartes, sauf le pouce et l’index qui tiennent un pan du peplos a la hauteur de la tete. Les cheveux sont coiffes en chignon au dessus de la nuque. Un rang de perles, termine par deux ornements allonges, pend le long du cou, sur le haut de l’epaule g.

Cabinet des Medailles, n°. 296. Raoul-Rochette, op. cit., pi. 1, n°. 7. Planche VII, 2

BMC, Thessaly, p. hi, n°. 6.

ANS.

b. Meme legende. Dione drapee, assise a g. sur un trone.

Le sceptre est place obliquement, et vient toucher l’O de la legende. La tete est relativement grosse, et le calathos petit. Rang de perles tombant du chignon, Ion-

LE ROI PYRRHOS

63

geant le coil et le haut de l’epaule g. On distingue le pendant sur 1 epaule dr. Les pieds du trone se terminent par deux toupies. Le haut du dossier est surmonte d’ornements en forme de W. Le sceptre se termine par un fleuron.

BM. Lloyd Bequest.

c. Meme legende. Dione drapee, assise a g. sur un trone.

La tete est penchee en avant. Le long du cou, une tresse de cheveux. Le haut du dossier du trone est orne de cones renverses. Les pieds du trone se terminent par deux toupies.

Cabinet de Medailles, no. 295. Planche VII, 3

d. Meme legende. Dione assise a g. sur un trone.

Le sceptre passe obliquement entre l’O et l’Y de la legende. La tete est petite, sur un tres long cou. Le long du cou, un rang de perles. Les pieds du trone se terminent par deux toupies. Le haut du dossier est orne de fleurons.

BMC, Thessaly, p. in, no. 7.

e. Meme legende. Dione assise a g. sur un trone.

La tete est relativement grosse. Le sceptre est indique par un grenetis. Les montants du trone sont surmontes de petits triangles. Les pieds du trone presen- tent trois ressauts circulaires. A Tex. A.

Cabinet des Medailles, no. 294. Planche VI I, i

BRONZE

Tete de Zeus dodoneen, couronne de chene, a dr. R/ BAIIAEfil TTYPPOY. Foudre. Couronne de chene.

BMC, Thessaly, p. 114, no. 40. Au dessous, A. Autre exemplaire, au rev. B (YP.

Fiorelli, Naples.

Imhoof Blumer, Monnaies grecques, n°. 48.

Coll. Weber, n°. 3020, B (YP. N°. 3021.

SNG, Copenhagen, Epirus n°. 100, A-102, B. Coll. Lambros et Argyropoulo.

Eckhel, n°. 6. Theupolus.

1' it /.william Mus., no. 5174, au rev. BA FTYP, cheval au galop. Mionnet, no. 26. Theupolus. Mionnet, Suppl., no. 29, t£te de Zeus laure, ag. au rev. monogr. couronne de chene, coll, du marquis de Lagoy - Mionnet, Suppl. no. 16, tete de Zeus laure, a g. Gessner, tab. IV, fig. 20. Cabinet des Medailles, nos. 3^ 3x8, 319, 320.

64

JEAN BABELON

L’image de Zeus dodoneen, couronne de chene, evoque d’une fa^on precise le plus saisissant des paysages epirotes et le culte qui s’y etait etabli. C’est a Dodone au pied des sommets escarpes du Tomaros que, des le temps des Pelasges, Zeus avait re<pi le foudre, et s’etait manifesto par l’oracle issu du bruissement propage par les feuilles des chenes seculaires. Zeus Naos de Dodone etait essentiellement le dieu des eaux qui ruisselaient entre leurs racines puissantes. Une ancienne tradition, recueillie par Etienne de Byzance, et remontant a l’historien Kineas, se refere sur ce point, a la ville de Phegon, en Ihessalie, d oil 1 oracle de Zeus Phegonaios avait emigre en Epire. Apres la bataille d’Asculum, en 274, Pyrrhos ayant reporte son quartier general a Tarente, pour rhivernage, invoque 1 appui de Zeus Naios, et envoie son butin au temple de Dodone.14

Ouant au type plastique des tetradrachmes au nom de Pyrrhos basileus, on remarque a premiere vue le style exceptionnel qui les distingue de toutes les representations analogues, et dans 1’ alliance de la couronne de chene et de la chevelure, le gout baroque de la composition. Le lien de la couronne enserre etroitement la tete jusqu’au dessus de la nuque, et contient mal des boucles tumultueuses. Le volume des muscles faciaux est modele avec soin, l’oeil enfonce sous un bourrelet de chair, le front ondule 011 convulse, le pommettes saillantes, le nez bossue. La barbe est disposee en trois rangs de boucles etagees, la moustache dessine une lourde virgule. Nous sommes loin ici du Zeus olympien fige dans une majeste immobile. LTn nouveau genie s’est empare du theme traditionnel purement statique, 011 plutot, il lui a substitue le fremissement meme du feuillage qui transmettait les ondes du message divin.

Ouelle que soit son originalite, cette effigie, dont nous ne saurions designer hauteur, si novateur, n’est pas isolee dans hart monetaire. Mais si elle se rattache a des types similaires, quand nous la plains a cote de ceux-ci, son caractere particulier s’accuse avec nettete. C’est bien a Elis que nous pouvons situer le point de depart. Alexandre le Molosse, en Epire, avant ses campagnes d Italie, en 334> avait fait copier sur ses monnaies les types des monnaies d Olympie.10 Apies avoir conquis les colonies eleennes d’Epire, il etait reste en etroites lelations avec

14 Pausanias, I, 13, 2. Voyez le buste de Berlin, qui represente Zeus vocos (sic), couronne de roseaux, les cheveux et la barbe ruisselants. Welcker, Griechischc Gotterlehre, I, p. 203. Cf. 1 inscription publiee par Carapanos, 45, 16, 47, 18: All Naou Kai Afova [6 8eiva] epcoTa tov Aia Kai tccv Aicovav. L archi- pretre portait le nom de vaiapxos, et des jeux dits Naia avaient ete institues. Le dieu primitif des sources fut identifie plus tard avec Zeus. Voy. Kern, cite par Kruse, dans RE. Comparez la tete barbue trouvee a Chios: Corolla Ludwig Curtins, I937> PP- I04 ss-> ph 25> 2&> 27-

15 M. P. Vlasto, “Alexander Son of Neoptolemos of Epirus”, NC, 1926, pi. X, 7, 8. Autre type, 9, 10.

Plate VII

LE ROI PYRRHOS

1-3, TfHradrachmes de Pyrrhos. — 4, Statere des Bruttiens. — 6, Staterr d’Ah xandir I< Moloss. 0, Stature de Locres. — 7-9, Didrachmes dr Pyrrhos.

LE ROI PYRRHOS

65

leur metropole, et avait fait figurer en symbole sur son numeraire la tete d’aigle d’Elis. Sur les monnaies frappees probablement en Epire vers 342-334, Zeus apparait couronne de chene, la barbe epaisse, le nez droit et long. Vlasto citait a ce sujet un passage de Denys d’ Halicarnasse: "Mfi -rrapepycos 6e pr|8’ ccutov tov crrt- cpocvov TrapeAGt)S ocms elrj - o08e yap ev toutco yevopevos aTropf|CTEis ETraivou- Tr|v pev 8p0v, oti bpa A 10s, Kai oti f) TrpcoTT^ Kai TTp£a(3uTdTri Tpo9q tcov dvGpcbircov Kai oti ovk &9COVOS, &AAa £90£y£ar6 ttote ev AooScbvq.”16 PLANCHE VII, 5

Le foudre etait done un symbole epirote, de meme que l’aigle et le fer de lance, insigne aeacide, comme l’a note Evans, en se referant a une epigrammede Leonidas (CXXIY ,4) relative a Pyrrhos: "AiypriTai Kai vOv Kai -napos AiaKiSai.”17

Etabli par la suite a Tarente, de 334 a 332, Alexandre le Molosse fait frapper des monnaies d’argent avec pour types Zeus dodoneen, couronne de chene, la barbe courte et frisee, les cheveux longs, le tout en liaut relief. Ernest Babelon ecrivait a ce propos: “C’est uniquement par l’appreciation des merites artistiques qu’on peut admettreque les pieces d’or et d’argent, qui sont du meilleur style, ont ete frappees en Italie, peut-etre dans l’atelier de Tarente ou celui de Locres, en Bruttium, tandis que les pieces de bronze et celles d’argent d’un style inferieur, sont epirotes.”18 Mais nous savons que les graveurs etaient itinerants qu’on les attirait de ville en ville ou d’Etat en Etat, pour se parer de leur adresse, et qu’une fois en place, ils pouvaient etre secondes par des artistes locaux qui s’etaient mis a l’ecole des maitres etrangers.19 D’autre part, nous citions a l’instant les monnaies

d’Alexandre le Molosse, frappees vraisemblament en Epire, et dont le style nesaurait nullement etre qualifie d’inferieur. Rien ne s’opposeen fait a ce que les artistes de la Grande Grece aient traverse la mer Ionienne pour se rendre soit a Dodone meme, soit a Ambracie, ou en quelque lieu choisi par Pyrrhos pour y etablir son atelier royal.

On a rapproche aussi le type de Zeus dodoneen des monnaies d’Alexandre le Molosse, de la tete voilee dite Persephone, Gaia, Aphrodite ou Amphitrite, telle

18 Ars rhet. I, 6.

17 A. J. Evans, ‘‘The Horsemen of Tarentum ’, NC, 1889, p. 142, n°. 172. Le fer de lance est le type des monnaies epirotes autonomes. Cf. les monnaies d’or et d’argent frappees A Tarente pendant 1 hegemonie de Pyrrhos. On peut comparer la tete de Zeus des monnaies d’Alexandre le Molosse avec la tete barbue du guerrier combattant une Amazone, sur un fragment de bronze de Siris, conserve au BM, oil Ton a voulu reconnaitre la cuirasse de Pyrrhos, mais qui remonte plutbt a Alexandre le Molosse.

18 Babelon, Traitt , II, 4, p. 152.

19 Macdonald, JUS, 1907, p. 149.

) Centennial

66

JEAN BABELON

qu’elle apparait sur les monnaies cl’or de Tarente, avec la signature KAA. II faudrait encore en trouver le prototype sur les stateres d’argent de la ligue arca- dienne, frappes vers 363-343, ou le type leonin de Zeus revele l’influence de Lysippe.

Une autre relation doit etre etablie avec les monnaies des Brettiens qui ont pour types la tete de Poseidon et au revers Amphitrite, et surtout avec le statere d’or des memes Brettiens, ou la tete de Zeus est proche parente de celle du Zeus dodoneen de Pyrrhos. Toutefois, un examen attentif nous oblige de constater dans le premier cas moins de souplesse dans le maniement du burin, un style plus rigide et plus froid, notamment dans le modele de la chevelure disposee en trois boucles alignees. Planche VII, 7

Dernier element de comparaison, les monnaies frappees par les Mamertins de 278 a 270. Cette fois, l’influence des evenements consecutifs a la campagne de Pyrrhos est patente. Les monetaires changent d’etalon ponderal et adoptent pour type la tete de leur dieu national Ares, tournee vers la gauche et non plus vers la droite. Le style en est plus hellenise que par le passe, et se rapproche du Zeus de Dodone.20

La figure de Dione, qui forme le type de revers des tetradrachmes de Pyrrhos fait contraste avec le caractere baroque de la tete de Zeus. On notera le caractere archaisant de la composition si severe. Peut-etre - mais on ne saurait le prouver - s’agit-il de la reproduction libre d’une statue, comme il arrive pour l’Athena Alkidemos qui parait sur d’autres monnaies de Pyrrhos. Le motif se presente toutefois sous un aspect assez variable, si nous comparons les coins differents qui sont sous nos yeux, la deesse etant ici plus modelee, dans un metal plus souple et plus gras, penchee en avant, la plus lineaire et dessinee d’un trait plus dur, l’attitude plus raide, le sceptre indique par un grenetis secheinent grave.

Ou done ont ete frappes les tetradrachmes du roi Pyrrhos ? Ouand il s est agi de localiser cet atelier inconnu, on a pu tirer argument de la similitude que ces monnaies presentent avec celles de Locres. Sur ces dernieres, l’effigie de Zeus decele un style anterieur a celui du Zeus dodoneen de Pyrrhos. En tout etat de cause, on ne saurait attribuer a un merne artiste, contrairement a une opinion qui s’est repandue parmi les numismates, la gravure de ces deux categories de coins.

Nous sommes amenes par la a passer en revue le monnayage de Locres, pour lui comparer les monnaies de Pyrrhos qui nous occupent presentement. Locres, la cite florissante situee au Sud de la presqu’ile italienne, sur la mer de Sicile, 20 Margit Sarstrom, A Study in the Coinage of the Mamertines (Bale et Lund, 194°)

LE ROI PYRRHOS

67

emit de bonne heure des monnaies qui attestent son importance commerciale, comparable a celle de Tarente. Les premieres remontent au milieu du IVe siecle. Le prestige en est reste tel que Barclay Head attribuait la haute qualite artistique des monnaies d’Alexandre, fils de Neoptoleme, a des graveurs originaires de Locres. Vlasto considerait au contraire que les monnaies de Locres etaient des copies du statere de T arente. Mais cette derniere opinion ne s’accorde guere avec la chrono- logie, si nous admettons que les monnaies dor frappees en Italie par le Molosse, en 332-330> avec pour type la tete de Zeus dodoneen, s’alignent sur celles de Tarente, de Locres et de Syracuse, dont les graveurs devaient appartenir a la meme ecole.

Une des monnaies les plus significatives de Locres est le statere qui porte la tete de Zeus, comparable au Zeus d’Elis, et au revers, avec la legende EIPHNH AOKPftN, un type inspire sans doute des monnaies de Terina. Cette piece se place aussitot apres le milieu du IVe siecle.21

Quant aux stateres bien connus de Locres, qui nous presentent au droit la tete de Zeus couronne de laurier, et au revers, avec la legende AOKPflN, Rome assise sur un trone, designee par le nom de PftMA, et une figure qui lui rend hommage, soulignee par la legende niZTIS!, ils se rapportent aux evenements consecutifs au debarquement de Pyrrhos en Italie. Apres la bataille d’Heraclee, Locres tombe aux mains du vainqueur. Le succes incertain du roi d’Epire provoque la reaction du parti romain, suffisamment outrage par le pillage du temple de Persephone auquel Pyrrhos avait laisse ses troupes s’abandonner. La monnaie fut frappee pour affirmer la fidelite de Rome envers ses allies parfois defaillants. La tete de Zeus qui en fox me 1 un des types n’est pas celle du Zeus dodoneen, puisque le laurier ceint les tempes du dieu, mais l’effigie est de tres pres analogue a celle dont nous avons examine les traits sur les tetradrachmes de Pyrrhos: l'imitation est fiagrante, et il suffit de comparer ces stateres a ceux que nous citions ci-dessus pour constater qu’un nouveau style avait ete des lors introduit dans l’atelier locrien. Planche VII, 6

C’est sans doute pour des raisons analogues que le revers du statere dor des Brettiens repiesentant Amphitrite semble une replique tres fidcle du revers du didrachme de Pyrrhos, ou Ton voit Thetis assise sur le cheval marin. PLANCRF vii,7^ Nous avons vu que les stateres d’Alexandre le Molosse, avec en symbole la tete

BMC, Italy, p. 364, n°. 7. HN2, p. 102. D’autres dates ont ete donnees: Milani, peu apr£s 400; Regling, vers 38S ( Klio , VI, p. 514). Le style de ces pieces nesoutient aucunement cette datalion. Vers le milieu du IV s. les monnaies corinthiennes circulent a Locres, et de meme les monnaies syracusaincs. Orsi, A ot. Scavi, 1912, p. 372, et RIN, 1908, pp. 339-40, trouvaille dc Geracc.

>•

68

JEAN BABELON

d’aigle, ont du etre frappes en Epire plutot qu’a Locres. D’autres ont pour type la tete de Zeus dodoneen et le foudre, avec en symboles le fer de lance et l’etoile, types specifiquement epirotes. Giesecke remarque a ce sujet qu ’Alexandre se montre ici le precurseur de Pyrrhos, en ceci qu’il tente d’introduire en Italie une monnaie nouvelle, internationale, et qu’il voit par la le moyen d’eterniser son nom en donnant a son numeraire l’attrait de la qualite artistique. Sur d’autres stateres du meme poids, ygr. 60, apparait la tete de Zeus, comparable a celle des tetra- drachmes de Pyrrhos, mais lauree, et d’un style nettement different, avec au revers l’aigle devorant un lievre, pareil a celui des decadrachmes d’Agrigente datant de 413. 22 Citons enfin les monnaies de bronze frappees a Locres: Tete d’ Athena coiffee d’un casque corinthien a panache, R/ AOKPftN Persephone assise sur un trone tenant un sceptre et une patere.230n retrouve dans le type du droit les traits de la tete d’Achille que nous presentent les monnaies de Pyrrhos. Ce monnayage evoque les monnaies de Hieron et des Bruttiens anterieures a l’arrivee de Pyrrhos.24

Les trouvailles ne nous apportent guere d’argument concluant sur le lieu d’emission des monnaies de Pyrrhos. Celle de Gerace, l’antique Locres, qui remonte a 1838, a pu paraitre decisive. Raoul-Rochette a rapporte les circon- stances de la decouverte du tresor, par des femmes, dans le ravin de Milligri. Le tout etait contenu dans un vase en terre. Outre vingt tetradrachmes de Pyrrhos, il y avait des monnaies de Syracuse, des monnaies de Palerme a legende pheni- cienne, et des Pegases. Presque toutes celles de Pyrrhos vinrent aux mains de la niece du cure, Francesco Mani, avant d’etre deposees a Naples. La nouvelle en parvint a Raoul-Rochette par l’intermediaire de Pasquale Scaioli.25

Un autre tresor a ete trouve a Gerace, en 1904. D’apres Paolo Orsi, qui l’a publie, il aurait ete enfoui vers 282. Mais nous savons que Pyrrhos nedebarqua en Italie qu’en 280. Ouelques tetradrachmes au type de Zeus dodoneen faisaient partie de ce depot.26

22 Giesecke, Italia numismatica, p. 90, pi. 11, nos. iG, 17.

23 Garrucci, Monete dell’ Italia antica, pi. CXI I.

24 Giesecke, Italia numismatica , p. 109, pi. 15, n°. 1.

25 Sydney P. Noe, A Bibliography of Coin Hoards 2, NNM, No. 78, p. 54, cite la trouvaille de Carlen- tini, en Sidle, en 1909: Une monnaie d’or de Pyrrhos. Voy. P. Orsi, Not. Scavi, 1912, p. 372 - Gerace, 1838. Les vingt tetradrachmes de ce depot ont ete repartis entre plusieurs collections: Santangelo, Durand, Millingen, Stewart.

26 Paolo Orsi, KIN, 190S, pp. 339, 340. Les pieces ont ete vendues a Hirsch par Virzi,de Palerme. Six exemplaires ont ete acquis pour l’Etat italien par Corrado Ricci. SNG, Lloyd Collection, n°. 656.

LE KOI PYRRHOS

69

Enfin E. J. Seltsam signale “une monnaie importante de Pyrrhus, roi d’Epire” L’exemplaire en question a ete trouve a Gerace, dansle lit d’un torrent. II fut acquis en Janvier 1897, par Chapman, qui faisait des fouilles dans la region, puis il fut vendu a Frank Sherman Benson. Chapman, en 1897, avait decouvert l’emplacement du temple dorien du V° siecle a Locres.27

Rien de tout cela ne nous oblige a croire que les tetradrachmes de Pyrrhos aient ete frappes a Locres. Pyrrhos a du preparer son expedition en frappant de l’argent pour payer ses troupes, sees bataillons d’Epirotes et les mercenaires de differentes origines qu’il avait recrutes. II lui fallait en meme temps assurer sa propaganda politique et ses interets economiques. Nous pouvons supposer sams forcer la verite, qu’il aura fait venir au prealable des graveurs de Grande Grece en Epire. Un depot effectue lors de la campagne en Italie meridionale peut bien etre le tresor d’un capitaine ou la caisse d’un payeur aux armees, ou simplement le magot, d’un soldat, laisse sur le terrain precisement sur le lieu des operations.

L’activite de l’atelier monetaire de Locres, attestee par denombreuses emissions, a fait supposer que Pyrrhos en avait employe l’outillage et le personnel pour faire frapper ses propres monnaies,28 mais des considerations de technique et de style, particulierement en ce qui concerne les tetradrachmes qui ont pour type Zeus dodoneen, frappes sur un flan large et mince, avec un relief relativement peu ressenti, nous donnent a penser qu’au contraire un materiel nouveau fut alors mis en usage.

Les circonstances historiques n’eclaircissent guere le probleme, si meme dies ne contribuent pas a en rendre la solution douteuse. Apres son debarquement et son premier sejour a Tarente, Pyrrhos se rendit maitre, en effet, de Locres qui, durant les guerres samnites, etait passee du cote des Romains, inaugurant ainsi les tergiversations, pour ne pas dire les trahisons successives qui marquent son histoire. Ce n’est pas par la violence que l’Epirote s’empara de la ville. Autant que des armes, il avait de l’or et de l’argent dans les mains, et son tresor de guerre regorgeait de metal precieux. Il s’etait muni de ressources considerables qu’il repandait a foison. Ainsi s’explique le monnayage d’or des Bruttiens ou de Tauromenium, com me celui de Tarente ou de Syracuse, pendant le sejour de Pyrrhos en Italie et en Sidle.29 Ces regions ne disposaient pas, on le sait, de mines

27 RBN, 1912, p. 5.

28 Lenormant, CRAI, 1880, p. 480. HN2, p. 104. G. F. Hill, Historical Greek Coins, p. 126.

29 Giesecke, Sicilia numismatic a, p. 113 et ss.

70

JEAN BABELON

d’or a exploiter.30 Ce n’est qu’en 268 que Crotone, Locres, Thurium regurent des garnisons romaines munies des premiers quadrigati d’argent frappes sous l’autorite de Rome. L’innovation apportee par Pyrrhos consista precisement dans remission de pieces d’or et de bronze.

En 281, la flotte romaine se presenta devant Tarente,31 et les troupes romaines firent tete a Pyrrhos, mais en 280, la bataile d’Heraclee qui assure tout au moins la gloire du roi d’Epire, sinon le succes definitif de ses armes, provoque bien des defections, notamment celle d’Heraclee. Locres suit le mouvement. Sur la route de Sidle, Pyrrhos s’y arrete pour etablir une base, et il y laisse son fils Alexandre. Locres et Tarente sont done les deux points d’appui dont il s’assure. Or, aussitot apres son depart, en 277, Locres fait retour aux Romains.32 Ellen’aura done ete aux mains du roi que durant un bref laps de temps, deux ans au maximum, ce qui est suffisant pour exploiter un atelier monetaire, mais cette exploitation dut etre entravee par le peu de confiance que devaient inspirer les Locriens et les mille preparatifs d’une campagne difficile. Enfin ces observations laissent inacte la question du financement du premier debarquement en Italie.

D’autre part, on songera que la capitale de Pyrrhos etait Ambracie, l’actuelle Arta, sur le golfe d’Arta, a l’embouchure de l’Araththos, a quatre-vingt stades de la mer.33 Au Sud Est s’etendait le territoire des Amphilochiens, a l’Ouest, la lagune de Tserkalia et l’Oropos. C’est d’Ambracie qu’en 342, Alexandre, fils de Neoptoleme, etait parti pour sa campagne d’ltalie. Au milieu du IV0 siecle, Neoptoleme etendait sa domination sur une partie tout au moins de la contree, bien que plusieurs territoires epirotes fussent sous la domination de la Macedoine, ou eussent conserve leur independance, tels les Atintanes et les Athamanes, Ambracie ne se rangea sous la domination des Aeacides qu’a l’epoque de Pyrrhos, mais celui-ci semble y avoir etabli le siege de son pouvoir royal. Il avait exige et obtenu la cession de la ville par Alexandre, fils de Cassandre, en recompense de l’aide qu’il avait pretee a ce dernier contre Antipater. Philippe de Macedoine avait installe une garnison a Ambracie, qui avait recouvre son autonomie sous

30 Lippold, dans RE, 1337. Beloch, Griechische Geschichte, III, I, p. 562. Denys d’Halicarnasse, XXII. Appien, IV, 9. Niese, op. cit., II, p. 32.

31 Diodore de Sicile, XXII, 7, 5. Beloch, op. cit., I, p. 284.

32 Zonaras, VI 1 1 , 6, place la rentree de Pyrrhos a Locres en 277. Niese donne une date posterieure : 276.

33 Klotzsch, Epirotische Geschichte, p. 176. Polybe, XXII, 13. Tite Live, XXXVIII, 9. Strabon, VII, 325. Oberhummer, Akarnanien, Ambrakien, Amphilochien, Lenkas in Altertuvi (Munich, 1887). L. Cesano, “Monetazione e circolazione nel suolo dell’ antica Albania”, Atti e memorie 1st. ital. di num., 1932, pp. 47SS.

LE ROI PYRRHOS

71

Alexandre le Grand. Des 295, Pyrrhos y avait a son tour installe une garnison, et il avait pris soin d’orner la ville de monuments et d’oeuvres d’art. Un palais, le Pyrrheion y avait ete edifie, et l’acropole avait re$u de puissantes fortifications. Des privileges s’attachaient a cette dignite qu’avait re^ie la cite, et le droit de monnayage lui avait ete concede.

Par le passe, l’activite de l’atelier monetaire d’Ambracie s’etait abondamment manifestee. Vers 480, il emettait des stateres au type corinthien du Pegase, avec l’initiale de la cite, A,34 et au revers la tete de Pallas Chalinitis. Ces stateres con- tinuerent a etre frappes de 480 a 432, puis de 432 a 338, quand l’atelier fut fermc a la suite des conquetes de Philippe de Macedoine.35 De nouveaux stateres portant le monogramme (AFIEIPSTTAN), sont places par Ernest Babelon a l’epoque de Pyrrhos, peut-etre des 285. Plus tard, a partir de 238, des drachmes sont emises a Ambracie avec les initiales A M ; la tete de Dione, et au revers un obelisque orne de bandelettes.

Si la lettre et le monogramme O0E que Ton releve sur les tetradrachmes de Pyrrhos (Cabinet des Medailles, n°. 294, type A), ne se pretent a aucune inter¬ pretation, toute notre attention se porte sur la lettre A placee en evidence a l’exergue soit du droit (n°. 295), soit du revers (n08. 294, 296). Il est plausible d’y voir l’initiale de la ville d’Ambracie, et ce qui nous confirme dans cette opinion, c’est la presence de la meme lettre sur les stateres corinthiens frappes precedem- ment par la capitale de l’Epire.

34 Babelon, Traite, IV, p. 125, n°. 267.

35 Babelon, Traite, IV, pp. 141, 142. O. Ravel, The Colts of Ambracia, NNM, No. 37 (New York, 1928). BMC, Corinth, p. no, n°. 58, pi. XXIX, 12.

A HOARD OF SILVER COINS OF THE EMPIRE

OF NICAEA

Alfred R. Bellinger

In the collection of the late Thomas Whittemore, now in the Fogg Museum at Cambridge, is a hoard of scyphate silver coins of Theodore I, Lascaris, Emperor of Nicaea from his coronation in 1206 to his death in 1222. There are 99 specimens of one type [BMC, Vandals, pp. 193T, nos. 1, 2) and 139 of another unknown until Tomasso Bertele described and illustrated six specimens from his collection in a note entitled “Monete inedite d’argento di Tesdoro I Lascaris (1204-1222),” Numismatica, II (1936), pp. 91-93; there is no mention of their provenance. It has seemed advisable to publish this hoard separately in anticipation of the complete catalogue of the Whittemore Collection with that of Dumbarton Oaks planned in collaboration with Professor Philip Grierson of Caius College, Cambridge. I am indebted to the Director and staff of the Fogg Museum for facilitating this study.

There is nothing in the present records of the collection to show where the hoard was found or when it passed into Mr. Whittemore’s hands, but fortunately t lie main points can be supplied. On April 19th, 1946, the Rev. Iain C. G. Campbell, now an honorary curator in the museum of the University of Nebraska, then a teacher at the English High School in Istanbul, wrote to the late John Kingsley Birge about a hoard of silver coins in the hands of George Zakos, a collector and dealer whose headquarters were in the Grand Bazaar. It had been found at

73

74

ALFRED R. BELLINGER

Torbali, a small town between Izmir (Smyrna) and Aydin (Tralles) in western Anatolia. It is said to have been found by a man ploughing; its container was lost or destroyed but the hoard itself was believed to be complete. Mr. Campbell’s note was accompanied by an admirably detailed description of the hoard which at that time consisted of ioi coins of the first type, 146 of the second. Mr. Birge sent the material on to me hoping that I might be able to buy the coins for the Yale collection. Negotiations were still going on on July 29th, 1946, which gives a terminus post quern for Mr. Whittemore’s acquisition of the hoard. Mr. Campbell would put it in August or September, 1946, certainly before June, 1947.