Studies in Early Byzatine Gold Coinage

Editor
Metcalf, William E., W. R. O. Hahn Hahn, W. R. O. Hahn, Wolfgang Reinhard Otto. Hahn, Wolfgang Reinhard Otto. Hahn, Wolfgang Reinhard Otto.
Series
Numismatic Studies
Publisher
American Numismatic Society
Place
New York
Date
Source
Donum
Source
Worldcat
Source
Worldcat Works
Source
HathiTrust

License

CC BY-NC

Acknowledgement

Open access edition funded by the National Endowment for the Humanities/Andrew W. Mellon Foundation Humanities Open Book Program.

Export

Table of Contents

FRONT

BODY

INTRODUCTION

Wolfgang Hahn and William E. Metcalf

The far-reaching historical implications of the late Roman and Byzantine gold coinage, from the fifth century through the eighth, have made this period the focus of a great deal of research in recent decades. A great wealth of material awaits collection and interpretation, and it is not yet possible to produce a synthetic monograph on all aspects of the subject. This volume brings together contributions on various topics in the hope of demonstrating how current progress has been made possible by new or refined methods as well as by the evidence of new finds. The following introductory remarks are intended to provide a summary of the current state of scholarship and a rough sketch of the problems that remain to be solved.

Numismatic considerations, which are closely related to the political history of the period, define the termini of this volume. The permanent division of the empire between Arcadius and Honorius was reflected by different developments in the coinages of the East and the West. Various attempts to assimilate the western half of the empire during the sixth and seventh centuries all failed in the end, but they provide a fertile field for the study of monetary politics. By the end of the eighth century the empire had lost most of its western dependencies—notably the African and Italian exarchates—and with those losses provincial mints virtually ceased to exist.

These centuries have fared unevenly at the hands of numismatists. The period of Anastasius I, in modern times regarded as the beginning of the Byzantine coinage proper, has recently drawn the attention of several scholars, but the fifth century has been poorly served by modern studies, especially as the framework that could be provided by RIC 10 is still lacking. The old studies of Sabatier for the East and Cohen for the West lack photographic documentation and are in any case inadequate, since coins of the eastern and western emperors are listed separately and without regard to the mints that produced them.1 Eastern and western coins are confused, and those of the empresses are ascribed to their husband, with the consequence that (for example) the eastern issues of Galla Placidia and the western ones of Pulcheria are concealed, as is the entire system of mutual recognition pieces so characteristic of fifth century coinage. The distribution of issues of empresses who struck under more than one ruler has also been obscured.2

Tolstoi's work on the eastern emperors is written in much the same tradition and presents a wealth of material, but is confined mainly to coins in his own collection and that in the Hermitage with additional references to Sabatier. For the West Robertson's recent catalogue of a much more limited collection unfortunately follows the same princple.3 Hahn has dealt with the eastern gold and silver coinage of the reign of Theodosius II, and a similar study covering the period through the death of Zeno is in preparation; but for the West the work of Ulrich-Bana stands as a solitary milestone in a field otherwise tilled by amateurs.4

The most urgent desideratum remains the continued accumulation of numismatic material to permit the most detailed possible study of the coinage. The systematic listing of hoards and stray finds will provide a reliable picture of monetary circulation and should confirm mint attributions based on stylistic criteria. Sorting out the attributions of gold coins not distinguished by specific mint marks and allocating them to different mints will yield the outlines of monetary administration.

Moreover, though the studies of W. E. Metcalf, C. Morrisson, and D. M. Metcalf below are virtually unparalleled in the Byzantine series, the compilation of similar die studies for other mints will permit insight into their activities and provide at least a relative idea of output, which is perhaps the most important question for economic historians.

The connection between civil, fiscal, and monetary administration has been set out by J. P. C. Kent and Hendy.5 As we now have a good idea where and when to look for gold mints, the possible attributions have been considerably reduced and are more securely based. The guiding principle is the division of the gold coinage into the regular production of the four praetorian prefectures (Orien, Illyricum, Africa, and Italy) plus some extraordinary cases (the Crimea, Spain, and Sicily) in which the existence of a gold mint is at least a possibility. Under special conditions military mints seem to have been active temporarily in the East during the troubled first quarter of the seventh century.

ORIENS

The overwhelming preponderance of gold coinage was stuck in and for the eastern prefecture, where the metropolitan moneta auri operated continuously with a number of officinae; usually ten were charged with the production of solidi, the main denomination. One must suppose that fractions and ceremonial pieces (in gold and silver) were struck in a single officina, which did not need to mark its coins, although it is uncertain whether this officina (or these officinae, if each denomination came from a separate workshop) is to be identified with one of the ten solidus officinae or is in addition to them. During the fifth and (rarely) the seventh centuries a portion of the Constantinopolitan solidi, in many types, lack an officina number. These unmarked pieces may have been struck at the beginning of every issue, since they are relatively much more numerous in short issues than in longer ones.6 Perhaps this is to be explained on the assumption that special orders were given for the striking of distribution pieces, outside the normal quotas of the officinae, on the occasion of accessions or anniversaries or for other ceremonial purposes connected with the introduction of new types.

Several explanations have been proposed for cases in which solidi of Constantinopolitan fabric display officina numbers higher than ten. In the seventh century the additional letter was probably an issue mark.7 In the rare cases of sixth century solidi with higher numbers, other interpretations may be required: additional officinae detached for use in other regions,8 indictional dates,9 cooperation between two officinae, or reference to a special weight standad.10

Where secret or issue marks are involved, their proper purpose is debatable. Possible explanations include marking of an extraordinary issue with a numeral indicating its date, amount, or occasion, a special destination ordered by one or another department of the state for distribution purposes, or a special metal source. The mark ⊖ is especially frequent.11 It should be noted that the occurrence of these secret marks suddenly increases under Heraclius, when the government had to resort to loans. Dates were twice introduced to the regular solidi of Constantinople, then abandoned: 567/8 and 635-50; while a temporary preponderance of the fifth and tenth officinae over the others (especially around 610) is explained by C. Morrisson, who postulates the connection of different officinae with certain state departments.12 The high standard of weight and fineness (97 to 99 percent) was maintained in Constantinople throughout the period with only a minor decline under Constans II and Constantine IV.13

Stylistically, the Constantinopolitan gold coins display great uniformity: doubtless there were several die engravers working simultaneously with shared punches. This often makes comparison of obverse dies very difficult.14 A broader range of style can be observed when a single type is struck in large quantity or over a long period;15 this is due to changes over the course of time rather than to the existence of different eastern mints. Under normal circumstances there were no mints other than Constantinople in the eastern prefecture. The fanciful attributions of D. Ricotti Prina (Nicomedia, Cyzicus, Antioch)16 are untenable, and the acceptance of Grierson's incorrect attributions of sixth-century solidi to Antioch is unforunate.17

A more serious candidate for a gold coinage in the East is Alexandria, capital of the diocese of Egypt, which had long been isolated economically from the empire and which had its own system of local copper coinage. The irregular use of AΛᵻOB on solidi of Justin II of Constantinopolitan fabric reveals a temporary gold production with dies provided by Constantinople. Although the signature was corrected to the prescribed universal mark for gold, CONOB, production seems to have lasted for some years, since it had to be sustained by locally cut dies of distinct style. These Alexandrian solidi of Justin II, which were differentiated from Constantinopolitan issues by other secret marks, might be connected with the presence of the emperor's nephew Justin as augustalis of Egypt in 566. A similar situation, with a member of the imperial family being in charge of Egypt, occurred under Heraclius, when his cousin Nicetas had to fight the Persians and may have been short of money (616-18).18 The issue of solidi from Alexandria during the revolt of Heraclius (608-10), as proposed by Grierson, is less secure.19

There are two other groups of questionable attribution. One is of Carthaginian style and may belong to the special large-module series there (see below).20 The other shows the peculiar style of an itinerant military mint which also produced copper coins with mint marks, first of Alexandria, then of Cyprus. As Cyprus was the proper base for military campaigns in the eastern Mediterranean basin, it is plausible that it was the home of an extraordinary gold mint to meet military needs.21 The activity of this mint began under Phocas22 and continued with the unusual consular solidi of the two Heraclii, which were supplemented by fractional issues in the name of the deceased emperor Maurice Tiberius. The solidi are marked with immobilized indictional or regnal dates or by a sequence of the letters I, IX, or IΠ, which may refer to a lustral cycle.

After Heraclius' coronation there were further issues, apparently for his Persian campaigns. Their attribution to Cyprus and their dating (up into the 620s) are not universally accepted and other suggestions, such as Alexandria or Jerusalem (from which we know copper coins from the siege of 614) have been made.23 The evidence of provenance is of little help because the finds, which range from Egypt to Turkey, correlate with the extended theater of war.

End Notes

5
J. P. C. Kent, "Gold Coinage in the Later Roman Empire," Essays Mattingly, pp. 190-204; M. F. Hendy, pp. 129-54. See also the latter's important new Studies in the Byzantine Monetary Economy, c. 300-1450 (Cambridge, 1986), particularly pp. 386-423.
6
See Hahn (above, n. 4), p. 106.
7
MIB 3, pp. 84-85.
8
Following J. Lafaurie, "Un Solidus in�dit de Justinien I frapp� en Afrique," RN 1962, pp. 167-82. W. Hahn (MIB, p. 51) suggested Justinian I for Carthage, but this is refuted by C. Morrisson, below, p. 52.
9
N. Fairhead and W. Hahn, below, pp. 33-38.
10
See MIB 1, p. 50, n. 20, for light weight solidi from Justinian I to Maurice.
11
Date: MIB 2, p. 32 for Justin II; amount: p. 61 for Maurice; occasion: N for nalalis [dies imperii], p. 61; special distribution: P. Grierson, "Solidi of Phocas and Heraclius: The Chronological Framework" NC 1959, p. 137. With respect to the θ, a reference to a treasury (thesauros) is tempting (MIB 3, p. 129, and Hendy, "Studies" [above, n. 5], pp. 411-12, n. 169) but does not fit in all cases.
12
MIB 2, p. 32, for the earlier dates; for the later dates, MIB 3, pp. 86 and 124, and C. Morrisson,"Le trésor byzantin de Nikerta," RBN 118 (1972), pp. 29-91.
13
Weight and fineness: C. Morrisson, "L'or monnay� de Rome × Byzance," Comptes-rendus Acad. Inscr. 1982, pp. 203-23; decline: C. Morrisson (above, n. 12), pp. 54-55.
14
Occasionally, however, obverse die links have indicated that a reverse die may have moved from one officina to another. See P. Grierson, Coins mon�taires et officines × l'�poque du bas-empire," SM 11 (1961), pp. 1-8; C. Morrisson (above, n. 13), pp. 39-40; W. E. Metcalf, below, p. 25.
15
As under Justinian I and Maurice, MIB 2, pp. 60-61.
16
D. Ricotti Prina, La monetazione aurea delle zecche minore bizantine . . . (Rome, 1972).
17
See DOC 1, p. 133. Grierson now prefers Thessalonica: see Coins, p. 53.
18
MIB 2, pp. 45-46, for Justin; MIB 3, p. 95, for Nicetas. The Alexandrian origin of the Justin II solidi was recently doubted by Hendy, Studies (above, n. 5), p. 404, who was not aware of the Egyptian hoard evidence.
19
DOC 2, p. 207.
20
MIB 3, p. 79.
21
MIB 2, pp. 85-86, and MIB 3, p. 89.
22
MIB 2, p. 97.

ILLYRICUM

The second praetorian prefecture, Illyricum, had its small gold mint at Thessalonica. Little can be added to D. M. Metcalf's study in this volume except the hope that uncertainties regarding the fractions and the possible issues of Phocas will be elucidated by further discoveries.24 The fact that the output of this mint is estimated to be small raises the question of the extent to which it was designed to meet the needs of the whole prefecture or whether it had only to serve a small civilian department at Thessalonica (perhaps the idike trapeza of the area praefectoria).25 It is also pertinent to observe that dies provided by Constantinopolitan mint engravers were used from time to time; on the other hand, the mint showed relative independence in maintaining elements of older, out-of-date typology at least until 562, when it was brought into line at the same time that the local copper currency system was abolished.

AFRICA

The beginnings of gold coinage in the African prefecture from its creation in 534 to 578 are dealt with here by Morrisson. Before the Justinianic reconquest Carthage had never struck gold coins.26 It seems that the Vandals possessed enough gold emanating from imperial mints that they could afford to respect the gold prerogative of the emperors. The new Carthaginian gold mint was virtually a solidus mint, from which we know only a few fractional issues, struck for ceremonial purposes, from the time of Heraclius onwards. The coins were normally dated by indictional years and sometimes by regnal years as well (under Tiberius II and Maurice). It is now clear that the numerals on the early solidi of Justinian I are dates, not officina designations. The continuous dating enables us to follow the curve of the output: its fluctuation, as Morrisson has recently shown, can be connected with lustral taxation.27 Technical considerations now explain the puzzling tendency to a globular fabric from the end of the sixth century on: it obviated the need for hammering the blanks and speeded the process of striking, for less effort was required when smaller blanks were used.28 The globular solidi are, however, accompanied by limited issues on larger flans, and their deviation from the usual pattern and their accompaniment by rare fractions demonstrate their special ceremonial character.

It can easily be supposed that the strange looking globular solidi would not easily have entered the circulating medium outside Africa, and in fact the evidence for external finds is scanty.29 The African hoards have been carefully recorded by Morrisson, and to some extent these can be connected with historical events. Especially for the last years of Byzantine domination, the coins provide new evidence: they even pinpoint the date of the first Arab conquest of Carthage at the very end of 695.30 The splitting-off of the Sardinian branch mint antedates this event,31 but the circumstances remain uncertain.32 Another study by Morrisson has shown that there was a slight decrease in the weight of the African gold coins under Constans II (642-68), but that the high fineness was retained until the end.33

End Notes

23
For Alexandria, DOC 2, p. 32; for Jerusalem, see Grierson, Coins, p. 93, and Hendy, pp. 415-16. Hendy presumes a transfer of the staff of the Antiochene mint by Phocas' general Bonosus in 608; the style of the portraiture, however, shows no resemblance to the copper of Phocas from Antioch or that of Heraclius from Jerusalem.
24
S. Bendall, "A New Mint for Phocas," NCirc 92 (1984), pp. 256-57, has assigned to Thessalonica under Phocas several tremisses hitherto given to Sicily. There remains the difficulty of the contrast with Thessalonican issues of Maurice and with the Thessalonican copper of Phocas himself. Heraclius' solidus MIB 2 should probably be added to Thessalonica. The tremisses of Justinian I almost certainly have to be augmented by specimens of a fabric like MIB 191 as can be seen from comparison with the drawings of the imperial bust and diadem/hair styles on Thessalonican copper, D. M. Metcalf, The Copper Coinage of Thessalonica under Justinan I (Vienna, 1976), p. 22, 3a, and p. 24, 4d. To this group belong the following specimens: BMC Vandals, pl. 16, 16, 1.47; DOC 19.1, 1.49; Mechitarist coll., Vienna = MIB 191; Rauch 36, 20-22 Jan. 1986 = Hahn coll., 1.48; Frankfurter Münzhandlung 90, 2 Mar. 1943, 28; Kress 112, 22 June 1959, 969; Hirsch 68, 1-3 July 1970, 916; Ciani and Vinchon, 6-7 May 1955, 514, 1.51.
25
For the subdivisions of the area praefectoria, see J. Karayannopoulos, Das Finanzwesen des fr�hbyzantinischen Staates, S�dosteurop�ische Arbeiten 52 (Munich, 1958), pp. 80-84, and Hendy (above, n. 5), pp. 411-12. J. M. Carrie, Collection de l'école Française de Rome 77 (1986), p. 129, has advanced the idea that the Thessalonican moneta auri was active only in order to check the weight of the solidi paid to the treasury and to recoin the light ones; but, in fact, Thessalonican solidi tend to be somewhat lighter in weight.
26
See P. Grierson, Münzen des Mittelatters, trans. A. P. Zeller (Munich, 1976), p. 16, illus. 2-3, his supposed Vandalic solidus from Sardinia is very doubtful. See also C. Morrisson, "La Circulation de la monnaie d'or en Afrique a l'epoque vandale, bilan des trouvailles locales," M�langes de numismatique offerts × Pierre Bastien..., ed. M. Huvelin et al. (Wetteren, 1987), pp. 325-44.

ITALY

The gold of the adjacent Italian prefecture has other antecedents. The old central mint of Rome retained its role as an institution of the regular administration from the time of the western empire through the Ostrogothic dominion and into early Byzantine times. The court mint of Mediolanum, first opened as a moneta comitativa or traveling mint, had been moved to heavily fortified Ravenna in 402. Aetius reactivated the Milan mint in ca. 450 for military purposes. Thus the magistri militum of the western empire had their own money supply during the second half of the fifth century. It was only here that Odoacer could maintain a continued coinage in the name of the last western emperor, Julius Nepos, whom Zeno demanded that he recognize in 476.34

Theodoric closed the Milan mint in the late 490s, perhaps as a result of his treaty with Anastasius I, in which imperial prerogatives were newly defined.35 The same fate befell Ravenna, a demonstration that the western court had ceased to exist. A typological caesura in the Italian gold marks a new stage in the gradual assimilation to Constantinople under Theodoric.36 The distribution of fifth-century solidi among the three Italian mints (and a fourth at Arelate) is fairly clear-cut, since most of them bear mint marks, but that of the unsigned fractions is more problematic. Only the Milanese tremisses display a distinctive style.37 The differentiation between Rome and Ravenna must rest on a close comparison of dies with those of the mint marked silver pieces of small module. The lack of a monograph in this field is keenly felt. The reconstruction of the framework of the fifth-century gold issues of Italy will provide important insights into the relationship between the eastern and western courts, inasmuch as the western mints struck a certain number of coins in the name of the eastern emperor at times of mutual recognition.

From about 500 onward the mint of Rome was the only Italian moneta auri until Justinian's Gothic war once more involved Ravenna. While the Ostrogothic kings had to move their mint first to Ravenna (536-40) and then to Ticinum (540-52), the Byzantines reorganized the Rome mint in the newly installed Italian prefecture, dividing it into ten officinae. This corresponds with the Constantinopolitan model.38 The mint soon followed the praetorian prefect to Ravenna as a consequence of the second siege of Rome (545/6). The distinction between the Roman and the early Ravennate solidus issues is problematic. Although the Roman pieces under Byzantine auspices started out with a clearly cut star of six rays—differentiating them from the eastern solidi, which are marked with a star of eight rays—this characteristic disappears later in Ravenna. Again the comparison with the mint marked copper provides a guide. As activity decreased during the later part of Justinian's reign, the indication of officinae became meaning less, and the letters in question seem to have become dates. The date at which this occurred and what kind of dating was employed are still under discussion.39 Even less is understood in the case of Tiberius II, when the western authorities could not decide whether to count his years as Caesar or his years as Augustus.

The contemporary gold coinage of the last three Ostrogothic kings—Hildebad, Totila/Baduila, and Theia—served mainly to pay their Frankish auxiliaries, and therefore evoked Frankish imitations which are not always easily recognizable as such. In the last third of the sixth century the Ravenna attributions need to be purged of similar coins which should be given to Sicily (see below). There are further difficulties in sorting out the early Lombard imitations which also use the Ravenna model. In addition to the royal mint in Ticinum, several smaller workshops of varying quality operated in the centers of the duchies.40

Rome's resumption of the striking of gold seems to coincide with Ravenna's growing isolation. The attributions in the earlier period (from Tiberius II to Heraclius) have hitherto been very hypothetical,41 but from Constans II onward they become more coherent: at this time Rome's issues begin to exceed those of Ravenna in quantity. From Constans II to Justinian II they show small monograms which probably refer to church treasuries or to a magistrate who had something to do with coinage.42 In any case there is no connection with the papal monograms which made their appearance on the small silver coins of Rome from the pontificate of Constantine (708-15) onward.43

In the last three decades of the seventh century a third Italian gold mint emerges at the seat of the Neapolitan dukes. Among a group of more or less imitative coins, its products are at first difficult to identify, but they become intelligible as soon as we can trace a sequence of issues on which style and administrative marks can be seen to correlate.44

In the eighth century the characteristic feature of the Italian coinage is its visible decline in weight and fineness. This has attracted the application of newly developed or refined methods of metallurgical investigation: see the contribution of Oddy to this volume. In general a short age of gold in the Mediterranean world accounts for its limited striking. The fluctuations in the gold content of the Rome coins, which resulted in a temporary token coinage of copper solidi, are extremely puzzling. These coins can be dated precisely on the basis of a continuous sequence of indictional years.45 The end of this old currency came with the Carolingian reform in Italy and with it the end of Byzantine hegemony over Rome. The latest Byzantine issue of copper solidi, as recognized by R. Denk,46 dates from 777/8. The end of gold coinage in Ravenna had come a little earlier. After the Lombard conquest of the city in 751, King Aistwulf continued to strike there in the exarch's mint,47 but under Frankish pressure he lost the city to the Popes and with the departure of the court there was no longer any need to maintain a gold mint. The differentiation between Rome, Ravenna and Naples under Leo III and Constantine V is less difficult, because the very distinct style of Rome is also seen on the small silver coins with papal monograms mentioned above, and also because a mint signature is occasionally used. On the other hand we can observe affinities between coins of Naples and those of Beneventum, where the Lombard dukes are found striking coins with ducal monograms from about 705 onward.48 Naples is the only Byzantine mint in Italy where a limited production has to be reckoned with after 800.49 The remnants of Byzantine territory in southern Italy were by now provided with coinage from Syracuse.

Having dealt with the gold mints under regular administration, there remain attributions to possible mints with special status. Three localities are considered: the Crimea (Cherson), Spain (Cartagena), and Sicily (Syracuse). Of these only the Spanish and the later Sicilian attributions (after 640) are secure; the Chersonese and the earlier Sicilian attributions are affected by various uncertainties.

End Notes

27
C. Morrisson, "Estimation du volume des solidi de Tib�re et Maurice × Carthage," PACT 5 (1981), pp. 267-84.
28
For hammering, MIB 3, p. 91. For smaller blanks, see F. Delamare, P. Montmitonnet, and C. Morrisson, "A Mechanical Analysis of Coin Striking: Its Application to the Evolution of Byzantine Gold Solidi Minted in Constantinople and Carthage," Journal of Mechanical Working Technology 10 (1986), pp. 253-71.
29
There are only a few intruders in later Sicilian hoards, at a time when Sicilian gold coins were struck on flans of similar shape.
30
For the sixth century, see C. Morrisson, "Le trésor byzantin de Souassi," BSFN 37 (1982), pp. 214-15, and her contribution in this volume; for the seventh, R. Guery, C. Morrisson, and H. Slim, Rougga, le trésor de monnaies d'or byzantines (Rome, 1982), p. 71. For the last issue see MIB 3, p. 167.
31
MIB 3, p. 167.
32
C. Morrisson, "Un trésor de solidi de Constantin IV de Carthage," RN 1980, p. 159; contrast MIB 3, p. 153.
33
C. Morrisson, J. Barrandon, and P. Poirier, "Nouvelles recherches sur l'histoire mon�taire byzantine: �volution compar�e de la monnaie d'or × Constantinople et dans les provinces d'Afrique et de Sicile," J�B 33 (1983), pp. 267-86, esp. pp. 274-75.
34
J. P. C. Kent, "Julius Nepos and the Fall of the Western Empire," Corolla memoriae E. Swoboda dedicata (Graz, 1966), pp. 146-50, and W. Hahn, "Die letzten Jahre der Mediolanenser Münzpr�gung vor der Schliessung der Münzst�tte durch Theoderich," G. Gorini, ed., La zecca di Milano. Atti del convegno internazionale di studio Milano 9-14 maggio 1983 (Milan, 1984), pp. 229-40.
35
MIB 3, p. 56, and Hahn (above, n. 34), p. 235.
36
MIB 1, pp. 77-78.
37
For documentation, see O. Ulrich-Bana, Moneta Mediolanensis 352-498 (Venice, 1949) and W. Hahn, "Die Münzst�tte Rom unter den Kaisern Julius Nepos, Zeno, Romulus Augustus und Basiliscus (474-91)," RIN (forthcoming).
38
MIB 1, pp. 53-54.
39
MIB 3, pp. 66-67, and the contribution by N. Fairhead and W. Hahn in this volume.
40
For Ticinum see E. Bernareggi, Il sistema economico e la monetazione dei Longobardi nell'Italia superiore (Milan, 1960); comparative material for the unattributed pieces is assembled in MIB 3, pl. 55. See also P. Grierson and M. Blackburn, Medieval Coinage, vol. 1 (Cambridge, Eng., 1986), pp. 55-66.
41
W. Hahn, "More about the Minor Byzantine Gold Mints from Tiberius II to Heraclius," NCirc 87 (1979), pp. 552-55.
42
MIB 3, pp. 131 and 154.
43
See M. D. O'Hara and I. Vecchi, "A Find of Byzantine Silver from the Mint of Rome for the Period A. D. 641-752," SNR 64 (1985), pp. 105-40.

CHERSON

Cherson, a trading outpost under Byzantine hegemony, had long retained some sort of autonomy, as is documented by its anomalous copper coinage. Under Heraclius and Constans II a few solidi of eastern appearance, but differing from the Constantinopolitan fabric and marked by the sign X, have been given tentatively to Cherson.50 The lack of find evidence for them is hardly surprising since almost no gold coins have been recorded as found in the Crimean peninsula.51 Six corresponding issues are attested by single specimens;52 it seems that most of them were dated, although the elucidation of these dates is somewhat problematical.

End Notes

44
MIB 3, pp. 155 and 169-70; M. D. O'Hara, "A Curious and Interesting Solidus for the Mint of Naples under Justinian II," NCirc 96 (1988), pp. 43-44.
45
DOC 3, pp. 87-88.
46
R. Denk, "Zur Datierung der letzten byzantinischen Münzserien aus Rom," LNV 1 (1979), pp. 139-43.
47
See P. Grierson (above, n. 26), p. 46, and above (n. 40), p. 65.
48
MIB 3, p. 188. For this series see also W. A. Oddy, "Anaysis of the Gold Coinage of Beneventum," NC 1974, pp. 74-109, and P. Grierson and M. Blackburn (above, n. 40), pp. 66-72.
49
DOC, 3, pp. 84-85.
50
W. Hahn, "The Numismatic History of Cherson in Early Byzantine Times: A Survey," NCirc 86 (1978), pp. 414-15, 471-72, 521-23. The Russian specialists in the coinage of Cherson have not discussed the attribution of this solidus, but see I. V. Sokolova, Moneti i petshati vizantijskogo Chersonesa [Coins and seals of Byzantine Cherson] (Leningrad, 1983), p. 28.

SPAIN

At the opposite end of the Byzantine world, in the West, there was a military district on the Spanish coast which was loosely attached to the African prefecture. Since this had been Visigothic territory and was surrounded by the Visigothic kingdom, where the tremissis was the only denomination struck, a need for this denomination had to be met. Carthage did not supply them, since it was not a tremissis mint, and was in any case remote. They were therefore manufactured locally from the beginning of the Byzantine presence under Justinian I (554) to its end under Heraclius (615-24). They have a pronounced and peculiar style with reminiscences of Visigothic coins in their lettering and even occasional typological borrowings. The evidence of provenance makes their attribution even more secure. Since the group was first identified by Grierson,53 the whole sequence of emperors from Justinian I to Heraclius is illustrated by a handful of specimens. Those of Justin II and Tiberius II have been discovered only recently.54

SICILY

By far the most important extraordinary gold mint was that of Sicily, which became the second most active mint in the empire from the later seventh century onward. Until recently not only its beginnings but the entire first hundred years of its supposed existence were more or less obscure. The spectacular Monte Judica hoard, presented in this volume, has shed new light on the early period of Sicilian gold coinage. The position of this island had always been of exceptional importance, so it is hardly surprising to find that for a long time it was not subject to the regular administration but was more directly controlled from Constantinople. Whenever the money supply from the capital was insufficient or its maintenance was considered inadvisable, a local mint had to serve as a substitute. Its products were not, at first, intended to reveal their non-Constantinopolitan origin, and we have the same difficulties contemporaries must have had in distinguishing them. This is, incidentally, also partially true of sixth-century copper from Sicily. Constantinopolitan as well as Ravennate stylistic influences hinder the attributions. While the attribution of Sicilian solidi now seems secure throughout (with minor exceptions under Heraclius),55 the fractions offer serious problems from Maurice into the first half of Heraclius' reign. A distinct group of them, which has been given tentatively to Sicily before 610, has recently been moved to Thessalonica without satisfactorily solving all the problems connected with this transfer.56

The Sicilian style proper and the use of specific administrative marks begins late in the reign of Heraclius. Occasional dating occurs under Heraclius and Justinian II.57 The effect of the temporary stay of Constans II in Syracuse (663-68) on Sicilian coin production is still a matter of dispute: it is connected with the question of a traveling field mint accompanying the emperor and with the supposed coinage of the usurper Mezezius in Sicily, the authenticity of which has been contested.58 The weight was reduced by Justinian II in his first reign (685-95) and again under Philippicus (711-13) following the western tendency toward lower weight standards.59 The fineness fell during the reign of Leontius (695-98) and after, but after some fluctuation it was stabilized by Leo III in the 730s at a level about 15 percent lower than in Constantinople.60 When the Arabs began their piecemeal occupation of the island (a process which extended from 827 to 878) and the Byzantine presence there became more and more isolated, a final debasement in weight and fineness began which led to coins consisting of one half copper. The first metallurgical investigations pioneered by the French team have shown the outlines, but more results are desirable to follow the development in detail.

This brief survey of the history of gold mints should make evident the need for further research. To understand the correlation between monetary circulation and transactions new hoard registers are needed. Regional projects have been initiated to replace Mosser's outdated and cursory bibliography of Byzantine hoards. Aside from the African inventory mentioned above, large scale efforts have been made to begin a regional examination of all fifth to seventh-century hoards from the Balkans,61 with a view toward achieving a better understanding of the Slavic incursions. Another attempt is underway in Italy,62 where smaller museums probably house unknown hoard material. All this work should be undertaken in light of new attributions resulting from recent research. Unfortunately the eastern empire cannot be expected to receive similar coverage. We have to rely on the occasional discovery of recently unearthed hoards, which rarely come from controlled excavations63 but are mainly known through information from the trade. The dawning recognition of the importance of preserving provenances offers a faint ray of hope. Several responsible dealers are to be thanked and encouraged for their efforts in this regard.

A second approach to understanding the coinage is metrology in its broadest sense, which involves investigation not only of weight and fineness, but of the exchange rates between metals — in other words, the price of gold expressed in copper or in kind (adaeratio). Aspects of this question border on the territory of the economic historian, but the numismatic implications are primary, since the coins themselves provide the evidence for monetary reforms or debasements as documented by the introduction of new denominations such as the light weight solidus. Since Adelson's monograph on these peculiar coins of the sixth and seventh centuries, much new material has come to light,64 so that a new corpus of dies might be a promising venture. Meanwhile we have been able to include in this volume a contribution devoted to one aspect of the subject, their distribution. The importance of the light weight solidi for the reckoning of exchange rates has been argued in MIB: they enabled the payment of various sums in carats.

The differentiation between coin units and units of account is a problem which has caused much confusion when related to contemporary texts. Two examples recently discussed in the light of exchange rates (i.e. solidus prices) are the Abydos inscription and the Donori inscription.65 Both give insights into Byzantine taxation policies. Similar problems are encountered when the numerous references to payments recorded in papyri are used for numismatic analysis.66

It has been supposed that the output of the mints was dictated in advance by the authorities of the tax administration, in close connection with taxation policies. This would imply that the volume of an issue was calculated by the expected demands, an idea that has not gone unchallenged.67 Where the solidi are dated by years, we should be able to check the fluctuations in mint activity as soon as we have calculated more reliable estimates of output through die corpora. A comparison with fiscal periods such as taxation cycles and census terms might also provide new insights.68 The financing of the augustaticum, which was to be paid by the emperor to his soldiers on his quinquennial anniversaries, must have been burdensome, since Anastasius I had abolished the collatio lustralis auri argentive in 498.69 Raising of such large sums seems to have been facilitated by monetary measures, such as altering exchange rates from time to time.70 In any case, the economic situation of the empire will be elucidated by better knowledge of the quantities of gold needed in circulation to support taxation on the one hand and, on the other, for the payment of troops, civil servants, subsidies, and tributes. A new compilation of the limited evidence referring to such budget figures would be useful and would facilitate reference to complementary evidence.71

As always, new questions may seem to retard the progress of numismatic research. The authors and editors of this volume have not hesitated to advance their ideas and speculations; nor have they concealed their feelings whenever the assembled numismatic material seemed insufficient to solve inherent problems. Although they differ in their style of presentation, they hope that their comments on the numismatic evidence (of which, it is hoped, some will endure) will be noticed by historians.

End Notes

51
I. V. Sokolova, "Nakhodki vizantijskich monet VI-XIvv v Krimu" [Byzantine coin finds of the sixth to twelfth centuries in the Crimea], VV 29 (1969), pp. 254-68.
52
Add to the five specimens listed in MIB 3, pp. 218 and 242, a piece which came to light in the Schulten sale of 2-3 Nov. 1983, 990, and reappeared in Schweizerische Kreditanstalt sale, 27 Apr. 1984, 700, with commentary.
53
P. Grierson, "Una ceca bizantina en Espa�a," NumHisp 4 (1956), pp. 305-14.
54
Justin II, MIB 2, p. 41, 19; and W. J. Tomasini, The Barbaric Tremissis in Spain and Southern France, Anastasius to Leovigild, ANSNNM 152 (1964), p. 171, pl. D, 7; Tiberius II, in the museum of Seville, see F. Percz-Sindreu, Catalogo de monedas y medallas de oro, gabinete numismatico municipal (Seville, 1980), p. 29, 45.
55
The group MIB 3, 101-3, has once more been claimed for Carthage by C. Morrisson, "Note de numismatique Byzance × propos de quelquss ouvrages r�cents," RN 1983, p. 218.
56
See W. Hahn, "Some Unusual Gold Coins of Heraclius and Their Mint Attribution," NCirc 85 (1977), pp. 536-39, group C, and Hahn (above, n. 41), group C; see also S. Bendall (above, n. 24), pp. 256-57.
57
MIB 3, pp. 94, 168, and 193.
58
W. Hahn, "Mezezius in peccato suo interiit. Kritische Betrachtungen zu einem Neuling in der Münzreihe der byzantinischen Kaiser," J�B 29 (1980), pp. 61-70; MIB 3, p. 159. The coin in question was taken as genuine by P. Grierson, Coins, p. 139, and C. Morrisson expressed skepticism regarding Hahn's conclusions (at least with respect to the British Museum specimen) in "Note de numismatique byzantine × propos de quelques ouvrages r�centes," RN 1983, p. 215, n. 8. See now P. Grierson, "A semissis of Mezezius," NC 1986, pp. 231-32, where a coin in Dumbarton Oaks from the collection of Hayford Pierce (acquired in 1946), omitted from DOC, is reattributed and taken as suggestive of the authenticity of Mezezius' solidi. Grierson assumes that the coin is of Sicilian fabric, but it shares the Constantinopolitan style with the solidi and therefore cannot resolve the question. The involvement of an older Italian forger (Tardani ? — see RIN 9 [1896], p. 150) cannot be ruled out.
59
DOC 2, p. 17.
60
See Morrisson (above, n. 33), pp. 275-76.
61
Directed by V. Popović, Belgrade.
62
The project "Ripostigli monetali in Italia, documentazione dei complessi" was begun in 1980 by the Civiche raccolte numismatiche di Milano and aims to publish hoards of all period, not only the Byzantine.
63
There are, fortunately, exceptions, such as the hoards of Nikertai (above, n. 12) and Rougga (above, n. 30), published by C. Morrisson, and of Hajdučka Vodenica in Serba, published by N. Duval and V. Popović, Le Trésor de Hajdučka Vodenica, Collection de l'école Française de Rome 75 (1984), pp. 179-82.
64
The 23-carat solidi were not recognized by Adelson, but were identified by E. Leuthold, "Solidi leggieri da XXIII silique degli imperatori Mauricio, Foca ed Eraclio," RIN 1960, pp. 146-54; for the introduction of this denomination see also W. Hahn, "A propos de l'introduction des solidi l�gers de 23 carats sous Maurice," BSFN 36 (1981), pp. 96-97.
65
Abydos has recently been discussed by J.-P. Callu, "Le tarif d'Abydos et la r�forme mon�tairee d'Anastase," T. Hackens and R. Weiller, eds., Actes du J congr�s international de numismatique Berne, sept. 1979 (Louvan, 1982), pp. 731-40; W. Hahn, MIB 3, pp. 36-39; and J. Durliat and A. Guillou, "Le tarif d'Abydos (vers 492)," BCH 108 (1984), pp. 581-98. For Donori see J. Durliat, "Taxes sur l'ent�e des marchadiseses dans la cit� de Carales-Cagliari × l'�poque byzantine," DOP 36 (1982), pp. 1-14.
66
J.-M. Carr�, "Monnaie d' or et monnaee de bronze dans l'�gypte protobyzantine" Les 'd�valuations' × Rome: �poque r�publicaine et impériale 2 (Rome, 1980), pp. 253-70, and "Comptes et depenses en or," M. Manfredini, ed., Trenta testi greci da papiri letterarie e documentari editi in occasione del XVII congresso internazionaee de papirologia (Florence, 1983), pp. 112-19.
67
MIB 1, p. 17; MIB 3, p. 85, n. 5, challenged by D. M. Metcalf, "New Light on the Byzantine Coinage System," NCirc 82 (1974), p. 15.
68
C. Morrisson (above, n. 27), and below, pp. 50-51 and 54-55.
69
For the practice of collatio lustralis in connecionn with the augustiaccum see RE 4, col. 371 (Seeck, Karayannopoulos (above, n. 25), pp. 129-30, and Hendy, Studies (above, n. 5), p. 647. For the abolition of this special kind of taxation see T. N�ldecke, "Die Aufhebung des Chrysargyron durch Anastasius," BZ 13 (1904), p. 135; for Anastasius' innovations, J. Karayannopoulos, "Die Chrysoteleia der Iuga," BZ 49 (1956), pp. 72-84. Contrast R. Delmaire, "Remarques sur le chrysargyre et sa periodicit�," RN 1985, pp. 120-29.
70
See Hahn (above, n. 65), pp. 96-97. Alterations of the exchange rates are denied by Hendy, Studies (above, n. 5), p. 493, and J. Durliat, La valeur relative de l or, de l'argent et du cuivre dans l'empire protobyzantine," RN 1980, pp. 138-54. Contrast W. Hahn, "Das Römerreich der Byzantiner aus numismatischer Sicht. West-�stliche Wahrungspolitk der Byzantiner im 5.-8. Jahrhundet," SNR 65 (1985), pp. 175-86.
71
For the eighth and ninth centuries see the detailed study of W. T. Treadgold, The Byzantine State Finances in the Eighth and Ninth Centuries (New York City, 1982). For the earlier period one had to rely on older studies, e.g. E. Stein, "Zur byzantinischen Finanzgeschichte," Studien zur Geschichte des byzantinischen Reiches vornehmlich unter des Kaisers Justinus II. und Tiberius Constantinus (Stuttgart, 1919) and BZ 1924, pp. 337-87, where references to earlier, sometimes controversial studies are to be found. See now Hendy, Studies (above, n. 5), pp. 164-81.

End Notes

1
H. Cohen, Description historique des monnaies frapp�es sous l'empire romain, vol. 8 (Paris, 1892).
2
This includss the empresses Pulcheria, Galla Placidia, Eudoxia II, and Ariadne. For the last see W. Hahn, "Die M�zpr�gung für Aelia Ariadne," Festschrift für H. Hunger (Vienna, 1984), pp. 101-6.
3
A. S. Robertson, Roman Imperial Coins in the Hunter Coin Cabinet, University of Glasgow 5 (Oxford, 1982).
4
W. Hahn, "De �stliche Gold- und Silberpr�gung unter Theodosius II," LNV 1 (1979), pp. 103-28; W. Hahn, "Die Ostpr�gung des R�mischen Reiches im 5.Jahrhundet" (in press). O. Ulrich-Bansa, Moneta Mediolanensis 352-498 (Venice, 1949). For an example of the latter category, see G. Lacam, La fin de l'empire romain el le monnayage d'or en Italie (Lucerne, 1983).

THE JOINT REIGN GOLD OF JUSTIN I AND JUSTINIAN I
Plates 1-3

William E. Metcalf

In the spring of 527, Justin I, who had ruled for nine years, fell ill; on April 1, under pressure from the Senate, he co-opted Justinian, whose career he had been promoting since his own accession. Three days later, on Easter Sunday, Justinian I was crowned by the Patriarch Epiphanios; that this event occured in the Delphax rather than, as usual, in the Hippodrome, bespeaks the gravity of Justin's illness. The old emperor (he was 75 or 77) finally succumbdd on August 1.1

The gold coinage of the brief joint reign of Justin and Justinian has attracted little systematic study despite its obvious allure.2 The elegant presentation of two enthroned emperors on the obverse stands out against the otherwise bleak background of the sixth-century gold: moreover, the very brevity of the reign and the rarity of the coins render them a controllable mass and insure that a high percentage of the population surviving above ground is either preserved in major collections or illustrated in sale catalogues.

Alfred Bellinger, who made no attempt at comprehensiveness, was able to assemble a corpus of 33 specimens in 1966; less than a decade later W. Hahn increased the number of recorded specimens to 40, and to 45 by 1981.3 Neither made a systematic study of the dies, although both attemped to classify the bewildering variety of obverse variants and obverse/reverse combinations in trying to educe a structure for the coinage. The purpose of this essay is to present a fuller listing and an ordered catalogue of the coins incorporating its extensive die linkage. It is not claimed that a rational structure can be perceived, and indeed it is plausible that that system, if any, would elude detection even if all the coinage survived to us. But it is possible to take the evidence somewhat further than either Bellinger or Hahn was able to do and, if one can generalize from the experience of a relatively brief episode in the early history of the solidus, to suggest some flaws in the way we now look at imperial mint organization.

CATALOGUE

In the catalogue, the major categories follow the presentation of Hahn in MIB 1, with numbers romanized for convenience. For varieties of his groups I and II, the obverse dies are prefixed "O" and numbered serially in order of appearance; reverse dies are prefixed with the officina numeral and, within each officina, numbered serially in order of appearance. In group III, the obverses are prefixed "C" (curved); reverses continue the numbering of groups I and II. Weights are given where known, merely for the sake of completeness, since the weights approach the normative 4.5 g; the dies seem to be oriented at 6:00 without variation. All coins are illustrated except nos. 4, 18, 29-30, 41, and 43.

Group I: Backless Throne

Ia. No cross, no globe

1. O1-B1 London, BMC 1, 4.48.
2. O1-Δ1 London, BMC 2, 4.44.
3. O1-H1 a. Padua, Museo Bottacin 2206; b. Bank Leu 13, 4 May 1976, 446, 4.49.
4. -H R. N. Bridge, "Some Unpublished Byzantine Gold Coins," NCirc 78 (1970), pp. 246-47, 4, 4.40.

Ib. No cross, globe

5. O2-I1 DOC 2, 4.46.

Ic. Cross, no globe

6. O3-B1 a. Tolstoi 132, 4.45; b. Hess-Leu, 12 Apr. 1962, 549, 4.50; c. Istanbul, Archaeological Museum.

Id. Cross, globe

7. O4-B2 Paris BNC 03/Cp/A//01, 4.46 (pierced).
8. O5-B2 a. ANS 1968.131.12, 3.52 (clipped); b. Christie's, 22 Apr. 1986 (Goodacre), 68, 4.42.
9. O6-ᒥ1 a. London, BMC 4, 4.39; b. Canessa, 28 June 1923 (Caruso), 657 = Hirsch 24, 10 May 1909 (Consul Weber), 3012, 4.48.
10. O5-Δ1 a. Tolstoi 138, 4.4 = Rollin and Feuardent, 20 Apr. 1896 (Montagu), 1096; b. DOC 5b, 4.40.
11. O7-A1 H. J. Berk, Roman Coins of the Medieval World, 383-1453 A.D. (Joliet, Ill., 1986), 41, 4.19.
12. O8-E1 Rollin and Feuardent, 25 Apr. 1887 (Ponton, d'Am�court), 873.
13. O5-S1 London, BMC 5, 4.50.
14. O9-S2 London, BM 1918-5–3-2 ex Dewick = Christie's, 5 May 1885 (Tomassini), 4.39.
15. O10-S3 a. Whitting = Glendining, 16 Nov. 1950 (Hall), 2212 = Naville 3, 16 June 1922 (Evans), 297, 4.25; b. Glendining, 9 Mar. 1931, 424 (plugged;; c. Kyrenia Girdle no. 3, pl. 8.4.
16. O6-S4 NFA 18, 1 Apr. 1987, 658 = Bank Leu 13, 30 Apr. 1975, 574, 4.48.
17. O8-S4 Oxford (Evans coll.) =Schulman, 28 Feb. 1939, 92, 3.99.
18. -S Photiades cat. 113 see Bellinger [above, n. 2], p. 91, 25, and BMC 5, n.).
19. O6-Z1 ANS 1977.158.1025, 4.33.
20. O11-H1 Hess 249, 13 Nov. 1979, 459, 4.48.
21. O12-H2 Münz. u. Med. 43, 12 Nov. 1970, 541, 4.07.
22. O5-⊖1 NFA 18, 1 Apr. 1987, 659, 4.48 = NFA [1], 20 Mar. 1975, 427, 4.47 = Numismatic Fine Arts, vol. 3, 2-4 (Autumn 1974), G155.
23. O8-⊖2 a. London, BMC 6, 4.48; b. Berlin 4052, 4.45.
24. O5-I2 a. Stack's, 20 Jan. 1938 (Faelten), 1712 = Canessa, 28 June 1923 (Caruso, 658; b. Hess, 30 Apr. 1917 (Horsky), 4692; c. Münzhandlung Basel FPL 13, Nov. 1938, 74 = Ratto 436; d. Hess-Leu 41, 25 Apr. 1969, 734, 4.47.
25. O5-I1 Münz. u. Med. 52, 19 June 1975, 823.
26. O8-I3 a. Paris, BNC 03/Cp/AJ/02, 4.24; b. Münz. u. Med. 12, 11 June 1953, 909; c. Ratto 437.
27. O13-I1 Oxford, Keble coll, 4.45.
28. O6-I4 Schotten, H�bl 3470 (pierced).
29. -I Turin.
30. -I Moustier 3959 (see Bellinger [above, n. 2], p. 91, 26, and BMC 6, n.).

Group II: Throne With Straight Back

IIa. No cross, no globe, no crossbar on throne, cushions

31. O14-Γ1 Kyrenia Girdle no. 2, pl. 8.3.
32. O15-Z2 a. DOC la, 4.35; b. Glendining, 17 June 1964, 265 = Glendining, 27 May 1941, 883 = Glendining, 8 Dec. 1922 ("Foreign Prince" [Cantacuzene]), 40.
33. O16-⊖1 Tolstoi 135, 4.5.
34. O17-⊖1 Klagenfurt, Dreer.
35. O14-I5 a. London, BMC 3, 4.45; b. Kastner 10, 18 May 1976, 320 = Sternbeg, 28 Nov. 1975, 568 4.45.

IIb. Cross, globe, no crossbar on throne, cushions

36. O18-I1 a. DOC 6b.1, 4.48; b. DOC 6b.2, 4.22.

IIc. Cross, globe, cushions

37. O19-A1 Numismatic Fine Arts, vol. 3, 2-4 (Autumn 1974), G154.
38. O20-A2 Kunst u. Münzen 12, May 1974, 845.
39. O21-B2 a. Bellinger coll., 4.37; b. The Hague, 4.47.
40. O22-Γ2 DOC 3, 4.41.
41. Vienna = Longuet, pl. 6, 89.
42. O23-S5 a. Paris, BNC 03/Cp/A//03, 4.45; b. Kricheldorf 28, 18 June 1974, 331.
43. -S Bucharest.
44. O24-Z2 ANS 1962. 170.1, 4.30.
45. O25-H3 a. ANS 1968.131.13, 4.34; b. Slocum coll. 1974, 4.45.
46. O26-H4 Kricheldorf 30, Apr. 1976, 382 = Kastner 6, 26 Nov. 1974, 430. 4.19.
47. O26-H5 Sternbeg, 28 Nov. 1975, 569, 4.45.
48. O ?-H4 Tunis, Bardo (El Djem).
49. O23-⊖3 Ratto, 26 Jan. 1955 (Giorgi), 1207 = Glending, 14 Jan. 1953, 184 = Ratto 438.
50. O24-I6 a. Berlin, Friedbaum, 4.47; b. Hess-Leu 28, 5 May 1965, 564 = Hesperia Art Bulletin 24, [1963], 82 = Münz. u. Med. 25, 17 Nov. 1962, 688.
51. O27-17 Cahn 35, 3 Nov. 1913, 578 = Cahn, FPL 24, Nov. 1912, 1938 (MIB "1338").

Group III: Throne with Curved Back

IIIa. No cross, no globe, cushions

52. C1-S3 a. Hirsch 31, 6 May 1912, 2095 = Hirsch 26, 23 May 1910, 881, 4.50; b. Bank Leu 10, 29 May 1974, 462, 4.47.

IIIb. Cross, globe, cushions

53. C2-S6 DOC 7a, 4.47 = Grierson, Coins, pl. 2, 19.
54. C3-H6 Tolstoi 141, 4.35 = Rollin and Feuardent, 20 Apr. 1896 (Montagu), 1095.
55. C4-I8 Tolstoi 143, 4.25.
56. C5-I9 a. Kyrenia Girdle no. 4, pl. 8.2; b. NFA 18, 1 Apr. 1987, 660, 4.51.

End Notes

1
Additional abbreviations used are:
Kyrenia Girdle P. Grierson, "The Kyrenia Girdle of Byzantine Medallions and Solidi," NC 15 (1955), pp. 55-70.
Longuet H. Longuet, "Die unedierten byzantinischen Munzen des Wiener Munzkabinettes," NZ 77 (1957), pp. 28-57.
The sources are conveniently summarized in A. A. Vasiliev, Justin I. An Introduction to the Epoch of Justinian the Great, DOS 1, pp. 95-96; for Justin's death, p. 414. For April 1 as the date of Justinian's installation, see J. R. Martindale, Prosopography of the Later Roman Empire 2 (Cambridge, 1980), s. v. Iustinianus 7 (pp. 645-48, esp. 647) and Iustinus 4 (pp. 648-51, esp. 650), apparently following the lead of E. Stein, RE 10, col. 1326-27, s. v. Iustinus 1.
2
The fullest treatment is that of A. R. Bellinger, "Byzantine Notes 3: The Gold of Justin I and Justinian I, ANSMN 12 (1966), pp. 90-92. See also MIB 1, pp. 44-45, and the Materialnachweise, p. 107.
3
The material assembled by Hahn is augmented in MIB 2, p. 23, and 3, p. 32.

OBVERSE DIES: SYNOPSIS AND KEY TO PLATES

Obverse dies show the two emperors (presumably Justin I on the left and Justinian I on the right) nimbate seated facing; the seat may be invisible or partly visible, or may take the form of a rectilinear or "lyre-backed throne. Unless otherwise indicated, the obverse legend is D N IVSTIN ET IVSTINIAN P P AVC. "Globe" indicates that a globe is held in the left hand of the imperial figures, the absence of a notation that the left hand is drawn up to the breast. Notes regarding the knee advanced refer to the figures l. and r. respectively: "r. and l. knees" indicates that the left figure's right knee and the right figure's left knee are advanced, "l. knees" that the left knees of both figures are advanced, and so on.

Group I: Backless Throne

Ia. No cross, no globe

O1: l. and r. knees, Pl. 1: 1-3.

Ib. No cross, globe

02: l. knees, Pl. 1:5.

Ic. Cross, no globe

O3: r. knees, Pl. 1:6.

Id. Cross, globe

O4: cushion on r. of throne, l. knees, Pl. 1: 7.
O5: l. knees, Pl. 1:8, 10, 13; Pl. 2: 22, 24, 25.
O6: cushion, l. knees, Pl. 1:9, 16; Pl. 2: 19, 28.
O7: r. knees, Pl. 1:11.
O8: D N IVSTIN ET IVSTINI P P AVC, I. knees, Pl. 1: 12; Pl. 2: 17, 23, 26.
O9: l. knees, Pl. 1:4.
O10: l. knees, Pl. 1:15.
O11: r. knees, Pl. 2: 20.
O12: I. knees, Pl. 2: 21.
O13: l. knees, Pl. 2: 27.

Group II: Throne with Straight Back

IIa. No cross, no globe, no crossbar on throne, cushions

O14: l. and r. knees, Pl. 2: 31, 35.
O15: l. and r. knees, Pl. 2: 32.
O16: l. and r. knees, Pl. 2:33.
O17: l. and r. knees, Pl. 2: 34.

IIb. Cross, globe, no crossbar on throne, cushions

O18: D N IVSTIN ET IVSTINAN P P AVC, l. knees, Pl. 3: 36.

IIc. Cross, globe, cushions

O19: l. knees, Pl. 3: 37.
O20: l. knees, Pl. 3: 38.
O21: N always two unconnected vertical strokes, l. knees, Pl. 3: 39.
O22: l. knees, Pl. 3 : 40.
O23: D N IVSTIN ET IVSTINAN P P AVC, l. knees, Pl. 3: 42, 49.
O24: l. knees, Pl. 3: 44, 50.
O25: D N IVSTIN ET IVSTINIANVS P P AVC, I. knees, Pl. 3: 45.
O26: l. knees, Pl. 3: 46, 47.
O27: l. knees, Pl. 4:51.

GROUP III: Throne with Curved Back

IIIa. No cross, no globe, cushions

C1: r. knees, Pl. 3: 52.

IIIb. Cross, globe, cushions, D N IVSTINV ET IVSTINIANVS P P AVC,

C2: l. knees, Pl. 3: 53.
C3: r. and l. knees, Pl. 3: 54.
C4: l. knees, Pl. 3: 55.
C5: l. knees, Pl. 3: 56.

THE OBVERSES

In the catalogue, the obverse dies are presented in increasing order of complexity of the type and within the major headings according to adjuncts. This preserves the essence of Hahn's system, which is based on the shape of the throne, the presence or absence of a cross, and presence or absence of a globe. These are by no means the only diagnostics that might have been chosen. Bellinger, who first drew attention to the varieties of throne type, also noted that the presentation of the imperial figures is not consistent throughout: either may have either knee advanced, and it may be the same as or opposite to that of his partner. Bellinger noticed the differences in obverse legend that are observed here without, apparently, attaching any significance to them; Hahn ignored the legends entirely. The presence or absence of a cross has elicited no detailed comment.4 After a lengthy discussion of such variations in the obverses, Bellinger concluded that they are not the idiosyncrasies of die engravers and that the divergencies are not demonstrably connected. He remarked, "We can only conclude that a number of engravers, working in a hurry, produced dies which conformed in general, to be sure, but which were not strictly controlled as to details."5 The "hurry" factor is often associated with short reigns, but except insofar as we suppose a pressure for immediate coinage at an imperial accession, Grierson is quite right to remark that short reigns are only short in retrospect, and that contemporaries could hardly have foreseen their brief duration.6

Yet within this broad context of supposed haste and consequent jumbling of attributes, some aspects of the types have been thought significant. Bellinger, for example, supported Wroth's distinction between styles 1 and 2, the former characterized by clasped hands, the latter by the left holding a globus. Bellinger remarks, "One hesitates to speak of a sequence of issues in a period of only four months but it may be suggested that the superior majesty of Type 2 was felt to be more appropriate than the superior piety of Type l."7 Hahn's discussion is brief and, in part, derivative from Bellinger's. Like Bellinger he sees chronological significance in the shift from folded arms to globus in left; and for him the variable seating arrangements (no throne/throne/lyre-backed throne) are the hallmarks of three different die engraves..

None of this will withstand examination. Hahn's own observation of the frequent linkage among obverse types, further elaborated here, disposes of any chronological significance they might be supposed to have, since no consistent pattern emerges. Note, for example, the following shared reverses:

B1: groups Ia and Ic

B2: groups Id and IIc

Γ1: groups Id and IIa

Δ1: groups Ia and Id

Z2: groups IIa and IIc

⊖1: groups Id and IIa

I1: groups Ib and Id

Dies paired with obverses of group Id are those most commonly linked with obverses of other groups, but by any measure that is the most common variety in the series and its linkage to the others would be expected to be most frequent. No single attribute or adjunct appears with sufficient consistency among the groups to be taken as a chronological signpost unless we suppose the simultaneous striking of three very different obverse styles; in that case we would be reduced to discussing nothing more than the relationship of the dies inter se, a bootless endeavor. What may be significant—though not necessarily in chronological terms—is the elaboration of the throne.

The most elaborate of the thrones that appear on the coinage of the joint reign is commonly called "lyre-backed," for obvious reasons. Until the study of Cutler in 1975,8 this form had attracted hardly any attention at all, perhaps because its representation is confined to coins. In a study of the appearance of the motif he concluded, largely from overwhelming ex silentio arguments, that there is no evidence for the existence of the lyre-backed throne as a piece of furniture, and that its significance is therefore to be sought in its symbolism. To summarize broadly (and without considering its possible orphic origins and implications), this throne may be seen as the medium by which the Logos, through intermediaries, harmonized the conflicts of the world. It originally had the connotation of wisdom, but this was eventually forgotten and the form degenerated so far as to become virtually unrecognizable.

This interpretation is certainly consistent with the coins of the joint reign, both politically and numismatically. The association of Justinian I in the rule was clearly intended to formalize his succession to the elderly emperor, who was either 75 or 77 years old at the beginning of 527. The coinage, like the events of Holy Week, simply confirmed a longstanding dynastic plan that would be the signal achievement of Justin's rule.

The view that the type is to be read only in such broad terms is reinforced by the casual admixture of other symbolic forms—globe and cross as attributes and adjuncts—as well as inattention to sometimes die-specific variants in legends meant to be perceived as formulaic, quite apart from the information they conveyed.9

End Notes

4
The cross which usually appears between the heads of the emperors on the obverse was called by Bellinger a cross potent, but if it is properly potent at all the termini of the cross are attenuated at best. This identification of the cross led A. Cutler, "The Lyre-backed Throne," Transfigurations: Studies in the Dynamics of Byzantine Iconography (Pennsylvania State University Press, University Park, 1975) to observe that the use of the cross potent during the reigns of Justin and Justinian seems to be confined to the coins; but while there are clear examples in the coinage of Justinian alone, the joint reign solidi are not the best evidence for the phenomenon.
5
Bellinger (above, n. 2), p. 92.
6
P. Grierson, "Coins mon�tairess et officines × l'epoque du Bas-Empire," SM 11 (1961), pp. 1-8, esp. p. 5.
7
Bellinger (above, n. 2), p. 91.
8
Cutler (above, n. 4), pp. 5-52, esp. 6-11 and 38-40. A. Grabar, L'empereur dans l'art byzantin (Strasbourg, 1936, rpt., London, 1971), pp. 199-200, had already noted the significance of the throne as a cult object dating back to early Roman times.
9
The omission of AN from the obverse of O8 is unique, as are the disconnected Ns of O18. It would be easier to take IVSTINAN as a simple misspelling if it did not occur twice (O16, O20). Similarly the form D N IVSTIN ET IVSTINIAN P P AVG, dominant in classes I and II, occurs a single time (C1) in class III, while that class's dominant fuller form occurs once in class II (O22). It should be recognized that such variations, which are often useful and used in cataloguing, may bear no relation to mint organization.

THE REVERSES

In contrast to the obverse, which stands out against the sixth century solidi, the reverse type continues that employed on the issues of Justin I, which was later to be carried over into the sole reign of Justinian and later still adapted by varying the ornament surmounting the cross in the angel's right hand. A quick survey of the coins of the flanking reigns of Justin I and of Justinian I seems to suggest that the reverses of the joint reign owe more to their predecessors struck under Justin than the coins of Justinian do to either in terms of style: by comparison, the coins of Justinian show a generally more squat angel whose wings do not rise as high vis × vis his head: the drapery is rendered with less care, and the placement of the feet often suggests a figure in motion rather than static.

It is possible—in fact, likely—that some reverse dies of Justin and Justinian were originally employed during the sole reign of Justin, and that some continued in use when Justinian achieved sole power in August of 527. A careful search would be tedious but would probably repay the effort. It has not been undertaken here because unless such links were to be found in numbers they would not significantly alter the picture of the coinage suggested here. What is needed is a more refined view of the internal chronology of the coinage than is possible in the present state of the evidence.

SUBDIVISION INTO OFFICINAE

The careful marking of reverse dies in the Byzantine coinage seems to attest the most rigid subdivision into officinae observed in any ancient coinage. Although the internal structure of the mint is not perfectly understood, it is at least clear that the concept of discrete, independently functioning workshops is to be avoided. This was first observed by Grierson, and has subsequently been reinforced by the occasional occurence of die linked solidi of different officinae in hoards.10

Here the instance of dies shared among officinae must surely be as high as is to be observed in any sample of Byzantine coins. Die O1 is found with reverse dies from B (1) Δ (2), and H (3a-b); O5 with B (8a-b), Δ (10a-b), S (13), ⊖ (22), and I (24a-b, 25); O6 with Γ (9a-b) and S (16); O8 with E, S, ⊖, and I (12, 17, 23a-b, 26a-c), and so on. In the present sample there is no evidence of a systematic association of pairs or triplets of officinae, and it has to be supposed that the distribution of dies among the workshops reflects their drawing upon a common pool of dies (a die box) for the production of reverses. It is of course equally plausible that the obverses, too, came from die boxes, since there is nothing which would suggest their attachment to specific officinae, and indeed the likelihood of this construct grows as obverses begin to appear linked with dies of three or more workshops.

In spite of the small size of the sample, it may be worthy of note that there is no certain instance of recutting of dies to be observed, either within or among officinae and despite the existence of die image3 with the number reversed.11

THE MINT(S)

Heretofore it has never been supposed that these coins were struck anywhere but Constantinople. There is clear evidence for the existence of a mint for gold at Thessalonica for the emperors from Anastasius I onwards—sometimes gold that would on the basis of its markings have been taken to be Constantinopolitan—but no one to date has hinted at the existence of a non-metropolitan mint for the brief joint reign.12

In fact if Cutler's analysis is credited and the throne is stripped of its pictorial literalism, and if this is combined with the discreteness of the lyre-backed group, it is possible, at least for now, to suggest the existence of a mint at Thessalonica. For none of the lyre-backed solidi display reverse die links to throneless or square-backed ones, while within those two groups die linkage is common.

It is dangerous to be dogmatic in enumerating stylistic differences on the basis of as few as seven specimens from five obverse dies but, for what it is worth, the obverses of group III generally display taller, more vertically oriented figures seated on the throne, which is itself rendered with more care and consistency than those shown on the coins with straight back of group II. The angel on the reverse maintains the same posture throughout, and his right wrist leading up to the cross is impossibly long. The star in the field is uniformly large, while some variation in size is to be observed on coins of the other two groups. Against the use of style as a dissociative criterion are certain features which, if not identical, at least display the same kind of variation as has been noted in groups I and II: for example, on obverse die C1 the outside knees of both emperors are advanced, while on all the other dies the right ones are forward; C1 also lacks the cross in field which is present on all other dies. Similarly C1 is without the globe, which is present elsewhere, and it has short form of obverse legend which is characteristic of groups I and II. I have not been able to examine the coins struck from this die, but just possibly the lyre-backed throne has been recut from a rectilinear one. If so, the hypothesis of a second mint is obviously weakened.

End Notes

10
P. Grierson (above, n. 6), p. 4. For die links among officinae see W. E. Metcalf, "Three Byzantine Gold Hoards," ANSMN 25 (1980), pp. 87-108, especially hoard 2, nos. 3 and 7, 4 and 6, 26 and 36, 34 and 39, 50 and 54, all pairing coins of different workshops; C. Morrisson, "Le trésor byzantin de Nikertai," RBN 118 (1972), pp. 29-91, esp. 40-43 and fig. 5.
11
But see below, p. 26.

CONCLUSION

At the very least the picture of the coinage of Justin I and Justinian I is more complicated than expected. Grierson had already pointed to the fairly frequent linkage among officinae, which is more common than the existence of a system for their numeration would imply; his view is confirmed here, with its implications extended into the sixth century. Similarly minor variations in legend and in distribution of attributes and adjuncts cannot be seen here—as they are sometimes seen elsewhere—as the trademarks of engravers or as secret marks that somehow, if decoded, might elucidate the working routine of the metropolitan mint.

Allowance must be made for the "hurry factor"; even if "short reigns are only short in retrospect," the illness of Justin gave more than the usual urgency to a proclamation coinage, and this might be reflected in unusual steps taken to hasten its production. Nonetheless, the coinage presents a surprisingly disorganized aspect which, if it may be generalized to the sixth century coinage at large, reduces the reliability of some time-honored simple solutions to complex problems, and invites the painstaking analysis, die by die, of a forbidding mass of material.

APPENDIX

The Approximate Size of the Coinage

A die study of this sort provides the opportunity to estimate the numer of dies used for this coinage from the number represented in the surviving sample. Any number of methods have been devised for this calculation, but the formula applied here is that of Carter,

image

where n = the number of coins and d = the number of dies observed and n < 2d or

image

where n = 2 to 3d.13

Six specimens (4, 18, 29-30, 41, and 43) have been omitted from the calculations since no photographs are available, leaving a total of 73 pieces from 37 reverses dies. For the obverses, two calculations must be made because of uncertainty regarding the identity of the obverse of coin 48: the first assumes that it is identical to another known obverse, the second assumes that it is not, so d1 = 32 and d2 = 33 obverses. Carter's formulae yield the following results: 1) projects an estimate of 47 × 4 obverse dies, 2) 50 × 5 obverse dies, while both yield an estimate of 61 × 7 reverse dies. Since these two calculations embrace the whole coinage, including two bodies of material that are immiscible (groups I and II do not and, in the case of the obverses, cannot share dies with group III) it is perhaps sounder methodologically to provide a separate calculation for groups I and II only: thus from 65 coins with 1) 27 or 2) 28 obverse dies, and 32 reverse dies, the results are as follows: 1) 38 × 4, 2) 41 × 4 obverses, and 51 × 6 reverses. Adding these projections to the actual number of dies observed for group III (which is too small to apply the method meaningfully) produces satisfactory consistency. An estimate of the total original die population at ca. 50 obverses and 60 reverses will not be far from the truth.

The joint reign lasted only 17 weeks, and there is no reason to suppose that the coinage began before it or continued after it. Even in this period, which might well have been one of heavy coinage in view of the imperial accession, somewhat less than an obverse a day was used if the work week consisted of five days.

End Notes

12
Prof. P. Grierson, who generously made available to me photographs accumulated over many years, had segregated the coins of group III known to him, and marked them with "Antioch (?)."
13
See G. F. Carter, "A Simplified Method for Calculating the Original Number of Dies from Die Link Statistics," ANSMN 28 (1983), pp. 195-206, esp. 201-2. The attraction of the Carter method is the simplicity of its application, especially where high precision does not yet seem attainable. For evaluation of this and other methods currently employed in estimating size of coinage, see W. W. Esty, "Estimation of the Size of a Coinage: A Survey and Comparison of Methods," NC 146 (1986), pp. 185-215.

THE MONTE JUDICA HOARD AND THE SICILIAN MONETA AURI UNDER Justinian I AND JUSTIN II

Niall Fairhead and Wolfgang Hahn

Plates 4-8

Some ten years ago on Monte Judica in the province Catania, Sicily, a hoard of sixth-century gold coins appeared. This hoard is essential for understanding the activities of the smaller Byzantine gold mints in the western half of the empire and is comparable only to the 1948 Thessalonica hoard's materials for mint attributions in the East. The coins had been concealed in an unglazed terracotta pot which, regrettably, has since been lost. The hoard was immediately split up, and some of the coins were dispersed in trade via Switzerland.1 We are are indebted to S. Bendali and H. Berk for certain information about the coins which has led to the following reconstruction of the hoard.

CATALOGUE

Anastasius I (491-518)

Constantinople, MIB 7, 507-18, solidus

1. Z 4.49 Munz. u. Med. FPL 434, June 1981, 19

Justinian I (527-65)

Constantinople, MIB 5, 527-37, solidus

2. Γ 4.46
3. Δ 4.50
4. ϵ 4.48
5. ϵ 4.45
6. ϵ
7. S 4.46
8. S 4.46
9. Z 4.48
10. Z 4.44
11. Z
12. 4.38
13. I 4.48
14. I 4.47
15. I 4.46
16. I

MIB 6, 537-42, solidus

17. A 4.47
18. A 4.46
19. A
20. A
21. B 4.46
22. Γ 4.47
23. S
24. Z
25. H 4.41
26. 4.49 = Schweizer. Kredit. 2, 27-28 Apr. 1984, 740
27.
28. I

MIB 7, 542-65, solidus

29. A 4.46
30. A
31. A
32. A
33. A
34. B 4.52
35. B 4.49
36. B 4.43
37. Γ 4.49
38. Γ 4.49
39. Γ 4.46
40. Γ
41. Δ 4.48
42. ϵ 4.52
43. ϵ 4.46
44. S 4.48
45. S 4.48
46. S 4.47
47. image 4.47
48. image 4.47
49. Rev damaged. H 4.45
50. 4.47
51. 4.49
52. I 4.47
53. I 4.48
54. IB 4.46
55. B or ⊖ 4.48

MIB 19, 527-65, tremissis

56. 1.46
57.

Thessalonica, MIB 23, 562-65, solidus

58.Obv. of Thes. hd. 23. 4.46

Carthage, MIB 25, 547/8, solidus

59. IA 4.43 = Schweizer. Kredit. 2, 27-28 Apr. 1984, 741
60. Obv. of DOC 277 a. 1. IA 4.49

Rome, MIB 31-32, 540-42, solidus

61. 4.44 = Schweizer. Kredit. 2, 27-28 Apr. 1984, 742

MIB, 34, 542-49, solidus

62. A 4.47
63. Rev., no star. A 4.45
64. Γ 4.47 = Schweizer. Kredit. 2, 27-28 Apr. 1984, 743
65. Rev. of BM and Birmingham specimens. Z 4.43
66. Rev. of Ratto 1955, 1209 4.40 = Schweizer. Kredit. 2, 27-28 Apr. 1984, 744

Ravenna, MIB 37, 549-65, solidus

67. Dies of Ratto 459 Γ 4.45 = Schweizer. Kredit. 2, 27-28 Apr. 1984, 745
68. Γ 4.48
69. Rev. legend error S 4.44 = Schweizer. Kredit. 2, 27-28 Apr. 1984, 746
70. Obv. of Ratto 461 г

Sicily, MIB 37 [Ravenna], 554/5?, solidus

71. Г 4.47

555/6 ?, solidus

72. Δ
73. Obv. of 72. Δ
74. Δ

MIB 41 [Ravenna], 542-65, tremissis

75.
End Notes
1
Münz. u. Med.'s FPL 434, June 1981, contained some coins of this hoard (as indicated in the catalogue) and Schweizerische Kreditanstalt 2, April 1984, included others. The references given in the latter were taken from an older version of this article and in the course of refinement, the numbers were changed. The discovery date of the hoard, claimed to be 1981 by some informants, could not be verified.

Justin II (565-78)

Constantinople, MIB 5, 567-78, solidus

76. A 4.44
77. A
78. Є 4.51
79. H
80. I 4.49

MIB 11a, 565-78, tremissis

81.

Thessalonica, MIB 16, ca. 570, solidus

82. 4.48 = Münz. u. Med. FPL 434, June 1981, 21

Ravenna, MIB 20c, 567-ca. 570, solidus

83. Obv. of Thes. hd. 110. image 4.44

MIB 21, ca. 570, solidus

84. P 4.47 = Schweizer. Kredit. 2, 27-28 Apr. 1984, 747

Sicily, MIB 21 [Ravenna], 568/9, solidus

85. Rev. CNONB. B 4.50 = Schweizer. Kredit. 2, 27-28 Apr. 1984, 748
86. Dies of 85. B 4.45

569/70

87. Γ = Münz. u. Med. FPL 434, June 1981, 20

570/1

88. Obv. of 85. Δ 4.45
89. Dies of 88. Δ 4.45
90. Δ 4.48 = Schweizer. Kredit. 2, 27-28 Apr. 1984, 749
91. Obv. of 90. Δ 4.46
92. Dies of 91. Δ 4.46 = Schweizer. Kredit. 2, 27-28 Apr. 1984, 750

571/2

93. Obv. of 90. Є 4.46

MIB 24 [Ravenna] ca. 570, semissis

94. Rev. VITC... 2.24

Tremissis

95. Rev. Victory facing 1. 1.50 = Schweizer. Kredit. 2, 27-28 Apr. 1984, 751
96. Obv. of 95. Rev. Victory facing r. 1.52 = Münz u. Med. FPL 434, June 1981, 22
97. 1.51 = Schweizer. Kredit. 2, 27-28 Apr. 1984, 752
98. 1.47

It seems probable that the hoard has been reconstructed as fully as possible. The 98 coins are distributed as follows.

Con. The. Car. Rom. Rav. Sic.
Anastasius I, 1 Sol. 1 - - - - -
Justinian I, 74 Sol. 54 1 2 6 4 4
Sem. - - - - -
Trem. 2 - - - - 1
Justin II, 23 Sol. 5 1 - 2 9
Sem. - - - - - 1
Trem. 1 - - - - 4
98 63 2 2 6 6 19

As can be seen from this summary there is a Sicilian gold coinage which can be traced back to Justinian I and indeed (as shall be shown) almost to the time of the reconquest of the island by the imperial forces. During the course of the Ostrogothic war Belisarius captured Sicily after a six month campaign from July to December 535 (Procop. Goth 5.5) and a Byzantine administration was installed in 537 (Novels 104). Sicily was given special status as it was not subject to one of the neighboring prefectures, but was governed by a praetor under the direct control of the quaestor sacri palatii in Constantinople: the province was administered quasi peculiare aliquid commodum imperatoribus. Under these extraordinary circumstances it appears that the monetary supply of the island was provided either directly by Constantinople or by an exceptional branch mint under Constantinople's control, and for the seventh century there is evidence for both practices.2

For the sixth century, however, the identification of Sicilian coins has been slow until now. The only clues have been stylistic comparisons and provenances, since the gold coins in question do not betray their non-Constantinopolitan origin by any kind of signature. Even the Sicilian coppers do not have their own mint marks until the 580s. Before the Monte Judica hoard appeared, the starting point of the Sicilian series was the partly mint marked copper coinage of Maurice. The most recent contributions to the subject sought to establish Sicilian gold attributions for as early as Tiberius II.3 Now we can move back the dating into the reigns of the two preceding emperors, Justinian I and Justin II, by reference to a number of unusual coins in this hoard which are of a fabric previously unrecognized or misattributed. The recognition of this fabric is made more difficult by its similarity to coins of Ravenna fabric of the later sixth century. The Monte Judica hoard, then, provides welcome corrections to the distinctions between Sicilian and Ravenna gold coinage. It has implications also for the copper, for which there are similar difficulties in the earlier part of the period of Byzantine domination.

The composition of the Monte Judica hoard is characteristic of an accumulation in a region where a provincial mint with limited output was operating. Only among the coins of the last years before the hoarding was completed do the local issues outnumber those of Constantinople,4 and die links occur only among the provincial issues of the local mint. The chronological pattern is normal and coins issued in the years not long before the date of deposit are represented in larger numbers in the hoard. In addition to Constantinopolitan coins the hoard contains examples from almost all the provincial mints. Their numbers reflect their proximity to Sicily: Thessalonica and Carthage are represented by only two coins each, whereas the Italian mints, first Rome and later Ravenna, are represented by 12 pieces.

The attribution of the non-Sicilian provincial gold of Justinian I and Justin II has been firmly established by modern numismatic research. Nevertheless one expects new material to bring forward novelties. The Roman solidus 63 is remarkable for its lack of a star in the right reverse field. Possibly this is no more than an engraver's error, but the same omission on a tremissis of approximately the same date, MIB 2, pl. 38, N36, is perhaps no coincidence.

The most important contribution of the Monte Judica hoard is the group of 19 unusual coins of Justinian I, 71-75, and Justin II, 85-98. To show the Sicilian origin of the group one has to argue from the latest part of the hoard, where the local coins have their strongest representation. There are nine solidi of Justin II, 85-93, having a fabric not known before and variously die linked among themselves. Certain features, especially the bulging rim, recall the Ravenna style; but, on closer examination, the emperor's bust is somewhat different from that of the well known Ravenna pieces (compare with 83 and 84). Moreover the terminal letter of the reverse legend, which gives a sequence of B, Γ, Δ, and Є on the new coins, is not otherwise attested on the Ravenna solidi of Justin II, apart from the problematical MIB 3, V20 with Δ, discussed below. There can be little doubt that these letters stand for dates, for the local mint was certainly too small to consist of five officinae. Furthermore the preponderance of Δ coins, the last year but one, is typical of a chronological sequence, while the last year, Є, is less heavily represented. The die links among solidi 85-93 indicate a high survival rate from almost certainly a very small number of dies and an obverse die is carried from one year to the next in two cases, while another skips year three.

image

Obverse Dies

Reverse Dies

The question of what kind of dates are concerned is easily resolved in favor of indictional years, because the solidi of the first two regnal years of Justin II should have had the star of the old indiction which ended in 567. This starred first emission of Justin II, known from Constantinople (MIB 1), is not found in the hoard, and the Constantinople portion drops during the later years (there are only five solidi of Justin II from the metropolitan mint). Using indictional dates, the Sicilian solidi 85-93 can be assigned to years 568-72; the latest coin in the hoard therefore was minted before 1 September 572. The appearence of dated solidi in the early years of the 567-78 indiction has parallels, incidentally, at Carthage, and there were also experiments made with dated solidi in Constantinople.5

In addition to the solidi, the hoard also has fractional coins of Justin II belonging to the local mint. There are four tremisses (95-98) and one semissis (94) of Justin II as well as one tremissis of Justinian I (75), of a fabric that was already known and that had been attributed to Ravenna (Justinian I, MIB 40.1 and 41.1, and Justin II, MIB 24), where two stylistic groups were recognized. This was explained by postulating two engravers. Now the problem is resolved by the recognition of two different mints, Ravenna and Sicily (See "Comparative Material," Plates 7 and 8). The stylistic relation between them is similar to that seen in the solidi: they share the bulging rim, but the obverse bust has some differences in its appearence, especially in the drawing of the folds of the paludamentum on the emperor's chest. Furthermore the Ravenna star is a little larger and often six-rayed; on the semisses it is missing altogether. As well as the coins in the Monte Judica hoard there are several other specimens which also belong to this newly recognized Sicilian fabric: there are four more semisses and one tremissis of Justin II.

image

Sicily

image

Ravenna

Semisses

128. Rev. Δ. MIB 3, pl. 56, 24b = Glendining, 9 Mar. 1931, 419
129. Rev. of 128. Milan Municipal coll.

Tremissis

130. Rev. star 1. Syracuse Archaeological Museum

Semisses

131. MIB 24 = BMC 290, 2.20
132. Lepczyk, 1-2 Apr. 1980, 1320 = Hess 249, 13 Nov. 1979, 461, 2.22

The semisses 128-29 are of special interest, because they have the letter Δ at the end of the reverse legend. This could be a date corresponding with that on solidi 88-92; but one wonders whether it is because the semisses of the other years were in fact not marked with a date, and other expected varieties are not in the hoard. It is tempting to dismiss the interpretation of Δ as year 4, and instead to connect the semisses 128-29 with another problematic coin, the solidus MIB 3, pl. 56, V20, which is recorded there together with semissis 128 (MIB 24b) under Ravenna. Of this solidus there are two specimens.

126. Rev. star 1., Peus 311, Oct./Nov. 1984, 833 = Lanz 16, Apr. 1979, 692, 4.34
127. Rev. of 126. Viennese private coll., 4.47

They have the star of the old indiction in the left of the reverse field and can therefore be dated 565-67. Although the Δ is apparently meaningless, it may have been copied (immobilized) from earlier Sicilian solidi, see 72-74 of Justinian I. Stylistically 126 and 127 stand between these and the Sicilian solidi of Justin II, 88-92. A further clue to the Sicilian attribution of the star issue is perhaps to be found in the tremissis 130 which has the star also to the left, as a consequence of which Victory has been shown with the globus cruciger and wreath reversed. This seems not to be an engraver's error, because the head looks left as on tremissis 95 in the Monte Judica hoard. As the obverse of 130 is very similar to that of the Δ semisses 128-29, it probably completes the series of denominations in gold for 565-67.

The reattribution of MIB V20 to Sicily makes the recent suggestions as to the meaning of the terminal letters Z and P on the Ravenna solidi6 untenable, especially since the presence of both in this hoard, 83-84, dates the change from Z to P before 572. The earlier explanation of these letters as pseudo-officinae7 remains preferable for Ravenna.

Returning to Justinian I, the Sicilian group within the hoard is represented by four solidi and one tremissis. Tremissis 76 offers no problems, being closely related in style to its successors under Justin II. More of Justinian's tremisses and semisses of this fabric are known from several sources and need to be removed from the Ravenna series MIB 39, 40.1, and 41.1 leaving only 40.2 and 41.2 for Ravenna.

Semisses

104. Rev. image (before 552). MIB 39 = BMCVandals 42, pl. 16, 13, 2.12
105. Rev. image. Tolstoi 45, 2.25
106. Rev. image (after 552). Basel Münzhdlg. 6, 18 March 1936, 2132; 2.24
107. Rev. image. Milan Municipal coll.
108. Rev. image. MIB 40.1 = Bucharest Akademy coll.

Tremisses

109. Peus 305, 12-15 Oct. 1982, 340 = Dorotheum 228, 26-29 Jan. 1960, 3, 1.40
110. MIB 41.1 = BMCVandals 43, pl. 16, 14, 1.47
111. Dorotheum, 8-9 June 1956 (Zeno), 2463, 1.45
112. Dorotheum, 8-9 June 1956 (Zeno), 2462, 1.45
113. Vinchon, 25-27 Apr. 1960, 683
114. Dies of 113. Udine Municipal coll.
115. Dies of 113. BNC 4/Cp/A//31, 1.46
116. Obv. of 113. BMCVandals 45, pl. 16, 15, 1.47
117. Bologna Archaeological Museum.
118. Rev. of 111. Birmingham University
119. Glasgow University
120. Dies of 119. ANS
121. Obv. of 119. Glendining, Dec. 1974, 331
122. Tremissis of Justinian I, Tolstoi 536, 1.45; 122a. DOC 19.2, 1.49; 122b. Cambridge = P. Grierson, Medieval European Coinage (Cambridge, Eng., 1986), 294, 1.41

The normal chrismon on the reverse of 104-5 (MIB 39) enables us to date the beginning of this mintage before 552, when it was reversed to image (MIB 40.1 and Justin II, MIB 24b).8

The Sicilian solidi of Justinian I in the Monte Judica hoard (71-74) are not as similar as is the tremissis to the Justin II pieces; we cannot follow the stylistic development so closely, either because the production of solidi was not continuous or because we merely do not know the connecting links. The differences from the Ravenna solidi (see 67-70) are likewise to be found in the detail of the emperor's bust, the heads being somewhat smaller and neater in Sicily. The trifolium on the crown has different forms: image in Sicily, image in Ravenna. Apart from the Monte Judica hoard we know of only two other solidi in the Sicilian fabric.

99. Rev. Δ. Berlin = Restle, pl. 5, Beyer Ak. Pal. 403; 4.279
100. Rev. A⊖. MIB 2, pl. 38, N37 (attributed to Ravenna)

This second specimen is undoubtedly of the same fabric as 67-70 of the Monte Judica hoard, but it has two terminal letters in the reverse legend, A⊖, where the others have Γ (67) or Δ (68-70), and the star is in the left reverse field instead of its normal position on the right. The sequence and the dating of these three varieties depend on the interpretation of the letters Γ, Δ, and A⊖. The succeeding issue, 126-27 of Justin II, has the Δ as well as the star left of the A⊖ issue. The solution might be sought from the copper coins.

As mentioned before, the newly identified Sicilian gold coinage of Justinian I and Justin II sheds fresh light on the hitherto very uncertain attributions of Sicilian copper. There are some curious folles signed ANNO XXX Δ and CON, but in a different style from that of Constantinople. A Sicilian origin was proposed by Hahn.10 We can now compare them with the Sicilian solidi and recognize the same bust and trifolium. Three specimens of Justinian's ANNO XXX Δ folles are known.

101. Rev. wreathed border. Naples National Museum (Santangelo coll.)
102. Dies of 101. Leningrad 510, 14.17
103. Rev. plain border. Leningrad 511, 16.91

If the year 30 is a real one (which does not seem improbable) the Δ could be an indictional year, as the signature of a fourth officina would be only a pseudo-officina letter to fill the space. A fourth indictional year ended on 31 August 556, and the thirteenth regnal year began on 1 April 556. The Δ on the solidi could refer to the same indictional year; the Γ would then be the year before (554/5), but what is A⊖? It is even uncertain whether we should read ⊖ or a vertical Φ; AΦ makes no obvious sense,11 while ⊖ could be the date, elucidated by A as year 9, or perhaps the date is expressed by the A, equalling 1, and the ⊖ is a separate administrative (secret) mark as occurs so often in the seventh century.12 The first year of this indiction was 552/3, and the ninth year was 560/1. Both are possible and the question cannot be resolved on the basis of current knowledge.

There is another coin in the hoard displaying two letters at the end of the reverse inscription, 54 with IB, well known from a number of other specimens and previously thought to be a product of a twelfth officina in the Constantinople mint.13 This explanation was based on the assumption that the opening of the Carthaginian gold mint in about 540 was done by two newly created officinae of the metropolitan mint which applied the marks IA and IB. These solidi stylistically developed into the locally identifiable Carthaginian issues similar to the standard type struck between 537 and 542 in Constantinople. After a while the twelfth officina was thought to have been returned to Constantinople and to have been active striking solidi of the type introduced in 542 until its dissolution somewhat later. This hypothesis is apparently invalid, as C. Morrisson has argued convincingly that the IA and IB of the Carthaginian solidi are indictional years.14 The IB marked coins of MIB 7 are certainly of Constantinopolitan fabric, and if they belong to a twelfth officina, we have no sign of an eleventh which may have been the silver officina not marking its products.

image

The situation is complicated even more by the recent appearance of a single solidus marked IЄ at the end of the reverse legend and of undeniably Constantinopolitan style. On this coin the Є was probably added to an already finished die of the tenth officina because there was no space left after the I and the Є, therefore, had to be cut over the exergual line. It is difficult to explain the combinations IB and IЄ in the metropolitan mint. J. P. C. Kent has suggested cooperation between two officinae in the somewhat similar case of the 22 carat, light weight solidi with ⊖S,15 but the question remains, why or under which conditions would they have cooperated. An early experiment to introduce indictional dating on the regular Constantinopolitan solidi is improbable because of the gap between the years 12 and 15. The assumption of indictional years would perhaps be more plausible, if the coins were struck in Constantinople but intended for a western destination, e.g. for Sicily. In this case the IB would stand for 548/9, because IЄ can only be 551/2 (the fifteenth year of the next indiction being later than the end of Justinian's reign). Although there is no further evidence for a Sicilian connection for these coins apart from the single example in the Monte Judica hoard,16 it is tempting to see them as a preliminary stage in the development of Sicilian gold coinage. The years 12 (548/9) and 15 (551/2) would have been provided by Constantinople, whereas the later years 3 (554/5), 4 (555/6), and 9 (560/1) or 1 (552/3) could have been struck in Sicily itself. The use of indictional datings on coins which might have been produced in the Constantinople mint for circulation in the West is also notable on the copper issue of 553/4 (MIB 98).

The wreath border on the reverse of follis 101 leads us to the question of the hitherto very uncertain origin of the numerous western 10- and 5-nummi pieces with the Latin value mark X (MIB 244) and V (MIB 246). Hahn proposed a Sicilian mint for them17 despite the absence of any solid evidence. Their drawing of the emperor's profile bust is often somewhat crude, but there are pieces (123-25) which are stylistically comparable to the fractional gold coins. The comparison seems to strengthen the Sicilian attribution of the coppers, especially the delineation of the hair image. The Monte Judica hoard thus offers more evidence for the production of copper coins in Sicily under Justinian I. In its later development, i.e. in the later years of Justinian I and under Justin II, further stylistic comparisions do not help, because the dies of the copper coins tend to become too crude for similarities to be detected. More hoards providing such evidence are needed to solve the remaining problems.

Due to the indictional dates on the Sicilian solidi of Justin II, the concealment of the Monte Judica hoard must have been in 572 or shortly afterward. We know of no threat to Sicily in these years. The last incursion had been made 20 years earlier by the Totilas Goths, while the Lombards had not yet reached the straits of Messina.

The value of the hoard amounts to 93 1/8 solidi, i.e. about 1 1/8 pounds of gold, almost exactly corresponding to that of Thessaloniki hoard of 1948 which amounted to 93 1/8 solidi. Compared with the payments and rents mentioned in the letters of Pope Gregory I (590-604) referring to Sicily18 this was a small fortune. Is it possible that there was a Jewish community there as there were in several other Sicilian cities of this time?19 Such communities were often harrassed by church authorities.20 In addition, 572, the probable year the hoard was closed, was the end of a lustrum, i. e. the end of a taxation period.

image

Number of Coins in the Monte Judica Hoard Divided by Possible Years of Issue

The age structure of the hoard shows a range of about 45 years, omitting the isolated specimen of Anastasius I. Divided into applicable periods21 and converted into yearly indices the composition curve has only a slight anomaly due to the overrepresentation of Justinian I's lustral issue for 537-42 (MIB 6, Constantinople). This might be explained by the course of events during the Italian war. The newly appointed praefecius praelorio per Haliam, Maximus, coming from Constantinople in 542 was stuck in Syracuse for a while (Procop. Goth. 7.7.1-3), and it is uncertain whether he reached Rome at all.

There is still only fragmentary understanding of the history of the Sicilian gold mintage in the decades following the burial of the Monte Judica hoard, i.e. after 572 and before the reorganization of the monetary supply under Heraclius and Constans II. What is known from the time of Tiberius II onwards is summarized in two articles by W. R. O. Hahn.22 A single tremissis of Tiberius II has survived (MIB 3, pl. 56, N14). There are five tremisses of Maurice (MIB 26) and, late in his reign, six solidi (MIB 28) and one light weight solidus of 20 carats (M1B 29). Both rulers resumed dating by indictional years. Phocas follows with similar solidi (MIB 97 and 98, three specimens recorded), semisses (three specimens),23 and tremisses (MIB 99 and 100, eleven specimens recorded). Under Heraclius the pattern is completed by a semissis (MIB V99, one specimen recorded) together with solidi (MIB 97 and 98, three specimens recorded) and tremisses (MIB 99 and 100, eleven specimens recorded). Then there seems to have been a break in the coinage, perhaps sometime in the early twenties. When coining resumed in the thirties,24 the coins were of a different fabric which developed into the characteristic Sicilian style well known from the time of Constans II onward.

During the entire period between Justin II and Heraclius Sicilian gold mintage was restricted to a small output apparently meeting only a part of the demands of the administration. This is obvious from the frequent die linkage and the small numbers of surviving specimens as well as from the predominantly Constantinopolitan composition of the few Sicilian gold hoards up to the middle of the seventh century.25 The Monte Judica hoard is somewhat exceptional insofar as it contains a relatively large number of local coins; but it is the only sixth century hoard recorded from Sicily, and only new discoveries can enhance our understanding of this period.

End Notes
2
MIB 3, p. 72.
3
W. Mahn, "More about the Minor Byzantine Gold Mints from Tiberius II to Heraclius," NCirc 87 (1979), pp. 552-55, and MIB 3, p. 71.
4
See the Thessalonica hoard of 1948 as revised by W. Hahn, "New Light on the Thessalonican Moneta Auri in the Second Half of the Sixth Century," NC 141 (1981), pp. 178-82.
5
Carthage MIB 3, p. 41; Constantinople MIB 2, p. 38.
6
MIB 3, pp. 66-67.
7
MIB 2, p. 41.
8
For an explanation of the use of the secret mark, see MIB 1, p. 17, and 3, p. 25.
9
Marcell Restle, Kunst und Byzantinische Münzpr�gung von Justinian I (Athens, 1964).
10
MIB 3, p. 55.
11
There is no stylistic resemblance to the exceptional Carthaginian solidus MIB 26 with ΛΦP instead of CONOB.
12
See MIB 3, p. 129. The form of the ⊖ with a prolonged horizontal bar occurs also on Roman coins of Justinian I (See MIB 34.8).
13
MIB 1, p. 48.
14
See her article in this volume.
15
J. P. C. Kent, review of Adelson, NC 19 (1959), p. 239.
16
Other provenances known are Albania (DOC 278 with incorrect attribution to Carthage) and Peloponnese. C. Morisson (below, p. 46) prefers a tentative attribution of these coins to a provisional mint in the Balkans.
17
MIB 1, p. 75, but see MIB 2, p. 49.
18
A. Holm, Geschichte Siciliens in Alterthum 3 (Leipzig, 1898), pp. 292-309.
19
B. Pace, Arte e civilt� della Sicilia antica 4 (Rome, 1949), pp. 137-38.
20
Holm (above, n. 18), pp. 310-11.
21
Tremisses 56 and 57 are included with the later solidi of Justinian I because of their flat relief, see MIB 1, p. 50.
22
W. Hahn, "Some Unusual Gold Coins of Heraclius and Their Mint Attribution," NCirc 85 (1977), pp. 536-39, and "More about the Minor Byzantine Gold Mints from Tiberius to Heraclius," NCirc 87 (1979), pp. 552-55.
23
Peus 308, 19-21 Oct. 1983, 593; NCirc 92 (1984), p. 257, 7-8.

END NOTE

Additional References for "Comparative Material"

123. BMC Vandals, pl. 9,8
124. Tolstoi 482
125. BNC 4/RO/�/23
133. Naples
134. Rome, National Museum
135. Schweizer. Kredit. FPL, Dec . 1984 , 7
136. Milan
137. Brussels
138. Milan
139. Copenhagen
140. Berlin
141. Münz. u. Med. 43, 12-13 Nov. 1970, 545

End Notes

24
The earliest coins of this group are dated 631/2.
25
W. Hahn (above, n. 22), "More about the Minor...Mints," p. 554.

CARTHAGE : THE MONETA AURI UNDER Justinian I AND JUSTIN II, 537-578

C�cile Morrisson

Plates 9-14

As long ago as 1956, J. P. C. Kent 1 analyzed the concentration of the striking of gold, following the Valentinianic reforms of 368,2 in the care of the comitatus and the Palatine officium or, insofar as the East was concerned, essentially at Constantinople. Nonetheless the existence of a number of issues from Theodosius II through Basiliscus,3 doubtless struck at Thessalonica, as well as the Byzantine series known from Ravenna in the sixth century, led him to postulate an association between the mintage of gold and the prefectures of Illyricum and of Italy. It was thus logical to conclude that, like Ravenna, Carthage, as the seat of the prefecture of Africa, would probably have received a Palatine detachment of the officium of the comes sacrum largitionum for that purpose after its reconquest under Justinian.4

The correctness of this hypothesis was confirmed some years later, after the acquisition by the Cabinet des Médailles of a solidus with the mark ΛΦP(ıkη) on the reverse. J. Lafaurie identified, on the basis of this exceptional exergue, a series of specimens of the same style with the normal marks CONOB and IA, IB, or IΓ at the end of the reverse legend, which had previously been confused with issues of Constantinople.5 The North African origin of certain pieces further supported this attribution,6 which was immediately accepted and supplemented in DOC (1966) by the identification, on stylistic grounds, of solidi with Γ7 and ⊖.8

The criteria adduced for this distinction are rarely defined with precision, and for good reason. Under Justinian, contrary to what one observes under Justin II and Tiberius II, there is no particular iconographic difference between the coinage of the province and that of the capital. Most of the solidi have on the obverse three pellets on the pectoral band of the cuirass, a wavy line under it, and an undulation embellishing the left vertical band, while the angel on the reverse is represented in a similar way except for a general elongation more marked in the figure at Constantinople.

The real originality of the African solidi must be sought in their manufacture. As Lafaurie saw, their diameter is slightly smaller; in addition, at Carthage the border of the die generally coincides with that of the flan, while at Constantinople the flan is often larger than the die.9 The line which forms this border is somewhat flatter and larger by some millimeters at Constantinople, while in Africa it is narrower and more pronounced. This is only one aspect of the more heavily accented relief which, together with the design of the eyes (often highlighted by a stroke beneath) constitutes the major characteristic of the gold coinage of Carthage. Nevertheless two distinctive styles must be differentiated within the mint. One, which I would characterize as "Constantinopolitan," is only distinguished from that of the capital by the singular workmanship mentioned above. The other I would call "African," and it is rougher and "stereotyped," so to speak, and of cruder engraving. It displays on the reverse a smaller star and angel10 whose cross (at least on the coins with Γ) encloses in its arms the I of VICTORIA, now reduced to a tiny stroke. On the obverse the last letter of the legend nearly disappears, a C reduced to a line so near to that defining the shield that it might almost be confused with it.

The Constantinopolitan style is heavily predominant. Virtually all the issues with I, IA, and IB share it (with one exception each for the two latter), as do two-thirds of those with Γ. The African style is represented by the remaining third with Γ, the three specimens with ⊖, and the IA/ΛΦP solidus. In its broad outline, this distinction did not escape Hahn, who saw an evolution over time leading from the first solidi, "quite naturally very similar to the Constantinopolitan style," to that group of peculiar style whose development might be attributed either to "comparative isolation from Constantinople or equally to the recruitment of local engravers."11 Yet in spite of the very significant place he assigned throughout his work to dating by regnal year or indiction, he continued, with Lafaurie and Hendy,12 to regard the letters on the reverses as an indication of supplementary officinae (IA and IB) created for this purpose, or of detachments from the third and tenth officinae of Constantinople (Γ and I). Consideration of the piece in Copenhagen with ⊖, however, led him to ask whether such a series of officinae was not too numerous for a provincial mint, only at once to underscore the difficulties of an indictional interpretation.13 We hope here, starting from the analysis of as great a body of documentation as possible (89 examples from Tunisian, American, and European collections and from sale catalogues) to present a solution which at once takes into account the elements of the numismatic documents and the historical events that may have influenced monetary output.

CATALOGUE

Justinian I

Obv. DNIVSTINI ANVSPPAVI. Bust facing with helmet and cuirass; in 1. shield, in r. globus cruciger.

Rev. VICTORI AAVCCC and letter. Angel standing facing; in r. long cross, in l. globus cruciger; in r. field, cross; in exergue, CONOB.

MIB 26, BNC 01, DOC 279: IA, in exergue ΛФp.

African manner. Two pellets in top of cuirass.

  • 1. BNC 01 (from Tunisia, 1960), 4.36 g. The I in VICTORI above 1. has disappeared either through being confused with the cross bar or because there was not enough room in the angle of the cross.

MIB 25, BNC 05, DOC 277: Γ

a) African manner. Two pellets in top of cuirass image Rev. above r., I reduced to short line wedged in angle of cross.

  • 2. Platt, 27 Nov. 1972, 92, 4.40 g. From coll. formed in Tunisia.
  • 3. Dies of 2. Tolstoi 12, 4.49 g (weight indicated by Mrs. Sokolova; 4.3 in Tostoi is incorrect).
  • 4. Dies of 2. Münz. u. Med. FPL 436, Sept. 1981, 35 = Bonham's 3, 3 Dec. 1980 (O'Hara), 35, 4.42 or 4.47 g. In spite of the difference between weights, it is the same specimen.
  • 5. Obv. of 2. Platt, 27 Nov. 1972, 90, 4.43 g. From coll. formed in Tunisia.
  • 6. Platt, 27 Nov. 1972, 93, 4.39 g. From coll. formed in Tunisia.
  • 7. Obv. of 6, illustrated MIB 25.4. Γ at end of rev. legend prolongs exergual line. BM (Graves, 4002), 4.02 g.
  • 8. H. Christensen, 11 July 1975, 736 = H. Christensen, 6-7 Dec. 1974, 1.
  • 9. Bucharest, Academy = Cahn 80, 27 Feb. 1933, 1045.
  • 10. From the vicinity of Jericho. Hoard reported in W. Hahn, "Eine Anfangsemission des Raisers Zeno und ein Schatz fr�hbyzantinischer Solidi bei Jericho," M�NG 18 (1973), pp. 1-4; this specimen comes from the same source but does not belong to the hoard itself. Hahn coll. 4.45 g, illustrated MIB 25.3.
  • 11. Bardo, Porto-Farina hoard 1944, 1, 4.47 g.
  • 12. Specimen differs from the preceding ones, has three pellets in central line of cuirass, angel has smaller, more elongated head. BM (Cracherode), 4.42 g, illustrated MIB 25.2.

b) Constantinopolitan manner. Three pellets in central line of cuirass, shield without loop. Rev. field star often small.

  • 13. Bardo, Dougga hoard, 1911, 4.42 g.
  • 14. Dies of 13. Bardo, Porto-Farina hoard, 1944, 2, 4.46 g, pierced twice.
  • 15. Γ at end of rev. legend appears as a C. DOC 227c (Baranowsky 1959), 4.45 g.
  • 16. Dies of 15. BM, 1849-11–21-19, 4.33 g.
  • 17. Dies of 15. Bourgey, 4 May 1972, 85.
  • 18. Γ at the end of rev. legend appears as a small C. Bologna, Museo civico archeologico (ex Palagi coll.), 4.43 g.
  • 19. Obv. of 18. Same shaped Γ in rev. legend. BNC 05 (from N. Kapamadji, ex Ricklin coll.), 4.50g. Au: 98%; Ag: 1.63%; Cu: 0.35% (proton activation analysis, JOB 1983).
  • 20. Same shaped Γ in rev. legend. Bridge coll., NCirc 78 (1970), p. 246, 6, not illustrated, 4.45 g.
  • 21. Obv. of 20. Γ reaches exergual line. N. K. coll. 121 (as Constantinople), 4.48 g.
  • 22. Γ reaches exergual line. Barthold-Poplavsky, 15 June 1976, 309.
  • 23. Dorotheum 412, 1-4 Dec. 1981, 882.
  • 24. Rev. of 23, dies very worn. N. K. coll. 132 (as uncertain mint), 4.45 g.

c) Constantinopolitan manner. Incipient loop on top of shield; angel face is smaller and more elongated. Rev. field star is small.

  • 25. Bardo, Dougga hoard 1911, 4, 4.40 g.
  • 26. Dies of 25. Dresden.
  • 27. Obv. of 25. BM (Blacas 1020), 4.43 g.
  • 28. Dies of 27. Stockholm, 3.83 g [sic].
  • 29. Münz. u. Med. FPL 370, Aug. 1975, 25, 4.47 g.
  • 30. Dies of 29. Banque Centrale de Tunisie (provenance Tunisia), 4.40 g.
  • 31. N. K. coll. 122 (as Constantinople), 4.44 g.

MIB 25: ⊖ at end of legend.a

African manner. Central line of cuirass is quite high, has three pellets; shield without loop; last C of legend, more rounded, touches shield. Rev. angel's head rather big in proportion to body height; on 1. I placed under crossbar.

  • 32. Carthage (provenance Tunisia), 4.41 g.
  • 33. Note similarity of obv. design to 12. Copenhagen (De Schousboe 5a), 4.46 g, Tangiers, 1834 (MIB 25.8, illustrated MIB 3, pl. 53).
  • 34. Rev. double struck, rough style, I in legend 1. in angle of cross. N. K. coll. 124, 4.43 g.

MIB 25, BNC 02-03, DOC 277e: I

Constantinopolitan manner. Three pellets in central line of cuirass. Rev. R in VICTORI lines up with crossbar.

a) The r. pteryges extend directly from shield image

  • 35. BNC 02 (from Tunisia, 1960), 4.50 g. Au: 97.3%; Ag: 2.3%; Cu: 0.42% (proton activation analysis, J�B 1983).
  • 36. Dies of 35. Platt, 27 Nov. 1972, 89, 4.47 g. From coll. formed in Tunisia.
  • 37. Platt stock 1976, 4.42 g.
  • 38. Schweizer. Kredit. 33, Winter 1980/1, 202, 4.46 g.
  • 39. N. K. coll. 126 (as Constantinople), 4.43 g (illustrated RN 1962, pl. 7, 5).
  • 40. Obv. of 39. DOC 277e.3, 4.39 g (ex Grierson coll. Christie's, 13 Dec. 1948, lot no. unknown).

b) The r. pteryges start slightly beneath shield image

  • 41. Vinchon, 17 Dec. 1973, 88.
  • 42. Dies of 41. Carthage (provenance Tunisia), 4.40 g.
  • 43. Obv. of 41. Münz. u. Med. FPL 332, Mar. 1972, 26 (from Carthage or its region).
  • 44. Birmingham (from Spink 1952 = Ratto 446?), 4.44 g.
  • 45. Tolstoi 16, 4.4 g.
  • 46. Dies of 45. Ciani-Vinchon, 6 May 1955, 512.
  • 47. Souassi hoard 1, 4.45 g.
  • 48. Dies of 47. Glendining, 16 Nov. 1950 (Hall 2), 2214, 4.41 g.
  • 49. BNC 03, 4.36 g.
  • 50. DOC 277e.l,4.44g.
  • 51. ANS 1968.26.2 (Nov. 1967, ex Ricklin coll. 90), 4.39 g.
  • 52. Bourgey, Dec. 1966, 55.
  • 53. Lower part of cuirass looks like lattice. DOC 277e.2 (ex Pierce coll.), 4.43g image

c) Incipient loop on top of shieldimage

  • 54. Platt, 27 Nov. 1972, 91, 4.42 g. From coll. formed in Tunisia.
  • 55. Dies of 54. N. K. coll. 127 (as Constantinople), 4.43 g.
  • 56. Obv. of 54. Baudey and Pesce (Lyons), 17-19 Oct. 1982, 685, 4.38 g.

MIB 25, BNC 04, DOC 277a: IA

Constantinopolitan manner. Three pellets in central line of cuirass. Rev. R in VICTORI lines up with crossbar.

a) Simple shield, r. pteryges start beneath shield. Rev. angel's head rather big in propor tion to body height.

  • 57. BM (acquired 1867), 4.47 g.
  • 58. Obv. of 57. From a hoard discovered on the Algerian coast around 1978, from which Mr. Zaidman purchased this coin and a solidus of Theodosius II VOT XX MVLT XXX B, a coin of Pulcheria, one of Leo and another of Zeno; the latter coins having been sold, no more precise identification was available. Nomisma stock, 1978, 4.47 g.
  • 59. DOC 277a.l (from Leu, 1958), 4.48 g.
  • 60. Obv. of 59. DOC 277a.2, ex Schindler coll., Grabow 18, 27-30 Oct. 1941, 532, 4.44 g.
  • 61. Obv. of 59. Monte Judica hoard 60, 4.49 g.
  • 62. Obv. of 59. Souassi hoard 4, 4.42 g.
  • 63. Florange stock, 1967 (ex Ricklin coll.), 4.44 g.
  • 64. Münz. u. Med., 6 Dec. 1946, 876.
  • 65. Superior Stamp and Coin, 19-23 Aug. 1975, 3113.
  • 66. Dies of 65. Münz. u. Med. 12, 11 June 1953, 911, 4.48 g (illustrated MIB 25.6).
  • 67. Obv. of 65. Souassi hoard 2, 4.45 g.
  • 68. Stockholm.
  • 69. J. Schulman 265, 28-29 Sept. 1976, 900.
  • 70. Vinchon 12, Apr. 1957, 33 (illustrated RN 1962, pl. 7, 3).

b) Loop or incipient loop on top of shield. Rev. angel's face is narrower and longer.

  • 71. Ratto 447.
  • 72. No pellets in central line of cuirass. Bonham's 3, 3 Dec. 1980 (O'Hara, ex Ricotti), 32, 4.45 g.
  • 73. Rev. of 72. Sternberg 8, 17 Nov. 1978, 781, 4.37 g.
  • 74. Emperor's face and neck elongated through end of series. N. K. coll. 144, 4.42 g.
  • 75. Cambridge, Fitzwilliam, Grierson coll. (from Tunis, 1967). Given the date of acquisition and the die links to 76, it may well have come from the Souassi hoard. 4.46 g.
  • 76. Dies of 75. Souassi hoard 3, 4.40 g.
  • 77. Birmingham G 573 (from Spink, Feb. 1952), 4.45 g.
  • 78. Obv. of 77. Carthage (provenance Tunisia), 4.45 g.
  • 79. Rev. of 78. BNC 04 (Ane. fonds 381, ex Duke of Orléans coll.), 4.41 g.
  • 80. Dies of 79. Frey 9, Jan. 1969, 13.
  • 81. N. K. coll. 143, 4.44 g (obv. appears incorrectly under no. 145 in catalogue).
  • 82. Obv. of 81. Glendining, 14 Jan. 1953, 185.
  • 83. Rev. of 82. Birmingham, G 934 = Glendining, 22 Sept. 1960, 914, 4.30 g.
  • 84. Central Bank of Tunisia (provenance Tunisia), 4.44 g.
  • 85. Dies of 84. The Hague.
  • 86. N. K. coll. 143, 4.45 g (rev. appears incorrectly under no. 142 in catalogue).
  • 87. Monte Judica hoard 59, 4.43 g (Schweizer. Kredit. 2, 27-28 Apr. 1984, 741).

MIB 25, DOC 277b: IB

Constantinopolitan manner.

  • 88. Incipient loop on top of shield. N. K. coll. 145, 4.44 g (obv. appears incorrectly under no. 142 in catalogue; illustrated RN 1962, pl. 7, 4, and MIB 25.7).
  • 89. DOC 277b (ex Grierson coll. from Platt, 1 Sept. 1950), 4.48 g. Simple shield without loop. Imitation?
  • 90. ANS 1968.26.1,4.41g.

Coin 90 was acquired from Bank Leu, November 1967, with another solidus of Carthage (above, 51), a Burgundian solidus, and two other Byzantine solidi but had no apparent relationship to them. According to Mrs. S. Hurter, whom we gratefully acknowledge for having searched at our request through her archives, these coins originated from a collection "which had been built up in Portugal." But the fact that 51 is known to have belonged to the Bicklin collection acquired either in North Africa or in Paris throws doubt on the archaeological value of the Portuguese origin of the lot — see the observations of J. P. Callu, in Crise ei redressement dans les provinces europ�ennes de l'Empire, mil. du Ill�-mil. du IV� s. ap. J.C. (Strasbourg 1983), p. 162, n. 38, about a similar origin attributed to an unpublished hoard of the early fourth century. Otherwise, 90 must be considered an intruder in the lot.

This coin differs much in style from the two preceding coins; it recalls the obverses of 33 and 34 (i.e. the same high placing of the central line and the disappearance of the final C in the reverse legend which is entirely confused with the shield border on 33 and 34). The reverse, less rough, recalls those of 2 through 12, as the reverse I is wedged in the angle of the crossbar. It might be an imitation or may simply have been a recurrence of the African manner.

Justinian's IB Issue

Contrary to P. Grierson's advice, A. R. Bellinger in DOC 1 included among the issues of Carthage a solidus (298) with IB and an angel holding a cross with christogram. Despite the corrections in BNC 1, p. 66, n. 5, and MIB 1, p. 48, n. 10, this attribution still causes some confusion. These non-African IB issues of MIB 7 only differ from other Constantinopolitan issues of the same type through the letters at the end of the reverse legend. No specimen has ever been found in Africa and the known provenances point to their circulation (and possible issue by a provisional military mint such as that which operated in Salona for copper?) in the Balkans. There are no die links among the extant specimens. The following short list has been compiled from my own photo file supplemented with information kindly provided by Wolfgang Hahn.

  • 1. Athens. Found near Nestani (Northern Arcadia, Peloponnesus) by a peasant. The same spot yielded a hoard of solidi of Justinian from Constantinople in 1948 (of which five are preserved in the museum). In spite of the indications given by A. Avramea, "Nomismatikoi 'thēsauroi kai memon�mena nomismata apo tēn Peloponnēso," Symmetika, Ethnikon Idryma Hereun�n, Kentron Byzantin�n Hereun�n 5 (1983), p. 62, citing this specimen as an issue of Carthage, it is not certain that it did belong to the hoard (my thanks to M. Caramessini-Oeconomides who provided the information and a cast of the coin).
  • 2. DOC 278 (ex Schindler, acquired in Graz in 1947, found in Albania).
  • 3. Tolstoi 40 (acquired in Rome).
  • 4. Ratto 463.
  • 5. A. R. Bellinger coll.
  • 6. Birmingham G666 (from Stanley Gibbons).
  • 7. Birmingham G933 (from Seaby, 1960).
  • 8. Birmingham G1387 (from Bank Leu, 1964).
  • 9. Istanbul.
  • 10. Bucharest.
  • 11. Brussels.
  • 12. Berlin.
  • 13. Vienna 204.414.
  • 14. Cahn 59, 14 Mar. 1928, 38.
  • 15. Münz. u. Med. 43, 12-13 Nov. 1970, 542.
  • 16. Kress 139, 19 June 1967, 2089.
  • 17. Peus 268, 24-26 April 1968, 287.
  • 18. Auctiones AG, Basel 7, 7-8 June 1977, 846.
  • 19. Wruck, Jan. 1972, 790.
End Notes
a
DOC 277d (⊖ engraved on Z) is a Constantinopolitan issue (MIB 6). Cf. the angel's beaded cross, a design never found in Carthage but sometimes seen in the capital (see above, p. 41, n. 8).
9
Carthage Carthage Constantinople Constantinople
Diameter Mean S.D. Mean S.D.
Flan 19.5 0.8 20.7 0.8
Obverse Linear Border 17.8 0.6 18.2 0.5
Difference 1.7 2.5
10
Or gives this impression because of the relatively great importance of the head in relation to the body.
11
MIB, p. 51.
12
Hendy, pp. 129-54, at p. 143.
13
MIB 3, p. 45.

Justin II

Obv. DNIVSTI NVSPPAV variously disposed. Bust of Justin II with helmet and cuirass, in 1. shield, in r. globus surmounted by Victory crowning him with wreath.

Rev. VICTORI AAVCCC and a letter; in exergue CONOB. Constantinople seated facing looking r. with helmet, tunic, and mantle; r. leans on spear, in 1. globus cruciger.

MIB l8a, DOC 190a: A

DNIVSTI NVSPPVI, three pellets in central line of cuirass.

  • 1. BM (Blacas 1023 = Madden 444), 3.82 g.

MIB 18a: B

DNIVS TI NVSPPAV two pellets in top of cuirass. image

  • 2. Three pellets in central line of cuirass, asymmetrical pteryges as 1; obv. legend unbroken, unlike 4 through 8; therefore it belongs like 3, at beginning of issue. Vinchon, 26 Oct. 1964, 63.
  • 3. Three pellets in central line of cuirass, pteryges tending toward asymmetry of following specimens; obv. legend unbroken. Ravenna 2399 (E. Ercolani-Cocchi, Imperi romano e bizantino attraverso le monete del Museo nazionale di Ravenna, nov. 1983-mar. 1984, fig. 150), 4.40 g.
  • 4. BMC 3 (Rollin 1904), 4.37 g.
  • 5. Obv. of 4. Stanley Gibbons FPL 1 (n.d.), 116.
  • 6. Vinchon, 20 May 1974, 22.
  • 7. Cuirass has winding strip dotted with pellets. BNC 01 (as Constantinople) (Anc. fonds 468, inv. in 1700 A.D.), 4.44 g. Au: 97.08%; Ag: 2.5%; Cu: 0.37% (proton activation analysis, see Cahiers Ernest-Babelon 2, 1984).
  • 8. Cuirass winding strip dotted with pellets. Platt, 27 Dec. 1972, 97, 4.44 g. From a collection formed in Tunisia.

MIB 18a, DOC 190b: Γ

DNI VSTI NVSPPAV, three pellets in central line of cuirass.

  • 9. BMC 4 (Rollin 1904), 4.40 g.

MIB 18a: Є

DNI VSTI NVSPPAV, three pellets in central line of cuirass.

  • 10. Obv. of 9. BMC 7 (Rollin 1904), 4.47 g.

MIB 18b, DOC 190c: Є

DNIVSTI NVSPPAV, three pellets in central line of cuirass.

  • 11. DOC 190c (Peirce coll.), 3.62 g, "No explanation of the abnormally low weight" (Grierson). See the similarly low weight of 1.
  • 12. Obv. of 11. Belgrade.
  • 13. Oxford, Ashmolean Museum.
  • 14. Helbing FPL 17 (n.d.), 655.

MIB 18b, DOC 190d: S

  • 15. Cuirass design similar to 11 and 12; Bonham's 3, 3 Dec. 1980, 47, 4.38 g.
  • 16. Pellets on cuirass weakly engraved, first two overlap line beneath; N. K. coll. 166 (as Constantinople), ex Ricklin, 4.43 g.
  • 17. Obv. of 16; BMC 8 (Rollin, 1904), 4.43 g.

MIB 18b, DOC 190e: Z

DNIVSTI NVSSPAV, three pellets in central line of cuirass; rev. Z overlaps exergual line except on 18 and 22.

  • 18. ANE, Barcelona (Feb. 1971), 191.
  • 19. Obv. of 18. Private coll.
  • 20. Obv. of 18. Brussels 49795, 4.44 g.
  • 21. Obv. of 18. Cambridge, Fitzwilliam, 4.47 g.
  • 22. DOC 190e (from Vinchon, 29 Oct. 1962), 4.40 g. North African provenance.
  • 23. Birmingham G566 (from Seaby, 20 Aug. 1956), 4.43 g.
  • 24. Bardo, Djebibina hoard, 4.41 g.

MIB 18b: H

DNIVSTI NVSPPAV, three pellets in central line of cuirass.

  • 25. Bardo, Djebibina hoard, 4.40 g.
  • 26. Cambridge, Fitzwilliam (Grierson coll., from Vinchon, 1967), 4.43 g.
  • 27. Dies of 26. Birmingham G969 (from Baldwin's, 1960), 4.37 g. From an Italian hoard.
  • 28. Obv. of 26. Münz u. Med. FPL 126, July 1953, 26 (not illustrated but described as "off. H. Provincial style.") = Cahn stock 1952 (from Grierson's photofile)
  • 29. Dies of 28. Vinchon, 29 Oct. 1973, 114.

MIB 18b, BNC 01, DOC 190f: ⊖

DNIVSTI NVSPPAV, three pellets in central line of cuirass.

  • 30. BNC 01 = Platt, 18 Mar. 1970 (Longuet), 320, 4.34 g.
  • 31. Ratto 757.
  • 32. Barthold, 4 Mar. 1982, 58.
  • 33. Glasgow, Hunterian Museum.

MIB 18b, BNC 02, DOC 190: I

DNIVSTI NVSSPPAV, three pellets in central line of cuirass.

  • 34. BNC 02 (from Tunisia, 1967), 4.06 g.
  • 35. Balvin coll. 865 = Münz. u. Med. FPL 126, July 1953, 27 (illustrated MIB 18b), 4.38 g.
  • 36. Birmingham G0025 = P. Grierson, "A Byzantine Hoard from North Africa," NC 1953, pp. 146-48, 2.
  • 37. DOC 190g = P. Grierson, "A Byzantine Hoard from North Africa," NC 1953, pp. 146-48, 1.
  • 38. Obv. of 37. ANS 56.25, 4.35 g.
  • 39. Lindpaintner coll., Innsbruck, 4.30 g. Acquired in Seville with 2 tremisses of Justin II from Carthage, MIB 19.
  • 40. Dies of 39. Hirsch 137, 29-30 June 1983, 420, now Hahn coll.

IA

DNI VSTI NVSPPAVC, three pellets in central line of cuirass.

VICTORI AAVCCCIA (A rather than A)

  • 41. H. J. Berk, Roman Coins of the Medieval World, 383-1453 A.D. (Joliet, Ill, 1986), 65.

DNIVSTI NVSPPAVC, three pellets in central line of cuirass.

VICTORI AACCCI

  • 42. Avignon, Calvet Museum (Inv. Byz. 184) 4.26 g.b
End Notes
b
Médaillier, Musée Calvet 1, les monnaies 1, ed. G. de Lo�e (Avignon, 1987), 184. Thanks are due to the Museum Director, M. Georges de Lo�e, and to J. C. Richard for enabling us to illustrate it here.

JUSTINIAN

Dating

The dating criteria at our disposal are not very numerous. On the one hand, the occurrence of solidi with facing bust, introduced at Constantinople in 537, and on the bronze beginning in year 12 (538) furnishes a terminus ante quern. On the other, comparison with the stylistic development of the folles and other bronzes of Carthage indicates a close relationship between the solidi with Γ of African style and certain half folles of year 13. Both are characterized by the symmetrical pattern of the upper part of the cuirass and its decoration with two pellets which generally replace the three of the horizontal band: image (compare DOC 294, BNC 27, Tolstoi 380 in bronze, Plate 9, 2-12).

In Byzantine numismatics, comparison with bronze coinage marked with a mint name has permitted the more or less secure identification of a certain number of provincial gold coinages. This principle is the basis of the pioneering studies of the mints of Italy and Sicily by Laffranchi and Ricotti Prina 14 and was used by Hahn in his MIB. In addition at Carthage the same engravers were responsible for the dies of denominations in all three metals. Even if, in the administrative plan, the moneta palatina, responsible for gold and silver, and the moneta publica, restricted to bronze, were distinct and often housed in different locations (as has been proved at Ravenna),15 most of the time the limited output of the provincial mints did not justify the employment of many artisans or teams of artisans specializing in the preparation of dies.

At Carthage application of this principle can be pressed further, and stylistic comparison can establish a fixed point for chronology. The distinction between the two styles established for the gold is in fact observed on the bronze as well as on the silver. In the latter metal, where the immobilization of the profile bust precludes comparison with the solidus, the general tendency is nonetheless clear. One moves from a first issue (MIB 51, VOT MVLTI MTI) bearing an image closely related to that of Gelimer16 and likewise of Hilderic by the characteristic stiffness and schematism of the diadem and the folds of the chlamys to a series (MIB 52, monogram of Justinian)17 where the quality of the rendering of the hair, for example, is reminiscent of that of coins from the capital. The evolution of the bronze—most interesting in that it is in part dated from 539/40 (Justinian's thirteenth regnal year) and even from 538/9 (the second indiction)—is exactly similar. The undated folles as well as the exceptional follis of the second indiction (MIB N185) show signs of the same stiffness characteristic of the African style. This continues on a part of the bronzes of regnal year 13, and especially on most of the folles and half folles with SO (MIB 194 and 196) which make up the first part of the issue.18 But the Constantinopolitan style dominates exclusively later in the course of the same year, on the folles and half folles marked S (MIB 195 and 197) and on all of the dekanummia (MIB 199) as well as on other later fractions (MIB 201, with circle, bust r.; MIB 203, with wreath and circle, bust r.).19

If we rule out the hypotheses of several separate officinae (Γ, ⊖, I) and of supplementary officinae responsible for the minting of gold at Carthage, which its importance does not justify, we must otherwise explain the letters on the reverse. The only possible solution is an indication of date, but which one? The problem is summarized in Table 1.

Table 1 Political and Monetary Chronology of the Reign of Justinian I in Africa
Date Significant Events a Indiction Regnal Emissions Emissions
Year Sept. 1-Aug. 21 Year Aug. 1-July 31 N
534/5 Campaigns of Solomon in northwestern Africa and Numidia 13 8 M
535/6 Summer 536: Belisarius killed in troop mutiny; departure of Solomon 14 9 K
536/7 Reorganization of the army and victory over Stotzas in Numidia; Aug. 31, 537; Novel 47 on the dating of imperial documents 15 10 I
ϵ
ϵᴵᴵ
Date Significant Events Indiction Regnal Emissions Emissions
Year Sept. 1-Aug. 21 Year Aug. 1-July 31 N
537/8 1 11 IA/ΛΦP ЄB
(African) ЄS
538/9 2 12 Mind II (M IB N185, African)
539/40 Summer 539: return of Solomon, prefect and magister militum;b victory of Bagai; reconquest of Sitifian Mauretania fortifications 3 12 Γ (1/3 African, 2/3 Constan.) M, K, I VOT XIII
540/1 4 14 I, VOT XIIII
541/2 5 15
542/3 6 16 I, VOT XVI
543/4 Summer of 543: beginning of plague; revolt of Mauri in Tripolitania. Spring of 544: defeat and death of Solomon at Cillium 7 17
544/5 Devastation of northwestern Africa 8 18
545/6 Suppression of revolt of Guntarith 9 19 ⊖ (African)
546/7 End of 546: Johannes Troglita magister militum 10 20 I (Constan.)
547/8 Campaigns and victories of Antalas over the Mauri 11 21 IA (Constan.) VICTORIA*I
548/9
551 Defeat of Byzantine force sent to Sardinia against the Huns of Totila 14 21
552 Recovery of Sardinia 15 25
ca. 552 Death and replacement of Johannes Troglita
562 Dec: Assassination of Cusinas by John Rogathinus 11 36
563 Revolt in Numidia
564 Pacification enforced by an army sent from Constantinople. Thomas, prefect 12 37
565 Nov.: Death of Justinian 14 39

The exclusive use of the facing bust on the gold coinage, which does not allow its introduction to be placed before 537, and the existence of solidi with Γ make dating by regnal year impossible. There remains indictional dating, which I favored in 1970 without being able to enter into detailed classification of the issues then known (Γ, ⊖, I, IA).20 The stylistic development of the dated bronze leads to the search for a parallel tendency in the gold coinage and to placing first the issues of African style, i.e. the unique piece in Paris with ΛФP (1) and the first group of solidi with Γ (2-12). The piece in the Bibliothèque Nationale must also be separated from other solidi with IA. But such a distinction is inescapable considering the close relationship between this first piece and the beginning of the issue with Γ (compare 2-5, 6-7) and, by contrast, the complete absence of elements common to pieces with IA CONOB, all Constantinopolitan with the three characteristic pellets on the pectoral band of the cuirass. On these last, the mark IA is that of the eleventh indiction. On the other hand, the Paris solidus bears witness to that liking for ambivalence so common in Byzantine coin inscriptions: here in fact IA simultaneously indicates the eleventh year and the the first indiction (ἰνδιϰτιῶνος πρώτης), a coincidence unique in the course of the reign but one which occurred conveniently as the same moment that Novel 47 was issued to determine the dating of imperial documents, and which obviated for the moneyers a decision for which they lacked instruction.

The first solidus struck in Africa in the name of Justinian is, then, the one that bears the mark of the name of the province, and not CONOB—nor even one of the compounds which combine the name of the mint with OB—which had been de rigueur since the reforms of Valentinian.21 From the reign of Zeno to that of Leo III,22 under whom it disappeared, CONOB was the only mark to appear in the exergue of gold coins, from whatever mint they originated. The only exceptions to this rule, aside from the piece which concerns us here, are the solidus of Justin II with ΛΛЮOB23 and a solidus and a tremissis of Justinian with ROMOB (MIB 28 and 35). These last deserve attention, for they constitute what is in effect the Italian counterpart of the African coinage. Like it, they are rare, and thus may belong to a special issue; second, the more common series which follow them revert to the traditional exergue; and last but not least they are the first gold coins following the reconquest. It therefore seems that the authorities wished to note the return to imperial control of the two most prestigious cities or provinces that had been seized from them. In Italy, the Valentinianic tradition suggested a repetition of the mark ROMOB,24 while at Carthage, where the yellow metal had never been struck, latitude was allowed to celebrate the name of the province rather than that of its capital.25

The autonomy of the African mint is also manifested about the same period in the bronze coinage by the striking of an exceptional follis, also known in only a single example, with the legend FELIX INDICTIO(NIS) II (MIB N185). This was probably a result of the dilemma posed for dating by Novel 47 for all imperial acts. In the absence of precise instructions and as it was impossible to mark the coins by all the methods—regnal year, consulate, and indiction—the mint authorities at Carthage chose the indictional date before conforming the following year (regnal year 13) to the Constantinopolitan pattern on the bronze.26 The local character of the piece is also evident in the obverse legend DOMNIIVSTINIANI PPA, where the expansion DOMNI as well as the use of the genitive in the titulary constitute a hapax legomenon.27 Another eccentricity is the exaggered importance of the cross with alpha and omega above the mark of value. It occupies almost as much space as the latter, whereas at other mints the usual marks (cross, chrismon, etc.) rarely extend more than a little beyond the top of the M.28 Finally, the style remains markedly peculiar, recalling in its rigidity (note especially the linear rendering of the diadem) the style of earlier undated folles like MIB 184b, with the chrismon on the chlamys and the officina mark Γ.29

This second indiction (538/9) does not seem to have had any great effect on the coinage, since in fact no solidus nor any bronze coins bear its mark. But it is very probable that the striking of folles and fractions of the undated series continued and that, the ΛФP issue having constituted only an exceptional series intended for distributions, the provincial authorities contented themselves with the use of gold consigned from Constantinople. A clue is provided by examination of the composition of finds including solidi of Justinian which have come to light in Tunisia.30

In spite of the very imperfect nature of the documentation, the table shows that the Constantinopolitan solidi of Justinian are the most heavily represented in the group, and that they make up an overwhelming preponderance in deposits slightly later than 542, which are probably contemporary with the troubles and devastations of the years 543 to 545 (El-Djem, Dougga, Lemta). In the absence of further information concerning the dispersed portion of the Souassi hoard, buried in or about 548, it is difficult to interpret it as proof of a growing role for the mint of Carthage in supplying the province with cash. But it is certainly no accident that the issue with IA, the most heavily represented in this deposit, is precisely the one that according to estimates (see below) had the greatest output. That the solidi of Justinian and Justin II from the capital eventually give way in favor of local solidi in later finds (Djebibina, post-573, Es-Sermita, post-615) is a phenomenon already discussed elswhere.31 On closer inspection, the distribution of the different issues of Justinian in the Tunisian finds confirms the chronology proposed: the solidi with Γ (MIB 25) are associated with the earliest coins of Constantinople (MIB 5) in the Dougga hoard, around which must be arranged, on the one side, the metropolitan issues of El-Djem prior to 542 and, on the other, the provincial strikings of Souassi dated 546 to 548. Such a pattern, imperfect as the picture may be in the fragmentary state of our knowledge, supports the idea of a relatively late beginning of the striking of gold at Carthage, as well as the distinction made between the IA/ΛΦP issue of 537/8 (1) and the normal IA issues of 547/8 (57-87). It also highlights the relative autonomy of the provincial mint inasmuch as it did not comply with the typological change carried out on the metropolitan and Italian solidi in 542. Carthage, as already underscored by Hahn (MIB, p. 51), never adopted the cross with christo-gram of the last class of Constantinopolitan gold, and kept to the simple cross reverse type down to the latest issue in 548/9.

Table 2 Solidi of Justinian and Justin II in Tunisian Hoards
Place of Discovery a Dale of Discovery Ruler Mint Reference Date Number of Coins
El-Djemb 1904 Leo I c Constantinople Tolstoi 7 457-74 1
Tolstoi 9 1
Zeno Tolstoi 21 476-91 1
Anastasius MIB 6, Zd 507-18 1
MIB 7, A 1
MIB 7, Γ 1
Justin I MIB 2, 1 518-22 1
MIB 3, B 522-27 1
MIB 3, ⊖ 1
MIB 3, 1 2
Justin and Justinian MIB 2c, H 527 1
Justinian MIB 5, Δ 527-37 1
MIB 5, Є 1
MIB 5, 1 2
MIB 6, A 538-42 1
MIB 6, S 1
MIB 3, H 1
MIB 6, ⊖ 1
20
Douggae 1911 Valentinian III, Arles(?)f Reinhart 1
Tremissis
Justinian Constantinople MIB 5, H 527-37 1
Carthage MIB 25, Γ: 539-40 2
=13.25 4
Lemtag 1909 Justinian Constantinople MIB 7, A 542-52 1
MIB 7, Δ 1
MIB 7, S 1
3
Souassih 1961 Justinian Carthage MIB 25, I =47 546-47 1
MIB 25, IA= 547-48 3
61, 66, 75 4
Derhalfla, Djebibinai 1917 Honorius Ravenna Robertson j 393-423 1
Zeno Constantinople Tolstoi 15 var. Δ 476-91 1
Justinian MIB 7. Є 542-52 1
Justin II MIB 3a, I 567-68 1
MIB 5a, A 567-78 1
Carthage MIB 18b, Z = 571-72 1
24
MIB 18b, H = 572-73 1
25 7
Es-Sermita, near 1945 Justinian MIB 25. Γ = 11 539-40 2
Porto-Farinak Tiberius II MIB 13 580 1
Maurice MIB 25, Z, S 588 2
MIB 25, H 589-90 1
MIB 25, IЄ 596-97 1
Heraclius MIB 92al">l, A 612-13 1
MIB 92a, Γ 614-15 1
9
End Notes
a
For the sources and the chronology, see C. Diehl, L'A frique byzantine 1 (Paris, 1896), pp. 15-93, 333-81; D. Pringle, The Defence of Byzantine Africa from the Time of Justinian to the Arab Conquest 1, BAR International Series 99 (Oxford, 1981), pp. 16-39.
b
There was not at that time a "reorganization of the prefecture by Solomon," as Hahn suggests. The text cited (Vand. 4.19. 1-4) says only that "the emperor entrusted all the affairs of Libya to Solomon," Βασιλενς Σολόμωνι αύθις απαντα Λιβύης τά πράγματα ένεχείρισε, meaning only that at that time Solomon was magister militum or praetorian prefect, as the inscriptions confirm, J. Durliat, Les D�dicaces d'ouvrages de d�fense dans l'Afrique byzantine, Collection de l'école française de Rome 49 (1981), pp. 1-59. But such a concentration of civil and military forces is already attested during the first prefecture of Solomon in 535-36. Even if the inscriptions cited by Diehl (above, n.a), p. 117, only agree with his second prefecture, as Durliat has shown, it is nonetheless clear that Solomon, as prefect (compare Novels 36-37), was in essence the commander of the Byzantine armies in Africa (Vand. 4.10.1, 4.11, etc.). The same situation exists under Justin II, and it is therefore improper to speak of a reorganization of the prefecture and the return of Solomon to Africa.
a
None of these hoards has been preserved and none has been catalogued in its entirety.
b
"Cette trouvaille a �t� faite par des ouvriers travaillant × la route ouverte ... entre El-Djem et la Smala des Souassi, × travers les ruines de l'antique Thysdrus. × 300 mètres du village pr�s de la voie romaine qui conduisait aux Thermes Les ouvriers ... avaient mis au jour des mosa�ques.... × quelques mètres de ces mosa�ques, on d�couvrit une citerne antique.... C'est dans cette citerne que les ouvriers trouvaient une urne en poterie refermant des pi�ces d'or.... × la suite d'une enqu�te ... 62 pi�ces d'or furent remises � l'Administration, sur les 200 environ qui auraient, para�t-il, compos� ce petit trésor (Commandant de Bray, "Notes sur cinq trouvailles de monnaies antiques faites en Tunisie," Bulletin Soc. Arch. de Sousse 5 [1907], pp. 102-4, at p. 102).
c
There are 20 pieces preserved at the Musée du Bardo with the provenance "El-Djem" and which I take to be part of the 62 pieces "turned over to the Administration." Following the description of Commandant de Bray, the composition of this lot may be reconstructed as follows (the number of pieces preserved at the Bardo for the relevant reigns is given in parentheses).
Anastasius I Constantinople MIB 6 7 (1)
Constantinople MIB 4, 5, or 7 4 (2, MIB 7)
Justin I Constantinople MIB 2 4 (1)
Constantinople MIB 3 6 (4)
Justin I and Justinian I Constantinople off. H 1 (1)
Justinian I ? MIB 5 ? 9 (4)
? ? 31 (4, MIB 6)
Commandant de Bray's presentation is very confused, often vague and even erroneous. In fact he attributes to Justin I "nine examples with bust holding globus cruciger and reverse Victory facing holding a globe and a long cross" (p. 103) which can only be solidi of Justinian I from Constantinople (MIB 5) or Carthage (MIB 25). Nonetheless the comparison of these descriptions with the composition of the Bardo lot inspires belief in the identity of the finds. Since the "administration" to which the pieces were submitted is not specified, it is possible that not all the pieces were sent to the Bardo. One difficulty remains: the absence from de Bray's inventory of the easily noted coins of the two Leos and Zeno; but the identity of the rare piece of Justin and Justinian would seem to eliminate doubts.
d
The officina is not previously attested for this issue.
e
The provenance of these four pieces is attested solely by the label they bear in the cabinet of the Bardo. Neither the museum's registry of acquisitions nor the report of finds for this year in the Bulletin du Comit� makes note of this find.
f
The tremissis in question, in the name of Valentinian III, is similar to those published by W. Beinhart in his classic study "Die Münzen des tolosanischen Beiches der Westgoten," DJN 2 (1938), pp. 107-35, at pp. 119-20 and pl. 4, 54. In place of this classical attribution to the Visigoths at Toulouse (during the reign of Theodored, 419-51 ?), Lafaurie has recently proposed a restoration of these pieces, which "undoubtedly circulated in abundance in central and southern Gaul" to the imperial mint at Arles ("Monnaies d'or frapp�es en Gaule dans la seconde moiti� du vc siècle," BSFN 38 (1983), pp. 270-71.
g
A. Merlin, BCTH 1909, pp. ccxxxii-ccxxxiii: "Pr�s de Lemta ... entre Salada et Sidi-Abd-es-Salem, des indig�nes ont recueilli, �pars × fleur de terre, un certain nombre de sous d'or — une dizaine, dit-on — provenant sans doute d'un petit d�p�t fait en ces parages. J'ai pu me procurer pour le Musée Alaoui, cinq de ces monnaies.. .."
h
C. Morrisson and R. Gu�ry, "Le tresor byzantin de Souassi," BSFN 37 (1982), pp. 214-16.
14
L. Laffranchi, "La numismatica di Leonzio II," Numismatica 4 (1938), pp. 73-74; 5 (1939), pp. 7-15, 91–92; D. Ricotti Prina, "La monetazione siciliana nell'epoca bizantina," Numismatica 16 (1950), pp. 26-60.
15
Kent (above, n. 1), p. 201. M. F. Hendy, "Aspects of Coin Production and Fiscal Administration in the Late Roman and Early Ryzantine Period," NC 1972, pp. 117-39, esp. 130-35. At Ravenna, as at Constantinople, the moneta auri was naturally located in the imperial palace. Ravenna: "in porticum sacrii palatii," see Marini, I papiri diplomatici raccolti ed illustrati, no. 120; Constantinople: "Chrysoplysia" or "Chryseps�teion," a workshop for the purification and striking of gold and silver located in the great palace, see Niketas Choniates, CSHB, p. 453 = CFHB, p. 347; Nicholas Mesarites, Die Palastrevolution des Johannes Comnenos, A. Heisenberg, ed. (W�rzburg, 1907), p. 25.
16
See MIB, pp. 55-56.
17
BNC, p. 103, for the interpretation of the monogram as that of Justinian rather than Mastinas (Grierson). The attribution to Carthage is not further discussed there (compare MIB, p. 56, n. 50) for it stands to reason when this reading and the known provenances are considered.
18
All the half folles with K/SO are of African manufacture, as well as certain folles; but the Constantinopolitan style is already found on other examples of MIB 194 such as BNC 20, DOC 291, and Tolstoi 298.
19
The case of the dekanummia and pentanummia with the legend VICTORIA A (MIB 200 and 201) is more complicated: certain dekanummia in fact bear a "round-head" bust which is very close to that of the later undated bronzes without officina mark or mint mark on the chlamys (MIB 185c and 187). Thus it is not possible to place the issue after 547/8. Possibly it is a question of a series parallel to the current type of dated dekanummion (MIB 199) and the corresponding pentanummion (MIB 203) designated for the distributions in 537 ? Certainly the series is very abundant. This does not exclude the possibility of payments to the troops before or after the successful campaign against Stotzas (Procop., Vand. 4.19.1-4).
20
BNC, p. 66.
21
Amandry et al. (above, n. 2), pp. 279-95. Outside of Rome and Constantinople the mints associating their names with the sign of purity OB were those of Treveri, Mediolanum, Aquileia, Siscia, Thessalonica, Nicomedia, and Antioch.
22
The later solidi with CONOB in the exergue are those of the beginning of the reign of Leo III (717-20) (BNC 1; DOC 1-2). The solidus and the semissis of Artavasdus (742/3, DOC 1 and 4) represent only a temporary and partial return to the old type (the traditional VICTORIA AVGV being replaced by the IhSϤUS XRISTϤS hICA of the miliaresion).
23
Represented by two specimens: one in the Leuthold collection, illustrated by Lafaurie (above, n. 5), pl. 7, 7, and a new acquisition in the British Museum (MIB 12, illustrated).
24
In the age of Valentinian even the mark ROMOB is an exception (RIc 9, p. 132, 60-61, dated 388-94) inspired by a style in use at that time only at Constantinople (J. W. E. Pearce, RIC 9, p. 113).
25
On the importance of the personification of the province, considered as a divinity in all of North Africa in a cult which flourished from the second century A. D., and her iconography in the classical period, see M. Le Glay, LIMC 1.1 (Zurich, 1981), pp. 250-55, s.v. Africa. Possibly the representation was more popular in a general way outside the province itself than that of the Tyche of Carthage.
26
MIB3, p. 52. The reverse legend is curious in the disagreement of the nominative Felix and the genitive indictio(nis), if the reading indicated by our British informants is exact (neither Hahn nor I have seen the piece). I know no example of the indiction being so described in the African inscriptions of this period. Denis Feissel has been kind enough to draw to my attention the frequent use of the expression in the papyri of the fourth through the sixth centuries (see W. Preisigke, W�rterbuch der griechischen Papyrusurkunden, [Heidelberg, 19241, e.g. col. 625 τῆς εὐτνχονς ίνδικτιώνος less often εὐτνχεστάτη) and, for an epigraphic example close in chronology, an inscription from Miletus: INΔĀ Ẹ[...]YXS + (H. Gr�goire, Recueil des inscriptions chr�tiennes d'Asie Mineure 1 [Paris, 1922], no. 219). The preference for indictional dating was always greater at Carthage in the sixth century than at other mints where, except for some pre-Justinianic exceptions, it all but disappeared. Apart from the series of gold coins, a certain number of bronzes are also explicity dated in that way (Maurice, MIB 2, 119, 123, 128; Heraclius, MIB 3, 236).
27
In the seventh century and beyond, the name of the emperor is almost always in the nominative (DOC 2, pp. 99-100). the only exceptions to this rule are dative legends on some Italian solidi of Constantine IV and Justinian II and especially on coins struck at Carthage under Justin II, Maurice, and Heraclius. Here the genitive may be a borrowing from seals.
28
The same unaccustomed equality of the mark of value and other symbol (the date) is observed at Carthage on the folles of Justin II (MIB 74; BNC, pl. 24, 30).
29
Compare for example BMC 364 (pl. 9, 16), DOC 286d.3, BNC 11. This close relationship does not necessarily entail a reconsideration of the chronology of the undated bronzes proposed by W. Hahn, who places first (533-37) the folles with officina mark. Nonetheless it its very likely that the same engraver was responsible for MIB N185 and for most of the folles like MIB 184b.
30
This table is expanded (below, pp. 54-57) from the summary inventory presented in BSFN 37 (1982), pp. 214-15, in connection with the Souassi hoard.
31
C. Morrisson in R. Gu�ry et al., Recherches archéologiques franco-tunisiennes × Rougga 3. Le trésor de monnaies d'or byzantines, Collection de l'école française de Rome 60 (Rome, 1982), pp. 62-65.

Political Events

Once this sequence and its proposed dating are accepted, they can be set against the pattern of political events which may have prompted the local issue of gold as well as of copper coins. Striking similarities in fact appear in the chronological evolution of the two metals, except during the first years following the reconquest. The most urgent task for the authorities was to replace the silver and copper currency of the Vandal kings with imperial types; and gold, which even in Vandal times had remained an unchallenged imperial monopoly, was easily provided from the capital, as we have seen above. After Justinian's reform in 538, the issue of a nearly complete series of bronze denominations, from the follis to the nummus, dated regnal year 13, parallels the Γ solidi of the corresponding indiction, while the rare half folles of regnal year 22 (MIB 198) may be the counterpart of the equally rare IB solidi. If we take for granted the attribution of the VICTORIA dekanummia and pentanummia (MIB 200 and 204) to the 547-52 lustrum, as argued by Hahn,32 they could as well parallel the issues of the last IA solidi in 547/8. It would thus not be fortuitous that when the issue of gold ceases in 548/9, no significant bronze is issued except the + I + /CON dekanummium (MIB 201), which the overstrikes clearly place after MIB 200.

Examination of the sequence of gold issues and political events reveals another more important coincidence. If only the major issues are considered, leaving aside the unique AΦP coin and the θ33 and IB strikings known from only three specimens each, two groups appear: one in 539/40 (Γ), the other in 546/48. The former is clearly related to the return of Solomon, praefectus and magister militum during the summer of 539;34 the latter to the appointment and later activity of John Troglita from the end of 546 onward. In both cases, the appointment of a new comman-er-in-cheif takes place after a series of disorders of varying importance in the army. In 538, Procopius mentions a conspiracy partly caused by a protest against long-delayed pay;35 in 545/6 a similar arrearage in pay is the principal reason underlying Guntarith's revolt. No doubt the new nominees managed to restore law and order by paying what was due the soldiers from the state,36 probably with coins newly struck for the occasion.

The above pattern is that shown by the relative rarity of preserved specimens. A more precise though fundamentally similar picture is given by die analysis of the material. Using the various methods devised by I. D. Brown, C. Carcassonne, and J. Möller, the original number of dies for each issue (which can be considered an index of coin production) is estimated (see Figure 1 and Table 3).37

The Γ (539/40) and the I (546/7) issues lie in the same range with a score of estimated obverse dies (23 and 18 respectively), while the IA (547/8) issue is clearly the most plentiful of all with 35 estimated obverse dies. It may well have been related to John Troglita's activity and his building a stronger army against the Levathae.38

image

Fig. 1

Table 3 Estimated Size of Justinian's Carthage Gold Issues (565-578)
Issue Die Links Coins in Sample Obv. Dies Pairs of do Rev. Dies Pairs of dr Estimated Number of dies Using 1) C;a 2) B;b 3) M.;c Estimated Number of dies Using 1) C;a 2) B;b 3) M.;c Confidence Range Confidence Range Percent of Issue Knownd Percent of Issue Knownd Ratio of Known Dies to Calculated Dies Ratio of Known Dies to Calculated Dies
Obv. Rev. n do zo dr zr Obv. Rev. Obv. Rev. Obv. Rev. Obv. Rev.
IA 1 1 1 0 1 0
Γ, 539/40 [2] 30 17 20 21 11 1)23 38 19-27 29-47 0.74 0.55
[3] 2)22 40 70 53 0.77 0.52
[4] 3)24 40 0.54
[5
[6
[7
8
9
10
11
12
[13]
[14]
[15]
[16]
[17]
[18
[19
[20
[21
22
23]
24]
[25]
[26]
[27]
[28]
[29]
[30]
31
⊖, 545/544 32 3 3 0 3 0
I, 546/7 [35] 22 14 10 17 5 1) 22 39 14-22 25-53 0.72 0.43
[36] 2)23 46 64 45 0.72 0.36
37 0.68 0.40
38
[39
[40
[41]
[42]
[43
44
[45]
[46]
[47]
[48]
49
50
51
52
53
[54]
[55]
[56
IA, 547/8 [57 31 20 15 23 8 1)32 54 24-38 34-64 0.61 0.45
[58 2)31 58 62 44 0.56 0.43
[59 0.59 0.44
[59
[60
[61
[62
63
64
[65]
[66]
[67
68
69
70
71
72]
73]
74
[75]
[76]
[77
78]
[79]
[80]
[81
82]
83]
[84]
[85]
86
87
IB, 548/9 88 3 3 0 3 0
89
90

The Africa IA issue (537/8) and the ⊖ (545/6) and IB (548/9) issues are too sparsely represented for any estimate to be possible: the number of preserved coins and dies then serves as the lower limit, well below the level for the three main issues. The historical context allows us to consider the Africa type (1) as a small ceremonial issue at a time when the mint practice for gold was still somewhat tentative; the production of coins with ⊖ (32-34) may have been restricted because of the revolt and other difficulties in 545/6, while IB (88-89) just represents the fading out ("queue d'�mission") of the abundant IA issue.

Whatever the exact relationship between local circumstances and gold production at Carthage during the reign of Justinian, there is no doubt that solidi (like bronze) were not issued on a regular annual basis, nor even at fixed intervals. This pattern is much different from that which prevails later at the African mint, and which begins to emerge during the reign of Justin II.

End Notes
i
A. Merlin, BCTH 1917, p. ccxxxvii, n.4: "M. Fanet, contr�leur civil de Zaghouan, m'a fait parvenir pour le Musée du Hardo, sept sous byzantins ayant appartenu × un trésor renferm� dans un vase de terre cuite (hauteur 0 m 08, diamètre max. 0 m 055) qui aurait �t� trouv� dans un douar de la r�gion du Djebel Derhafla (r�gion du Djebibina)."
j
A. S. Robertson, Roman Imperial Coins in the Hunter Coin Cabinet, University of Glasgow 5 (Oxford, 1982), pl. 88, 17.
k
G.-Ch. Picard, BCTH 1943-45, pp. 427-28: "In a vineyard situated at a place called Es-Sermita, 2km southeast of Rasel-Djebel and 1 km northeast of Si Rou Krima ... workers digging up the earth discovered, at some distance from the cisterns, large blocks of rock which they pulled out. Under these blocks there then appeared a small piece of broken pottery containing 33 Ryzantine gold pieces This hoard, deposited at the Directorate of Antiquities ... includes " Relow is the editor's inventory of the solidi, with Sabatier numbers, the possible identification of the pieces, and an account of the coins now preserved at the Musée du Bardo.
12, Justinian I, S. 2 MIB 5, Constantinople 2, MIB 25, Carthage
2, Justin II, S. 1 Carthage or Constantinople? None
4, Tiberius II, S. 1 Carthage or Constantinople? 1, MIB 13, Carthage
8, Maurice, S. 7 Carthage 4, MIB 25, Carthage
2, Phocas, S. 1 Carthage or Constantinople? None
1, Heraclius, S. 2 MIB 1-5, Constantinople None
4, Heraclius and Heraclius Constantine, S. 48 bis Carthage, large module, or Constantinople ? 2, MIB 92a, Carthage, large module
33 9
In spite of their number (less than a third of the whole), it is probable that the nine pieces preserved at the Rardo with the ticket "Porto Farina 1944" come from the hoard of Es-Sermita. The proximity of the places, the dates of record, and the close correspondence of the two inventories leave little doubt on this point.
a
Ch. Carcassonne, "Tables pour l'estimation par la m�thode du maximum de vraisemblance du nombre de coins de droit (ou de revers) ayant servi × frapper une �mission," Table Ronde Numismatique et Statistique, 17-19 Sept. 1979, Paris.
b
I. D. Brown: "Some Notes on the Coinage of Elizabeth I with Special Reference to Her Hammered Silver," BNJ 28 (1955-57), pp. 568-603; "On the Use of Statistics in Numismatics," NCirc 77 (1969), pp. 83-84; "Statistical Methods as a Tool in Numismatics," Cornucopiae 3 (1975), pp. 33-44.
c
J. W. Möller, "Estimation du nombre original de coins," PACT 5 (1981), pp. 157-72.
d
I. J. Good, "The Population Frequencies of Species and the Estimation of Population Parameters," Biometrika 40 (1953), pp. 237-64, where image = percent known.
32
MIB, pp. 70-71.
33
The rough style of this striking and the unexpected return to the African style may have been a consequence of the troubles surrounding Guntarith's revolt which ended, according to Procopius, "in Justinian's nineteenth year" (Vand. 4.28.34).
34
"In Justinian's thirteenth year," Vand. 4.19.1.
35
Vand. 4.18.9 on the unsuccessful coup of Maximinus. The reason is one already given by Procopius for the mutiny of Stotzas in 537 in the speech he attributes to the rebel (Vand. 4.15.55).
36
This is clearly stated by Procopius in the case of the patrikios Germanos who, following Stotzas' mutiny in 536, gave the repentant mutineers "their pay for the time during which they had been in arms against the Romans" (Vand. 4.15.5).
37
C. Morrisson, "Estimation du volume des solidi de Tib�re et Maurice × Carthage," PACT 5 (1981), pp. 267-84, with references to the earlier literature. Ch. Carcassonne has since published a report on her method with accompanying tables, 'Tables pour l'estimation par la m�thode du maximum de vraisemblance du nombre de coins de droit (ou de revers) ayant servi × frapper une �mission," 2 Symposi numismatic de Barcelona (Barcelona, 1980), pp. 115-28. Using my data, Mr. J�rg Möller was kind enough to calculate figures estimated according to his own method, based on the empirical frequency of dies, and I am very grateful to him for this invaluable help.
38
Corippus (Ioh. 7.150-242 and 272-73) mentions the recruiting of new troops and 150,000 Moorish allies. However exaggerated the figures, no doubt a greater financial effort was made in the winter of 547/8 than in the previous one. The increase in coin production may reflect an increase in military expenses.

JUSTIN II

Justin II's gold issues from Carthage have long been confused with Constantinopolitan coinage and were first identified by Philip Grierson in volume 1 of DOC. He pointed to four criteria (note to 190c) for distinguishing the coins of Carthage:

  • 1. The obverse inscription is not broken by the Victory as on other solidi;
  • 2. A is regulary used in the legends instead of A;
  • 3. The loop at the top of the shield is misunderstood and replaced by the curved elongation to the right of the two lines at the top of the cuirass;
  • 4. There are three pellets in the top horizontal bar of the cuirass.

His first observation is true of most issues except all B, Γ, Δ, and earlier ones (2-12) which have an unbroken inscription like the unique I coin (41). Point 2 applies only to a great majority of reverse legends, obverse inscriptions displaying A and A. The last two criteria are of almost general application. The only exception to the fourth criterion is precisely the B issue (already remarkable for its unbroken inscription) which shows a point on either side below the pteryges and above the horizontal bar of the cuirass, similar to Justinian's Γ in the African style (Justinian, 2-12). This is one of the rare obvious connections between the gold coinage of the two emperors, the other perhaps being the systematic design of the loop at the top of the shield as an elongation of the left pteryges that is derived from the loop (or incipient loop) on most of Justinian's later coins (I, IB). Another peculiarity of the Carthage legends under Justin II is the absence of the final B on the obverse: except on the two earliest types (A and B), the legends end in PPAV instead of the PPAVC used at Constantinople.

The change in iconography between Justinian I and Justin II and the choice of a new type for the reverse, as well as the replacement of the globus cruciger by a Victory on globe on the obverse, precludes close comparison of local styles in these two reigns. But a difference does appear in the design of the cuirass: under Justinian, the pteryges are asymmetrical, or at least separate, whereas under Justin II they always touch one another, often in the middle of the cuirass. During the latter's reign, the gold coins of Carthage do not display any originality in fabric, except perhaps a higher relief in the figure's design. They are also of the same module as metropolitan solidi, and the diameter of the border of dots also equals that at Constantinople. It would appear that the geometrical characteristics of the solidi from the two mints were now very close to one another, perhaps due to more precise specifications from the administration responsible.39

Carthaginian solidi do, however, retain distinctive stylistic characteristics besides the four typological criteria mentioned above, namely the higher relief of the figures, the more conspicuous lettering, and, above all, the peculiar treatment of the eyes. Unlike metropolitan coins, they also bear, as they did under Justinian, a date at the end of the reverse legend. That the letter at the end of this legend has a chronological meaning is indicated not only by the volume of production, which is far too small to account for the activity of ten officinae, but also by 41 and 42 with IA or IΓ.40 This last issue precludes the interpretation of the letter as an indictional date, since no thirteenth indiction occurred in the reign of Justin II.

Table 4 Politcal and Monetary Chronology of Justin II's Reign in Africa
Date Significant Events Indiction Regnal Emissions Emissions
Year Sept. 1-Aug. 31 Year Nov. 15-Nov. 14 N
565 14 1 A X and V PR ANNO (MIB 78, 81)
ca. 566 Construction of new defensesa 15 2 B K, I, facing bust, lustral type for 567-72 (MIB 75, 79, 82)
567 Financial help sent to Africa (?)b 1 3 Г
568 2 4
569 Theodorus, praefectus, slain by the Moors 3 5
570 Theoctistos, magister militum, defeated and killed by the Moors 4 6 S
571 Amabilis, magister militum, killed by the Moors 5 7 Z
572/4(?)-578 Thomas praefectusc 6 8 H M, K, ANNO VIII
573 The Maccuritae tribe renews obedience to the emperor 7 9 (MIB 75, 76) 1 NM (MIB 80) two busts
574 574-78, under Justin II and Tiberius, restoration and repairs of walls at Iunca Soriana (Bordj Younga)d 8 10 1 M, K, ANNO X, two emp. enthroned (MIB 74, 77)
757 9 11 IA
576 10 12
577 11 13 ΙΓ
578 Justin II dies, Oct. 5

Table 4 shows the extent to which the gold issues parallel the copper ones. In both metals, Justin's African coinage began with a primo anno dating. The undated bronze series with the facing bust of Justin alone is best attributed to the early years of his reign on the basis of the appearance of a similar type in other mints, from 565 to 569 in Antioch and from 565 to 568 in Thessalonica.41 In gold, with the exceptions only of years 4 (Δ) and 12 (IB), for which no specimens are yet known (although they could well appear in the future), all regnal years of Justin II are represented by one or more specimens. The issue of gold is placed on an annual basis from this reign onward, although it is struck in varying quantities. That of bronze also becomes more regular than it had been under Justinian, with two other dated issues in regnal years 8 and 10. The later pattern of the Byzantine coinage at Carthage, with regular annual issues of gold and lustral periodicity for silver and copper, thus takes shape progressively under Justin II.

The variations in output, which estimates of the original number of dies may indicate (Figure 1, above, and Table 5), show two significant peaks for the B and I issues and a smaller one for the Є issue. Both are paralleled in bronze according to our chronology, a fact which supports our interpretation of the letters on the solidus as regnal years. The issue of the second year covers part of the first indiction, and this may not be fortuitous. It should be noted that it is also partly contemporaneous with the financial "comfort" sent to the province from Constantinople in 56.42

Table 5 Estimated Size of Justin II's Carthage Gold Issues (565-578)
Issue Die Links Coins in Sample Obv. Dies Pairs of do Rev. Dies Pairs of dr Estimated Number of Dies Using 1) C;a 2) B;b 3) M.;c Estimated Number of Dies Using 1) C;a 2) B;b 3) M.;c Confidence Range Confidence Range Percent of Issue Knownd Percent of Issue Knownd Ratio of Known Dies to Calculated Dies Ratio of Known Dies to Calculated Dies
Obv. Rev. n do zo dr zr Obv. Rev. Obv. Rev. Obv. Rev. Obv. Rev.
A, 565/6 1 1 1 0 1 0
B. 566/7 2 7 6 1 7 0 1) 19 0.31
3 2) 21 42 0.28
4 2) 23 42 0.26
5
6
7
8
Γ. 567/8 9 1 1 0 1 0
ϵ, 569/70 10 5 4 1 5 0 1) 9 2-16 0.44
11 2) 10 60 0.40
12 3) 11 0.36
13
14
Γ + ϵ [9-14] 6 4 2 6 0 1) 6 3-9 0.66
2) 8 67 0.50
3) 7 0.57
Issue Die Links Coins in Sample Obv. Dies Pairs of do Rev. Dies Pairs of dr Estimated Number of Dies Using 1) C; 2) B; 3) M.; Estimated Number of Dies Using 1) C; 2) B; 3) M.; Confidence Range Confidence Range Percent of Issue Known Percent of Issue Known Ratio of Known Dies to Calculated Dies Ratio of Known Dies to Calculated Dies
Obv. Rev. n do zo dr zr Obv. Rev. Obv. Rev. Obv. Rev. Obv. Rev.
S, + 570/1 15 3 2 1 3 0 1) 2 1.00
16 2) 3 67 0.67
17 3) 4 0.50
17 3) 4 0.50
Z. 571/2 18 7 4 6 7 0 1) 5 3-7 0.80
19 2) 4 57 1.00
20 3) 5 0.80
21 or more
22
23
24
H, 572/3 25 5 2 6 3 2 1) 2 4 1.5-2.5 2.2-5.8 1.00 0.75
26 2) 5 80 80 1.00 0.60
27 3) 2 5 1.5-2.5 2.2-5.8 1.00 0.60
28
29
⊖ 573/4 30 4 4 0 4 0
31
32
33
I, 574/5 34 7 5 2 6 1 1) 9 19 4-14 3-35 0.55 0.31
35 2) 11 21 57 0.45 0.28
36 3) 10 23 0.50 0.26
37
38
39
40
IA, 575/6 41 1 1 0 1 0
IΓ, 577/8 42 1 1 0 1 0

The issue of the tenth year could have been prompted by a whole series of events: the association of Tiberius as Caesar on December 7, 574, coincided with the decennalia of Justin II on December 15 and thus suggested a distribution greater than the usual quinquennial donative. At the same time the recently appointed praefectus Thomas may have wanted and needed fresh coins from the mint. Unfortunately too little is known about the history of Africa under Justin II to inquire more precisely into the possible relation between monetary output and the political and military situation of the province.

End Notes

a
Thubursicu Bure (Teboursouk), between 565 and 569: J. Durliat, Les d�dicaces d'ouvrages de d�fense dans l'Afrique byzantine, Collection de l'école française de Rome 49 (1981), no. 25; D. Pringle, The Defence of Byzantine Africa from the Time of Justinian to the Arab Conquest, BAR International Series 99 (Oxford, 1981), pp. 243-44. Perhaps Tingica, see Pringle, pp. 270-72.
b
Our only source is the allusion, in Corippus' praefatio to the panegyric of Justin II, to the "generous comforts" (ampla ... solacia) sent by the quaestor and magister officiorum Anastasius to the province, and this may well be biased.
c
For a discussion of the date of Thomas' appointment, see Durliat, pp. 61-62.
d
J. Durliat, no. 26, p. 62, gives a new reading of the inscription found at Sidi Gherib; Pringle, pp. 202-3 and p. 328, no. 30, puts forward the view that the inscription, despite its having been found a few kilometers from the site, came from and referred to Iunca.
a
Ch. Carcassonne, "Tables pour l'estimation par la m�thode du maximum de vraisemblance du nombre de coins de droit (ou de revers) ayant servi × frapper une �mission." Table Ronde Numismatique et Statistique, 17-19 Sept. 1979, Paris.
b
I. D. Brown: "Some Notes on the Coinage of Elizabeth I with Special Reference to Her Hammered Silver," BNJ 28 (1955-57) pp. 568-603; "On the Use of Statistics in Numismatics," NCirc 77 (1969), pp. 83-84; "Statistical Methods as a Tool in Numismatics," Cornucopiae'� (1975), pp. 33-44.
c
J. W. Möller, "Estimation du nombre original de coins," PACT 5 (1981), pp. 157-72.
d
I. J. Good, "The Population Frequencies of Species and the Estimation of Population Parameters," Biometrika 40 (1953), pp. 237-64, where image = percent known.
39
Carthage Carthage Constantinople Constantinople
Diameter Mean S.D. Mean S.D.
Flan 20.8(19.5) 0.76 20.53(20.7) 0.68
Obverse dot border 17.7(17.7) 0.73 17.12(18.2) 0.48
Difference 3.1 3.41
When compared to the data for Justinian's coins (noted here in parentheses and above, n. 9), the trend toward uniformity is immediately clear.
40
The "probable re-engraving of ⊖ over I," on BNC 2 (34) alluded to as a supplementary argument by Hahn (MIB, p. 41) is not confirmed by close examination of the coin itself.
41
W. Hahn, MIB 1, p. 48.

CONCLUSION

This description of the establishment of the moneta auri in Carthage under Justinian I and Justin II can be compared and related to the picture already drawn for the reign of Maurice.43 The vacillation between regnal year and indictional dating during these two early reigns, which is significantly illustrated by the "bimodal" coin of the next ruler, Tiberius II, with H/I (regnal/ indictional) dating, is not overcome until 597/8, when the first indictional cycle of Maurice ends and thus suppresses the confusion between regnal years and indictions which had obtained since the accession of Maurice in 582. Indictional dating of gold coins, which then became the norm in Carthage for a century, is but an external aspect of the regularity in gold issue achieved in this reign. If the lustral variations in output are set aside, the average estimate for the number of obverse dies per issue amounts to 6.12 (S.D. = 3.56) for Maurice. Under Justin II, when the trend toward regular annual issue of solidi is already in evidence, the average is only 5.36, with the 5.17 standard deviation showing a higher dispersion of the data. The irregular issues of Justinian—six in a period of 32 years—naturally reach higher individual peaks on the graph (Figure 1), but considered as a whole they surely represent a less significant amount than the gold coinages of Justin II and, a fortiori, Maurice. Plotted against the number of years of Justinian's rule in Africa, the number of obverse dies used is only on average 2.6, as against 5 for the following reigns. The significantly larger size of the issues in gold by Justinian's successors reflects the increasing regionalization which affected Africa, like other Byzantine provinces, from the second half of the sixth century.

End Notes

42
W. Hahn draws my attention to the fact that the absence of any IB coin could be explained by a shift during Justin II's eleventh regnal year from regnal (IA) to indictional dating (ΙΓ). In support of this "indictionaľ' interpretation, which was already argued in MIB 2, pp. 38 and 40-41, he notes the absence in Carthage of the star mark in the left reverse field, characteristic of the first year issue in Constantinople. In his opinion this would point to no gold being struck before the first indiction in 567. Considering the tendency to autonomy already displayed in Carthage under Justinian, I would not give too much weight to this point: the rough parallelism between gold and bronze issues appears to me more important. It should be conceded, however, that the successive bronze issues in years VIII (corresponding to the beginning of a new lustral cycle) and X clearly show that the mint was operating on a dual quinquennial basis whose cycles were alternatively of a fiscal or political nature.
43
See Morrisson (above, n. 37).

End Notes

1
J. P. C. Kent, "Gold Coinage in the Later Roman Empire," Essays Mattingly pp. 190-204. F. Dworschak, "Studien zum byzantinischen Münzwesen I," NZ 29 (1936), pp. 74-77, assumed (arguing principally from the striking of silver at Carthage as at Ravenna) that the African mint also struck gold. Recalling this hypothesis in 1950, Grierson wrote, "His reasoning is not conclusive but the fact itself seems to me probable" (emphasis added), P. Grierson, "Dated Solidi of Maurice, Phocas, and Heraclius," NC 1950, p. 61, n. 16.
2
On the reorganization of the coinage of the three metals established by these measures, see M. Amandry et al., "L'affinage des m�taux monnay�s au Bas-Empire: les r�formes valentiniennes de 364-368," NumAntClass 11 (1982), pp. 279-95.
3
The only ones identified at that time; see D. M. Metcalf, "The Mint at Thessalonica in the Early Byzan tine Period," Villes et peuplement dans l'Illyricum protobyzantin, Collection de l'école française de Rome 77 (1984), p. 114.
4
Kent (above, n. 1), p. 203.
5
J. Lafaurie, "Un Solidus in�dit de Justinien Ier frapp� en Afrique," RN 1962, pp. 167-82.
6
Even if it must be admitted that the solidus with IB (Ratto 463) cannot be a Carthaginian issue (see below, pp. 46-47, Justinian's IB Issue) as Grierson noted already in DOC, p. 159, 278 n.
7
The attribution was earlier suggested by Lafaurie (above, n. 5), p. 177, n. 1.
8
DOC 277d, with X engraved over Z, is too different from African examples (Plate 10, 32-34) to be given to Carthage. It is assigned to Constantinople by Hahn, MIB 6-⊖. Note, in fact, the similarity of the reverse to M IB 5 and DOC 3, on which the cross is likewise represented with a succession of dots, which is never the case on the solidi of Carthage where it is always represented by a line.

THE MINTING OF GOLD COINAGE AT THESSALONICA IN THE FIFTH AND SIXTH CENTURIES AND THE GOLD CURRENCY OF ILLYRICUM AND DALMATIA

D. M. Metcalf

Plates 15-24

A quarter of a century ago J. P. C. Kent presented a clear and influential survey of the administrative organization which lay behind the supply of gold coinage in the later Roman Empire. Taking as his point of departure a balanced assessment by J. W. E. Pearce,1 he deliberately emphasized one aspect, namely the government's urgent need for coined gold to pay the stipends of its soldiers and functionaries and to fulfill the heavy periodic demands of accession donatives and quinquennial donatives. These demands were met through the collection of onerous taxes payable in gold. Kent went on to describe the gathering of tax by city councils, who forwarded it to diocesan treasuries. In the fourth century the work of provincial mints was linked to these treasuries, but from 366/7 onward the striking of gold became centralized. With a few exceptions it was permitted only at the comitatus or imperial residence. The gold collected in each province was carried to the comitatus in the form of ingots (of which there are late fourth century examples with the stamps of the Thessalonica treasury),2 and there a moneta auri, which was a palatine department quite separate from the older monetae publicae, converted it into coin.

Kent saw that, within the administrative framework he described, the minting of gold at Thessalonica in the fifth century was anomalous. He suggested that the clue to this puzzling situation was to be found in similar situations at Ravenna and probably also Carthage, where the praetorian prefects were expressly assigned a detachment from the officium of the comes sacrarum largitionum, to enable them to coin small quantities of gold.3 Kent's keen insight remains the basis of our understanding of the matter today: there was a palatine moneta auri at Thessalonica because the city was the seat of the pretorian prefect of Illyricum.

Hahn has now firmly ruled out the idea tentatively floated by Bellinger and others during the 1960s and 1970s that sixth-century solidi with the mint signature CONOB might have been minted at a variety of places, in particular those where there was a moneta publica striking copper coinage. Only the palace or the seat of a prefect might have a mint for gold, "at least under normal conditions."4 The saving clause may still permit debate around Kent's idea that in the fourth and fifth centuries the mint signature COMOB referred to the comitatus rather than the comes sacrarum largitionum, and that there might, exceptionally, be a traveling mint which This study ought in truth to have appeared jointly under Wolfgang Hahn's name as well as my own. His modesty and insistence should not be allowed to conceal the fact that much of the enterprise and much of the hard work were his. I am deeply grateful to him for his encouragement and friendship. A preliminary version of this paper was published as "The Mint of Thessalonica in the Early Byzantine Period," Villes et peuplement dans l'Illyricum protobyzantin, Coll�ge de l'école française de Rome, 77 (Rome, 1984), pp. 111-28. accompanied the emperor. That, however, is relevant to our theme only because of the scarce issues in the names of Theodosius II and Honorius signed COMOB which he attributed to Thessalonica in ca. 408. It will be argued below that they are not Thessalonican, but that some of them are from a separate Illyrican mint.5

The numismatic facts about Thessalonica upon which Kent relied in reaching his assessment that its gold coinage was anomalous have been changed substantially by the recognition that the mint continued to be active long after the time of Basiliscus, indeed throughout the sixth century and into the early seventh. Its later products were not readily identified because they were marked CONOB (rather than TESOB or THSOB). They are distinguished chiefly by the addition of stars in the reverse field or by other small variations from the design in use at Constantinople. Bellinger made the attribution in DOC 1, noting that the coins lacked an officina numeral on the reverse and that the tradition of adding two stars went back to coins of which the Thessalonican attribution was not in doubt.6 Fagerlie's monograph on the Scandinavian finds, published also in 1967, proposed the same attribution, and she should probably have the credit as the originator.7 Hahn presented the case more explicitly in MIB 1, referring to the highly important late sixth century hoard of gold found at Thessaloniki itself in 1948. This hoard has recently been fully published, with excellent illustrations of every coin, by Oeconomides and Touratsoglou.8 Most of the coins are from the local mint, and the hoard site may be thought to place their attribution beyond any reasonable doubt.

Gold was struck at Thessalonica, then, steadily from ca. 408 (although the imperial mint there had its origins a hundred years earlier) until at least 616/7.9 Towards the end of that time, political circumstances in the west Balkans were altered radically by the Avaro-Slavonic incursions and the mint did not survive Heraclius' reforms of the coinage in 629. By then, it had had a special status for more than two hundred years. Until Justinian's restoration of the prefectures of Africa and Italy, it was the only moneta auri, other than Constantinople itself, in the eastern part of the empire. The essential reason for the city's place in the imperial minting arrangements was not its great size or its commercial importance—several other cities would have had at least equal claim to a mint for gold if that been so—but the fact that it was a major center of fiscal administration. Thessalonica minted gold because it was the seat of the praefectus praetorio per Illyricum.

The prefecture was a territorial circumscription with fiscal and legal competence. It was responsible for estimating the annual needs of the military forces, the civil service, and public works within its boundaries, and for collecting taxes to meet those needs. More and more, those taxes were collected in gold. From Thessalonica the whole of Illyricum was administered: if reserves of gold were accumulated, it was presumably there that they would be held; if new money was needed, it was presumably there that the need was met.10

One item in Kent's account therefore provokes hesitation. If the raison d'�tre of the Thessalonica mint was connected with the administration of Illyricum, should we not expect to find Thessalonican gold coins making up an important part of the currency everywhere in Illyricum ? Why then were only small quantities of gold minted at Thessalonica ? And what are we to understand by small? Small absolutely, or just small in relation to Constantinople?

The prefecture, which was so vast at the date of the Notitia Dignitatum, stretching from the Danube to Crete, is still described as consisting of 13 provinces, covering the same wide territory in the early sixth century, but there may by then have been some administrative devolution, and Thessalonica may have been linked more especially with a province corresponding with the old diocese of Dacia—roughly Macedonia, the two Dacias, Moesia I, and the easterly part of Pannonia.11 This was still a large responsibility.

Admittedly the prefecture of the East, administered from the capital, was (as Hendy 12 has remarked) of wide extent and had a developed, largely urban economy. One would expect the budget of Illyricum to be small in comparison. Not, however, as small as some of the hoard evidence might appear to imply: the massive treasure from Szik�ncs, which has been seen as reflecting payments to the Huns, contained 1,433 solidi minted in the eastern part of the Empire, among which only one was from Thessalonica.13 This is doubtless a very misleading measure of the relative importance of the moneta ami and merely reflects the fact that the payment in question consisted of moneys emanating from the treasury at Constantinople. A better indicator is provided by the finds from Scandinavia. Among more than 750 late Roman and Byzantine solidi of the fifth and sixth centuries found in Sweden and Denmark, nearly 400 are from the mint of Constantinople, over 200 are from Italian mints, and only 23 are from Thessalonica.14 The relative scarcity in commerce of Thessalonican solidi compared with those from Constantinople is probably much greater than the Scandinavian ratio of 17:1 would suggest, but this may be because so many of the coins reaching western European markets originated from the Middle East.

Numismatists have referred until now to single specimens or to very small numbers of Thessalonican solidi and have certainly contrived to leave the impression, intentionally or otherwise, that the scale of the Thessalonican issues was small absolutely. This was perhaps a natural assumption. Even major public collections are generally able to show only one or two specimens from each reign, if indeed they are fortunate enough to hold any at all. Study of the series has been limited for the most part to a simple listing of the different types and varieties known to exist, reign by reign.

But in order to bring numismatic research into contact with historical studies of the fifth and sixth centuries, two further steps are needed. First, a more quantitative approach is desirable. We have to ask ourselves (even if no exact answer is possible) how much gold coinage was struck. Evidence exists, in that the solidi were struck from a number of different dies. By assembling a corpus of specimens and carefully comparing their dies we can estimate statistically how large or small that number was. The project has been limited to the period from ca. 408 onward, from which it has been possible to assemble an astonishing total of 468 specimens of Thessalonican gold, among which there are few die duplicates. The original total number of dies will obviously therefore have been several times larger than the sample. This material is presented in Appendix 2, The Coins: Types and Summary, and Appendix 3, The Coins: An Inventory.

Second, we can collect evidence of how the coins were used once they had been issued. If Thessalonica was the mint for Illyricum, to what extent was the currency of Illyricum supplied by that mint? And were Thessalonican coins carried further afield? Hoards, Appendix 4, part 1, and stray finds, Appendix 4, part 2, show us how far the coins circulated, and an assemblage of all the finds of gold coinage from a region can be analyzed in a variety of ways, including a topographical analysis, which may offer clues to the character of monetary affairs. Both steps need to be taken together, because the results of the die estimation may influence our reading of the find evidence.

One must, however, draw a sharp distinction between finds from within and beyond the imperial frontiers. An entirely different set of "laws" applies to the latter. The sometimes very large hoards from north of the Danube, such as that from Szik�ncs already mentioned, may reflect tribute or booty or imperial gifts designed to impress foreign rulers. Fagerlie has shown how the flow of solidi to Scandinavia reflects the presence of the Ostrogoths in Pannonia from 454 to 488;15 and Kovačević has estimated the tribute paid to the Avars between 558 and 626 at around six million solidi.16 Once it had left the lands of the Empire, the gold was not, of course, susceptible to being drawn back by imperial taxation, although it might well return through trade. Unsettled conditions beyond the frontier may have contributed to a much higher rate of loss there or, more precisely, a much higher rate of non-recovery of hoards, so that those hoards known to us may represent a significantly higher proportion of the regional stocks of gold than the comparable find evidence from within the Empire.

As we shall see, by no means all the gold coins from Illyricum were from the Thessalonica mint, and this raises the question how those from other mints reached the places where they were lost or concealed. When most of the fourth century gold mints had been closed, and the range of denominations simplified, the scope for analysis of the find material is reduced. Any regional movements of gold coins are in detail obscured when only the Constantinople and Thessalonica mints remained at work. It is worth glancing back, therefore, to some Illyrican hoards of the fourth century to see what mints are represented in them. Similarly the sixth century copper coinage affords a fuller opportunity to analyze regional monetary circulation, because more mints struck copper.

image

Fifth- and Sixth-Century Finds of Gold Coins from Northern Illyricum

A type of fourth-century hoard, characteristically found just within the frontiers, e.g. along the Danube, is marked by the presence of multiples, and may be associated with the army and even with the presence of the emperor—as in the case of the Ljubljana (Emona) hoard of 1956, consisting of triple and double solidi of Constans and Magnentius, minted at Aquileia (12 specimens) and Thessalonica (1 specimen).17 A Constantinian hoard from Sirmium18 is more mixed in its composition, but includes a half dozen multiples, from the mints of Nicomedia (4) and Thessalonica (2). Like the nearby Borca hoard, which also includes multiples, it has nearly a half pound of gold, and the point has been noted that this was the gratuity paid to discharged veterans.19 When a hoard from the frontier regions includes just one or two multiples of a less spectacular size, such as the piece of 1 1/2 solidi in the U�će hoard, yet again from the vicinity of Sirmium,20 one might similarly think of donatives or stipends paid to military officers, although the coins may have changed hands before being concealed, and the last owner would not necessarily have been an official.

The Ljubljana hoard should be discounted as evidence for the circulating medium, for the group of multiples had clearly stayed together since its issue, probably in the possession of the recipient; but it is less obvious how far the U�će hoard should be discounted, since the presence of an army is a textbook example of a way in which large amounts of money may be injected into a regional economy that would otherwise be largely self-contained. The Borca hoard in any case proves that within at most five years from their dates of issue, a small group of coins on the Danube frontier could include examples from seven mints as far apart as Antioch and Rome. Similarly the U�će hoard of only five coins includes examples from Alexandria, Nicomedia, Siscia, and Trier, and this raises in an acute form the question of how the majority of the gold coins in the finds reached the places where they were concealed. Indeed it may raise the further question of whether gold was (in the fourth century) regularly reminted before being reissued by the authorities. How rapidly did gold pass from hand to hand in private transactions, once it had been put into circulation through the payment of officials? For example, the Ljubljana hoard of 1910, from the mid-fourth century,21 is larger and therefore offers even better evidence of the way in which the issues of different mints were carried from province to province, from Italy to the Balkans, and even from Antioch to the Balkans, becoming mingled with remarkable rapidity. The bulk of the hoard consists of coins minted from 337 to 340 and although the nearby mint of Siscia is strongly represented, so are Trier, Aquileia, Thessalonica, Heraclea, Constantinople, Nicomedia, and Antioch. The one older coin in the hoard is from Rome. It may well be, then, that the reduction in the number of gold mints in the fifth and sixth centuries tends to conceal the degree of geographical movement of the coins.

As regards copper coinage, the Thessalonica mint sometimes supplied a major part, although never all of the Illyrican currency. Serbian sites have yielded a proportion of 58 percent for the years 565-82, but the figure was much lower than that both before and after. The proportion in central Greece under Justin II and Tiberius II, as Popović has pointed out, shows an exactly similar peak of 58 percent.22

Hoards and isolated finds of gold from well within the frontiers of Illyricum should be regarded as normative for the purpose of this enquiry. We shall also look at finds from Dalmatia, even though that province did not become part of the Illyrican prefecture until late in the sixth century.23 There are a surprising number of finds (mostly from the time of the Avaro-Slavonic invasions) but unfortunately few of them have been adequately published. If they have been dispersed without detailed record—and that means inter alia a record which permits individual dies to be checked against future finds—the finer points of their evidence have inevitably been lost.

Once cannot foresee when or how progress will be made: another major hoard from Illyricum, like that found at Thessaloniki in 1948, may come to light next year; or alternatively very little new evidence may turn up for two or three decades. Meanwhile, an outline view can be sketched which further discoveries are unlikely to modify substantially.

Kent's emphasis that gold coins enjoyed a short life in circulation, before being clawed back into the imperial coffers by the ruler's device of insisting that taxes should be paid in gold, may possibly be true of other parts of the Empire, but it would not be an obvious conclusion to draw from the Illyrican finds. Nor would they lead us to suspect that Thessalonica had any important role as the moneta auri for the official finances of Illyricum. If a short life had been the norm, one would have expected gold hoards to be generally small and few and, more surely, to consist largely of almost new coins, including numerous clusters of die duplicates. Also, one would have little reason to expect that coins would have had the chance to travel far (except when armies were on the move) before they had been recovered in taxation, to be melted down into ingots and in due course reissued as new coins, to begin another short life cycle.

Yet the age structure of the hoards reveals that a significant proportion of the currency consisted of solidi that had remained in circulation for more than ten years. The age structures are, in fact, not markedly different from those found in hoards of other types of coinage, where there is no reason to postulate urgent withdrawal. The six coins of Theodosius II which make up the Po�arevac hoard from Moesia I, for example, are of types dated by Hahn as follows: one from 408-22, three from 422-25, one from 430, and one from 443-50. Only one coin in the Po�arevac hoard—the latest—is Thessalonican.24

Various late sixth century hoards from Illyricum show a similar age structure. Three finds from Sadovets, in Dacia Ripensis, are made up as follows:

Hoard B (34 Ꜹ) Hoard B (34 Ꜹ) Hoard C (25 Ꜹ) Hoard C (25 Ꜹ) Hoard D (125 Ꜹ) Hoard D (125 Ꜹ)
Sol. Trem. Sol. Trem. Sol. Trem.
Anastasius - - - - 1 -
Justinian 5 2 3 1 22 59
Justin II 9 2 8 1 4 16
Tiberius II 2 10 6 2 6 14
Maurice 3 1 4 - 2 1
Total 19 15 21 4 35 90

Two interesting features of the hoards (which may well have been concealed in haste in the face of the Avar assaults of 584/5 and the capture of Singidunum and Viminacium)25 are the number of tremisses they contain and the wide variation in the proportions of solidi and fractions. Hoard D has a particularly extended age structure, and consists mostly of tremisses. Not a single coin in these three hoards is Thessalonican. If the suggested date of deposit is correct it shows that coins of Maurice reached Sadovets from Constantinople within at most two to three years of their issue.

Three late sixth-century hoards from Dalmatia, possibly unrecovered because of the dangers following the Avar attacks on Sirmium in 568 and 569 and the loss of the city in 582,26 yield very similar conclusions. The Grabovnik find (from ca.569?) contained four solidi of Justinian and 15 of Justin II. All except one pair were from different dies, and only one coin was Thessalonican.27 The Vid hoard, from the site of the ancient Narona, contained 24 solidi of Justinian, a further 24 solidi and six tremisses of Justin II, five solidi and five tremisses of Tiberius II, and one tremissis of Maurice. This last coin provides a terminus post quern of 582, and shows that a third of the coins were about 20 years old. At least two seem to have been from Thessalonica, but probably not more. Die duplication was not recorded, but the wide range of officinae represented suggests a well-mixed sample of currency. The Solin hoard (from ancient Salonae) consisted of one coin of Justinian, one of Justin II, and three of Tiberius. Again, if the suggested occasion of concealment is correct, the latest coins had reached the west Balkans swiftly.

From Greece there are remarkably few records of gold hoards dating from before the closure of the Thessalonica mint. The Athens (Osteotheke) hoard consisted of five solidi of Tiberius and one solidus and one tremissis of Maurice. The coins of Tiberius are all Constantinopolitan; the solidus of Maurice was not sufficiently described for its mint attribution to be certain.28 Other Greek finds, including those from the Peloponnese, Paiania, and Corinth,29 were incompletely or summarily published, at a time when no one had thought of the attribution of post-Zeno coins to Thessalonica.

The best systematic evidence for the gold currency of Byzantine Illyricum lies in a rather unexpected direction, namely in odd coins, usually without an exact provenance, bought by local museums long ago. A catalogue of nineteenth-century finds from Bosnia is particularly precious, because it preserves evidence from a period when the coins in question were worth little more than their bullion value, and when a curator with an active interest in local finds could hope to recover pieces that today would probably not be reported to the authorities. Karlo Patsch in 190030 was able to list 55 gold coins ranging from Arcadius to Justinian, most of which were bought in Sarajevo, while others came from a collection formed in Imotski, or were from Izačić, Kladanj, Skender Vakuf, �upanjac, Stolac, Glamoč, �urmanci, Knin, Donji La рас, the Vranplanina, or the Neretva valley. Many are probably single finds, while others may be from unrecorded hoards. A good proportion have been pierced, perhaps in modern times, for use as jewelry. Among the solidi in the Sarajevo Museum, only one was minted at Thessalonica.

When all the recorded provenances of fifth and sixth-century gold are marked on a map of northern Illyricum, a rather unexpected pattern emerges. The finds are scattered most densely in Dalmatia and its hinterland, far to the northwest of Thessalonica. The pattern is no doubt distorted by the fact that such a large proportion of our evidence was collected in Sarajevo. If there had been another Karlo Patsch in Belgrade, the map would look different. It is fair to point out, however, that the Sarajevo provenances mostly lie to the west of the Bosnian capital. The concentration on Lika, Dalmatia, and the Neretva valley is probably a genuine distribution. Moreover, the Ostrogothic coins in the Zagreb Museum, recently published by Żeljko Demo,31 show a similar dual distribution, with numerous finds from Dalmatia and others from Syrmia (administratively within the scope of the Zagreb Museum) but little from the intervening parts of Croatia.

Ostrogothic coins are unexpectedly frequent among the finds now in Sarajevo, from the time of Zeno to Justinian—long after the administrative link between western Illyricum and the prefecture of Italy had been severed. Dalmatia evidently maintained links with Italy, and the hinterland with Dalmatia. Specimens are recorded from Skender Vakuf, the Morača valley, and �upanjac. There is a contrast with the evidence of, for example, the Grabovnik and Vid hoards, which are close enough to the Adriatic to have shared in the pattern, but which consist essentially of coins struck in the third quarter of the century. One may wonder whether the inflows of Ostrogothic money ceased in about the 540s, or whether the hoards reflect official payments made exclusively in imperial coin. The former explanation seems preferable, especially where there are tremisses in the hoard. This makes the point that the finds are a more evenly spaced sample of the currency than the hoards, which are concentrated so much in the 580s and contain few coins from the first half of the century.

The Bosnian material also very usefully confirms the high proportion of fractions in circulation. Nearly half the catalogued coins are tremisses. This makes a very pronounced contrast with the Scandinavian finds, where fractions are most unusual. It is a familiar idea from many ancient and medieval series of coinage, of different centuries, that people preferred to exclude the fractions from sums of money set aside, and that, unless we are fortunate enough to be able to refer to isolated finds (accidental losses), there may be an important sector of the currency which remains "hidden."

The mint attribution of tremisses is a particularly delicate problem. Where previously none were given to Thessalonica, Hahn has found a few which show close stylistic affinities with copper coins of the same mint. The stray finds from Illyricum (and indeed the tremisses in the hoards) deserve to be reexamined carefully to see whether more local issues can be distinguished. But as most of the solidi are Constantinopolitan, there is no a priori reason why the tremisses should not be so too.

Another group of finds, smaller, but of equally solid value as evidence and again from the heart of Illyricum, comes from Leskovac, near the site of Caricin Grad, in the Morava valley above Ni�, where the museum staff published a list of their holdings shortly after the 1939-45 war.32 Ten gold coins ranging from Theodosius II to Phocas have provenances in the environs of Leskovac or, in one case, at Ora�će (Gornje Lusarje). Again, half of the pieces are fractions; two semisses and three tremisses. Only one solidus, of Theodosius II, was minted at Thessalonica.

image

Fig. 1

If, putting together the Sarajevo and the Leskovac material, we try to assess the chronological distribution of the finds by dividing the number of gold coins from each reign by the number of years of the reign, the resulting graph (Figure 1) will only be rough and ready, both because the totals of coins are small and because no allowance is made for the length of time over which different issues remained in circulation. The trend which can be discerned is nevertheless sufficiently pronounced to be worth considering: gold was relatively plentiful from the middle of the fifth century until about the third quarter of the sixth century.

The proportions are repeated in the holdings of the Split museum, built up since 1820, and in which Marović reports 11 gold coins of Honorius, 29 of Valentinian III, 22 of Leo I, 26 of Zeno, 26 of Anastasius, 84 of Justinian, 30 of Justin II, and 15 of Tiberius II.33 The coins of Valentinian, obviously, are Italian, and it would be interesting to know how many Ostrogothic coins there are among the sixth-century totals.

Single finds do not, of course, provide any information about the age structure of the currency, but there is one further hoard which confirms what has been said above and also reiterates the importance of the tremisses. The Thessaloniki find of 1948,34 concealed ca. 580, was almost certainly abnormal in its composition, in that it contained a very high proportion of coins of the local mint and among them a great many die duplicates or die linked coins.

Constantinople Constantinople Thessalonica Thessalonica Ravenna Ravenna
Sol. Trem. Sol. Trem. Sol. Trem.
Justinian 14 7 3 1 - -
Justin II 19 9 42 14 1 -
Tiberius II - - 4 1 - -

The figures set out in the table suggest that a progressive dispersion of the gold currency took place. From the reign of Tiberius II, all five coins in the hoard are from the Thessalonica mint, but from the preceding reign 3 percent of the solidi and 39 percent of the tremisses are from Constantinople, and one coin has found its way to Thessaloniki from Ravenna. From the reign of Justinian (that is to say, among coins that have been in circulation for some 15 years or more, which make up over a fifth of the hoard) 82 percent of the solidi and 88 percent of the tremisses are from Constantinople.

The extensive die linking of the Thessalonica mint coins of Justin II and Tiberius, both solidi and tremisses, in contrast with the absence of die links among the Constantinopolitan coins, shows that the hoard includes a sum or sums that had not changed hands much since they left the mint. The coins of Justin might in principle be from the very end of his reign, but it is unlikely that they all are, as there are several varieties among them. One needs to compare them with specimens from other sources in order to establish the character of the hoard.

We should hesitate to deduce from the hoard of 1948 that the composition of the gold currency in Thessalonica itself was greatly different from that elsewhere in Illyricum during the fifth and sixth centuries. Not only would such a scheme of monetary circulation be contrary to normal expectations, but the hoard itself reveals a progressive dispersal through the third quarter of the century. It is however possible that the hoard reflects an unusual confinement of circulation to the city of Thessalonica at the time of an incursion from the Sclavinias, to which Menander and John of Ephesus both refer, and which is most plausibly dated to 581.35

Discounting this hoard and leaving to one side the Hajdu�ka Vodenica hoard, the evidence is clear that, throughout Illyricum, the solidi of the Thessalonica mint made up only a small proportion of the available gold currency. The same is in all probability true of semisses and tremisses. (Elsewhere in the Empire, Thessalonican coins were even less significant, to the point that they have very rarely been recorded.) We are left, then, with unanswered questions about how the majority of the Illyrican gold finds, which are coins minted at Contantinople, reached the more westerly prefecture.

In terms of social structures and monetary responses there is a distinction to be drawn between the finds from Dalmatia and the Adriatic coastlands and those from north-central Illyricum, particularly from the time of Justinian onwards. One's impression is that in Dalmatia interregional trade exerted a strong influence on the circulation of gold. Along the Danube, on the other hand, and in the regions towards the frontier, finds of gold certainly tend to be concentrated in or near the major garrisons. The corollary is probably that long distance circulation was sluggish and local "pools'' of gold existed, as a spin-off from the military presence, in regions where a money economy was otherwise of limited scope. The same phenomenon has been observed in, for example, the Dobrogea.36 From Illyricum, we have noted hoards from Borca and U�će, both in the vicinity of Sirmium, and from Sirmium itself. Other finds are from Sisak (Sicia), Po�arevac (Viminacium), Gornja/Donja Vre�ina (a castellum near Naissus), and the Leskovac district (Prima Justiniana).

Elsewhere in north-central Illyricum, hoards tend to be from urban sites (e.g. Stobi and Bargala). Popović has drawn attention to a contrast between coin hoards up to the end of Justinian's reign (many of them from rural localities) and those from the time of Justin II and Tiberius (from fortified sites). To some extent, therefore, we may be looking at a chronological rather than a regional distinction.

The pattern seems not to have been unduly influenced by the choice of sites to excavate, and it is probably fair to say that the local availability of gold coinage in north-central Illyricum reflects the stationing of soldiers and officials, particularly from the 560s onward. One can see, therefore, how the composition of the currency could be different from one place to another, and (to return to the matter that prompted these comments) how the Thessaloniki hoard might after all be typical of the city's currency.

The survey of provenanced material also leaves us in a position to appreciate the implications for the monetary history of Illyricum of the observation that the dies of the Thessalonican solidi are almost all different. The corpus of specimens, Appendix 3, includes the following totals:

Sol. Sol. Trem.
Theodosius II 130 Justin I 9 -
Marcian 20 Justinian 85 3
Leo I 61 Justin II 52 15
Basiliscus 10 Tiberius II 9 2
Zeno 15 Maurice 6 7
Anastasius 33 Phocas 1 -
269 Heraclius 10 -
172 27

Some of these 468 specimens are in public collections, but most have been gathered from illustrated auction and sale catalogues from the photo file of the Austrian Academy of Sciences. In general, these coins have disappeared into private collections, and their present whereabouts are unknown.

Since there is no reason to doubt that the available coins can be treated (hoard material apart) as a random sample in respect to the different dies represented among them, statistical procedures can be applied to the material in order to derive an estimate of the total numbers of dies originally used. It will be prudent to explain these procedures in some detail, in order to make quite clear what is fact and what is conjecture and to show what reseach will be needed in the future if we are to make further progress.

The most convenient formula37 states that the number of obverses in a random sample that are linked to other obverses in the sample is a measure of the proportion of the total output reflected by known dies. Thus, if among 30 coins there were 4 unique obverses, the formula would indicate that the dies in the sample represented 26/30 or 87 percent of the original dies. Further, if 14 different dies were represented in the 26 multiple obverses responsible for 87 percent of the coinage, then each die struck on average just over 6 percent of those coins. To estimate the total number of obverse dies used for striking the series, at just over 6 percent per known obverse, then slightly over 16 obverse dies would have been needed for 100 percent of the coins, suggesting that 2 obverse dies may be still unknown. If the missing dies happened to be dies that were underused, 2 "equivalent" dies might be represented by 4 or more actual dies. In using the formula, we should be careful to speak of "equivalent" missing dies rather than actual dies as a measure of output.

If, on the other hand, an issue is represented by 8 specimens among which there is only one pair which share a die, the outcome of estimation is much less conclusive. The sample represents only 25 percent of the original output, and the numbers are too small for any confidence to be placed in their accuracy. If the next specimen to turn up yields another die link, the estimate changes from 25 percent to 44 percent. The sample, in a word, is too small. This is characteristic of the position in which we still find ourselves in studying the Thessalonican gold. In spite of being able to present a corpus of over 450 specimens, the total is insufficient because such a small proportion of the dies originally used are known to us. As new hoards come to light, it will be necessary to rework the corpus, checking systematically for die duplication, until an estimate is reached with a level of accuracy in which we can be reasonably confident.

Meanwhile, it is already clear, in spite of the wide margins of sampling error, that there were originally some hundreds of pairs of dies. If this is what is meant by saying that Thessalonica struck solidi in small quantities, what amount should we expect from the Contantinople mint? And if hundreds of pairs of dies were used, where did all the coins go, seeing that they make up such a small proportion of the currency of Illyricum? Was the total stock of currency in Illyricum so large ? Or were payments across the frontier a very severe drain on the Empire's resources ?

The volume of the imperial coinage in the fifth and sixth centuries is of some general historical interest as an index of the level of monetary activity in late antiquity. Gold was the major part of the money supply, in terms of its book value, and accordingly it should have first claim on our attention. Counting dies does not, unfortunately, provide as simple a way of measuring the size of the currency as might appear at first sight.

First, there is absolutely no way in which we can know how many coins, on average, were struck from each pair of dies. Technically, a reverse die was probably capable of striking on average 20,000 solidi or at least as many tremisses, and it is not easy to see why a mint would have gone to the trouble of preparing and engraving far more dies than were needed. If the prefecture's budget envisaged an issue of a certain size, it should have been possible to know in advance roughly how many dies would be needed. And if an issue was prolonged, further dies could be cut as needed. On the other hand, there is very clear evidence from English medieval coinage that the dies for gold were used far less heavily than those for silver, for no obvious reason.38 The conventional figure of 10,000 coins on average from each reverse die, or a million solidi (14,000 pounds) from a hundred dies, is well within their technical capacity. They could have been used to that extent; but we cannot prove that in fact they were. It may be more acceptable, therefore, to deal simply in numbers of dies and to make comparisons, for example, between Thessalonica and Constantinople in those terms—although, again, it would be very difficult to demonstrate that average output per die was similar at the two mints.

Second, it is the size of the currency rather than mint output which is of interest as a measure of monetary activity. If we cannot say whether the average life of a solidus was ten years, or five, or twenty, we introduce another factor of uncertainty of exactly the same size as that which arises from not knowing whether on average a die struck 10,000, or 5,000, or 20,000 coins.

Third, the velocity of circulation is just as much a factor in the effective money supply as the size of the monetary stock.

With all these reservations in mind, the estimated numbers of reverse dies used at the Thessalonica mint may be calculated as follows:

Sol. Trem Sol. Trem.
Theodosius, ca. 408-50 640 Justinian, 527-42 400? -
Marcian, 450-57 60/100? Justinian, 542-65 525? 5?
Leo I, 457-74 320 Justin II, 565-78 140 30/35
Basiliscus, 475-76 20/40 Tiberius II, 578-82 20? 5?
Zeno (second reign), 476-91 50/80? Maurice, 582-602 20? 20?
Anastasius, 491-518 300/400? Heraclius, 610-ca.617 30?
Justin I, 518-27 20/50? -
image

Fig. 2

Comparative mint output over the two-hundred year period offers a broad context within which to explore the probable reasons for minting gold in Illyricum. If our concern extends to the scale of the currency, the simplest way to present the figures is to divide the numbers of dies into the lengths of the reigns, to obtain a measure of average output per year. Figure 2 shows that it varied widely, and in practice the fluctuations were no doubt even greater, since the figures for periods of fifteen or twenty years will conceal years during which the mint was relatively inactive. To move from a stylistic analysis of the coins and a relative chronology toward an absolute chronology, and thus to refine the diagram, would be an extremely speculative and difficult exercise. In principle the only way forward is through the discovery of hoards which can be very plausibly associated with particular historical events part way through an issue.

From the reign of Theodosius II there are, fortunately, four separate issues of solidi at Thessalonica, which can be dated from their types and legends, in conjunction with a fuller series of issues at Constantinople. The sharp contrast between the level of output from 408 to 440 and in the 440s attracts attention and seems to call for some special explanation. Hahn's reading of the evidence39 is that there were two issues at Thessalonica in the 440s. The VIRTVS EXERC ROM issue, which is exceedingly scarce, is assigned to 440 or later. The GLOR ORVIS TERRAR issue, for which hundreds of dies were used, falls in the years 443-50.

Since there are problems about the attribution, and since its correctness is crucial to a monetary interpretation of the coinage, the difficulties had better be understood. Kent suggested40 that the GLOR ORVIS TERRAR issue belonged to the sole reign of Theodosius (423-25), and that it was more plentiful at Thessalonica even than at Constantinople because Valentinian III was proclaimed Caesar there in 424. Hahn however would date it from 442/3 onwards, noting that the same type was continued by Marcian. He finds many reasons for interpreting the IMP XXXXII COS XVII P P issue at Constantinople and elsewhere as being destined for the payment of tribute to the Huns—a tribute which reached catastrophic levels in 447. Perhaps the most cogent evidence for a late date for the GLOR ORVIS TERRAR issue (which could, after all, have been revived after an interval rather than directly continued by Marcian) comes from the Szik�ncs hoard and from a Bulgarian hoard published by Voirol, each of which includes none or hardly any specimens of GLOR ORVIS TERRAR against large quantities (247 in Szik�ncs) of IMP XXXXII COS XVII P P. The earlier issues, from the 420s and 430s, are present in sufficient numbers for the absence of the plentiful GLOR ORVIS TERRAR issue to be problematic if it were dated to 423-25. If on the other hand it was a very recent issue, its absence is perhaps explicable.

Still, the question remains why the Thessalonica mint should have struck a military type in great quantities during the years 443-50, having previously been far less active. The finds from beyond the imperial frontiers (including the Scandinavian finds) give no encouragement to think that the solidi were destined for the payment of tribute. Nor have the GLOR ORVIS TERRAR coins been found in conspicuous quantities in northern Illyricum (note the contrast between Figures 1 and 2). The reason may lie closer to home.

Vickers has sketched the extensive program of building which followed the transfer of the seat of the prefecture to Thessalonica, after the destruction of Sirmium by the Huns in 442/3. The city needed to be equipped with buildings appropriate to its new status: there had to be a new palace in a less vulnerable position than the relatively exposed site of the Tetrarchie palace on the city's eastern edge; up-to-date fortifications were required; and churches of imposing proportions had to be built. Spieser has, however, criticized Vickers' interpretation, questioning the extent of this building program. He would argue that, of the major works, only the walls are of mid-fifth century date, and that the conversion of the rotunda into a church and the construction of St. Demetrius took place more probably at the beginning of the sixth century. In any case Vickers went on to argue (and the point may still be valid) that the widely found brick-stamp A-ENT reflects a crash program of buildings works and refers to the first year of an indiction (Sept. 447/8) during which Hormisdas, later Pp Orientis (Feb. 448-Jan. 449; late 449-April 450) may have been Pp Illyrici.41 Our list of prefects of Illyricum is sketchy for the 440s, so that, although transfers from one prefecture to another were not usual, it is not impossible that the Hormisdas of the much discussed inscription in brick on the eastern wall of the city was the same man. It is puzzling that there should have been a delay until 447/8. A possibility, which seems not to have been considered so far, is that Thessalonica may have been badly affected by the earthquake of 447.42

Numismatists will be intrigued by the assumption that, until 442/3, the seat of the prefecture was at Sirmium. If that were so, should one not expect, in accordance with the theory of moneta auri discussed above, that the earlier coins of Theodosius would have been minted at Sirmium, and not at Thessalonica ? The evidence for the whereabouts of the prefect in the first part of the century is tenuous. Stein expressed the opinion that from 395 the seat of the new Illyrican prefect was at Thessalonica; and it is difficult to imagine where else it could have been, unless at Stobi, for Sirmium seems not to have been transferred to the East until some time between 424 and 437, and more probably the latter date.43 If that is so, Sirmium could have been the seat of the prefecture for only a short time before 441/2. It may be, therefore, that the only issue of gold coinage which raises a problem is the VIRTVS EXERC ROM type dated by Hahn to the years 440-43. At Constantinople it was minted by nine if not ten officinae, but at Thessalonica it is very rare. The other horn of the dilemma is to know why such an extensive program of buildings was required in 447/8, if the prefecture had only recently left Thessalonica.44

All these difficult and inconclusive arguments about events in 447/8 which might provide a context for the copious GLOR ORVIS TERRAR issue are prompted by the virtual absence of the type from Szik�ncs and other hoards, and its scarcity in Scandinavia. The choice of a military type is a serious obstacle to any theory that expenditure on public works accounted for the large size of the issue. Perhaps an explanation for both the type and the size of the coinage is to be sought in the first expedition against the Vandals.

The next problem is Leo's issue of consular solidi on a large scale at Thessalonica, but not at Constantinople. The coins cannot be assigned to one rather than another of Leo's four consulships (458, 462, 466, 471). The issue is specifically non-military in type, and thus neither the second expedition against the Vandals in 461 nor the installation of Anthemius in Italy in 467 has any obvious connection.

The large estimated number of dies for the twenty-month reign of Basiliscus also seems to call for some special explanation. There is no obvious event in the history of the city of Thessalonica itself to which the coinage could be referred. If the mint was active in order to provide accession donatives, Basiliscus' coinage may not be significantly different from that of longer reigns, where the average number of dies per annum may conceal a similar phase of early activity. It should also be noted that the estimate for Basiliscus rests on a very inadequate statistical basis.

Where more than one variety was issued during a long reign, such as Justinian's, one should resist the temptation to establish an absolute chronology which mirrors the relative chronology indicated by stylistic variation. Most of the dies might have been used within a brief period, or their use might have been spaced out evenly through the reign, and it is exceptionally difficult to obtain evidence one way or the other. The association of varieties with quinquennial periods on general or a priori grounds is a stimulating exercise provided its speculative character is recognized. For Justinian's Thessalonican solidi, stylistic arguments taken together with the hoard evidence cast doubt on the proposed sequence of MIB 20 and 21 and on Hahn's dating of them before and after 537 respectively (see Appendix 2). The date of the change to MIB 22 is likewise conjectural, although we may be confident that it was after 538. MIB 22 is absent from the Hajdučka Vodenica hoard, in which the latest Thessalonican coins (and they are present in considerable numbers) are of MIB 20.45 The site is on the south bank of the Danube, and the shattering attack by the Kutrigurs in 540, in the course of which thirty-two Illyrican fortresses were captured, might seem the obvious occasion for its deposit—were it not that the hoard apparently includes MIB 7, which Hahn dates to 542 onwards, in connection with a weight reduction of the copper coinage and consequent changes in the design of light weight solidi. One should think rather, therefore, of the second appearance of the Kutrigurs in Illyricum in 544, or even of later raids, such as that in 548 when the Slavs crossed the Danube into Illyricum and penetrated as far as Dyrrachium.46 There were raids almost every year at this time, and those of which we have records (e.g. in 550, past Ni�, or in 559, through Moesia to Macedonia and Thessaly) are not the only ones which could have affected the Danube frontier. Thus the chronology of the coins and the historical circumstances in which the hoard was concealed are both somewhat problematic.

Of events in Thessalonica itself in Justinian's reign we know very little. Perhaps we should be looking there, rather than on the northern frontiers of Illyricum, for a monetary context for the locally minted gold—as has been suggested above for the GLOR ORVIS TERRAR coins of Theodosius II. The relatively large quantities of gold minted at Thessalonica might reflect simply the general prosperity of Justinian's reign, or they might be linked more directly with the program of defensive building works. There is very little in the pattern of finds (other than the Hajdučka Vodenica hoard itself) which favors the latter theory: none of the Sarajevo or Leskovac coins of Justinian is from the Thessalonica mint; there are no other frontier provenances; and there are not even many specimens in the regional museums—out of 38 examples in MIB 22, only one is in Belgrade and another is in Zagreb.47 The reigns of Marcian and of Anastasius, though not as lengthy, are just as uncertain as regards the chronological spread of the solidus dies and the reasons for minting.

The pattern of the late sixth century is of declining gold output from the Thessalonican mint—and almost certainly a sharp decline relative to Constantinople. Since the issues of Tiberius II were certainly on a small scale, the change antedates the year 578, and thus would seem to antedate the various attacks and disruptions in the western Balkans to which the sources refer. Popović has pointed out that this is perhaps because the authors are silent about Sclavinian incursions in the period 558-78, rather than because there were no disturbances.48 He has drawn attention to no fewer than ten hoards with closing dates between 567/8 and 571/2. Five of the ten are from Caričin Grad (Prima Justiniana), where there is archaeological evidence of coins of the 570s from late walls.49

The Thessalonican gold of Maurice is equally scarce today, as is that of Heraclius. It is unlikely that it was ever plentiful, for it would have found its way into the hoards of the period—including, one must assume, a certain number of hoards dispersed without record. The curve of mint output thus offers few insights into the purposes for which gold was coined. It is to be hoped that the detailed analysis of future hoards will clarify the patterns of monetary circulation.

Another aspect of the mint's work deserves mention. The ratio in which reverse and obverse dies were used was not always the same. Under Theodosius II and Marcian the ratio seems to have been close to parity, but from then on several reverse dies were used with each obverse, until 491. Anastasius seems to have reverted to a 1:1 ratio, but under Justinian and Justin II it was certainly more, varying between about 3:1 and 5:1. The tremisses of Justin II and of Maurice, however, employ a 1:1 ratio. It is intriguing that there should have been different practices in the mint concurrently, and one wonders whether tremisses were not supplied to the public in small quantities on demand.

APPENDIX 1

Cruces in vestibus: The Question of a Comitatensian Mint in Illyricum, and the Events of 407/8

Kent discerned a moment of change at Thessalonica in a series of solidi which (in his view) make the transition from the mark COMOB to TESOB. "The first pieces," he wrote, "assignable beyond possible doubt to that mint are of the CONCORDI AAVGG issue of ca. 408-ca. 420 marked image and image. Thereafter, at least until the reign of Basiliscus, Thessalonica coined gold in small but continuous quantities."50

The COMOB coins in question, which are eastern, or at least not Italian, make up a small series with a highly distinctive style of portraiture with a chrismon superimposed on the emperor's cuirass. Lafaurie 51 and Hahn 52 have followed Kent in assigning them either to Thessalonica or at least to the Illyrican prefecture. The series began, in fact, under Arcadius.53 Lafaurie identified three varieties, the first with AVGGG (for Arcadius), the second with AVGG (for both Arcadius and Honorius), and the third with AVGGG and a star added in the field (for Honorius), which corresponds with Hahn, Theodosius II, 53a. There is a fourth variety in the same style of engraving, but in the name of Theodosius II (Hahn 52a). The chrismon has been interpreted in the light of a text attributed to Quodvultdeus, bishop of Carthage, which tells how Arcadius received a sign of victory against the Persians. As his soldiers were going into battle, crosses appeared on their garments. Afterward he commanded that a gold coin should be struck, bearing this same sign of the cross.

It is a matter of conjecture which has somehow become enshrined as fact that the crucigerous coins with COMOB and those with TESOB follow each other in a continuous series, as Kent suggested and Hahn endorsed. Yet the coins with the two different mint marks are not by the same hand. Those marked TESOB are not in the distinctive style of portraiture, and they do not bear the chrismon on the cuirass. Why, then, have the former been so confidently assigned to Thessalonica?

Lafaurie, while preferring to see the coins with the cross as "probably Thessalonican" (1973), offered in an earlier article (1963)54 the hypothesis of a third mint, "dont les �missions sont rigoureusement parall�les × celles de Constantinople et de Thessalonique." Given that the coins are not western, i.e. Italian or Gallic, this of course has implications for the theory that monetae auri are strictly palatine. Lafaurie does not shy away from the difficulty: "Cette permanence fait plut�t penser × un atelier, dont la stabilit� est attest� pour une quinzaine d'ann�es, qu'� un atelier itin�rant. On remarquera particuli�rement que, si le bouclier que tient l'empereur est toujours orn� × Constantinople par un cavalier qui parait �tre l'empereur chassant, × Thessalonique, pour la m�me p�riode, le bouclier est orn� d'une Victoire, distinction qui individualise les produits des deux ateliers concurrement avec la marque de revers. II est possible de penser que la croix chrism�e caract�risee un trosi�me atelier...." To this argument, it should merely be added that the Victory at Thessalonica is normally a naval Victory, as she has her foot on a prow,55 and that this detail is very much a temporary feature of the mint's issues.56

There are only two recorded provenances for the Cruces in vestibus series. One is a small hoard from Asturias, in which there was one specimen along with Italian and Gallic issues, concealed ca. 408-10. The other is a hoard believed to come from southern Tunisia, in about 1958, which considerably enriched the body of surviving materal.57 These two provenacess are already a considerable obstacle to an Illyrican attrbution.58 Dare one say that the obvious attribution is Carthage? Alone of the imperial Byzantine mints, its coins are characerizedd by a prominent cross or chi-rho on the cuirass after being recovered from the Vandals in 533; and Quodvultdeus was bishop of Carthage.

The dating of the series must rest, in default of better hoard evidence, on the assumption that AVGG and AVGGG were (as was still the practice at that date)59 used correctly. AVGG should refer to Honorius and Arcadius, 395-402, AVGGG or AVGGG* to Honorius, Arcadius, and Theodosius, 402-8, and AVGG* to Honorius and Theodosius, 408 onward. The series thus lasts a minimum of six or seven years. If it is Carthaginian, the natural date at which it might have begun is 398, when the African revolt ended following the defeat of Gildo, the authority of Honorius was reestablished, and there was a period of entente between East and West. Africa remained within the praetorian prefecture of Italy-Illyricum-Africa, and Carthage was not the seat of a prefect. For that reason, perhaps, the mint signature COMOB was employed (in what was, in any case, technically a western province).

No evidence has so far been presented which would justify an Illyrican attribution of the COMOB coins of the Cruces in vestibus style. There is, however, another, stylistically quite separate series of COMOB coins (which were amalgamated by Hahn with the crucigeous series and attributed, along with them, to Thessalonica). These appear to be relatively scarce. They are known for both Theodosius II and Honorius, with AVGG.

image

Fig. 3

This second small series of COMOB coins has a separate (and non-Constantinopolitan) pattern ornamenting the cuirass, in particular a row of dots on stalks (Fig. 3) and a very distinctive picture on the shield, namely a Victory on globus (although on one of the two specimens all that can be seen of the globus are the two dots, one on each side of the Victory's feet). There are two specimens which link COMOB and TESOB. One is a coin of Theodosius in Vienna, which has the row of dots on stalks ornamenting the cuirass, but in combination with a naval Victory on the shield. The reverse of this coins reads AVGG and the mint signature is TESOB. The other specimen is in the name of Honorius (Appendix 3, 21) which again has an obverse with elaborate cuirass with a row of dots on stalks. It need not be doubted that it is by the same hand as the COMOB coins. The reverse again reads AVGG TESOB. These two coins show unequivocally that distinctive obverse dies were being supplied from the same workshop to the mint or mints signing TESOB and COMOB, and that the obverses were carefully differentiated as regards the ornament on the emperor's shield. They suggest that the second COMOB series, at least (that is, the one without cross, but with dots on stalks), is Illyrican. Is it, however, Thessalonican? And if it is, why does it not employ the naval Victory on the emperor's shield ?

The question next arises whether in terms of stylistic development these COMOB coins can plausibly precede, or be inserted into, the TESOB series. The coins both of Theodosius and Honorius with TESOB and AVGGG (so far as one can make out the fine detail on the shield) have another, and slightly different rare ornament, namely a Victory crowning a diminutive (?) kneeling figure. The specimens which indubitably have a Victory and prow are signed AVGG, and the best of them is from an extremely elegant obverse die. But the Victory is soon replaced, in the TESOB series, by the standard ornament of a rider (who is spearing a fallen figure, and should therefore presumably be seen as Virtus Augusti), and the great majority of the Thessalonican solidi—as also, apparently, all those from Constantinople—are unremarkable. The use of one or another of the three versions of a Victory on the shield forms a brief and thoroughly unusual episode.

Rather than regard the alternation of AVGGG and AVGG as anachronistic (as Hahn was driven to do) should one not try to accept, at least as a working hypothesis, that the formulae were used correctly? If they were, then the known COMOB coins (with Victory on globus ornament) were struck after the TESOB mint signature had been introduced in association with a different Victory ornament. The straightforward interpretation appears to be that the COMOB coins are not Thessalonican, but are a parallel issue, from ca. 408 (but the series may begin a little earlier, as will appear in a moment) from a peripatetic mint, presumably in Illyricum. The contrast between the Victory on globus and the Victory with captive/naval Victory is in that case, in Lafaurie's words, "une distinction qui individualise les produits des deux ateliers concurrement avec la marque de revers."

A coin from a recut reverse die clinches the argument. The mint signature has been altered on the die, TESOB being cut over CONOB or COMOB. Traces of the middle letter, which was probably an M, remain between the E and the S.60 As the legend ends AVGGG, the coin offers indirect evidence that the COMOB mint was in existence while that ending was still in use. The associated obverse, of Honorius, has Victory and small (?)kneeling figure very clearly visible on the shield61—certainly not a prow. The general style of the imperial bust (ornament of the cuirass, etc.) conforms to the later coins of Honorius with AVGG and TESOB.62

The shield ornament is a crux: if TESOB/AVGGG and AVGG are the only issues on which we find the Victory and kneeling captive, then there are real difficulties in arranging the Illyrican COMOB and the TESOB coins into a single sequence. But when an issue is known only from two or three specimens, an element of uncertainty must remain.

With that reservation, the pattern seems to be that the COMOB die cutting workshop occasionally or briefly helped out with dies signed TESOB, and with appropriate obverse dies as well:

TESOB COMOB
AVGGG Victory with kneeling captive AVGGG ? (recut rev. die used by TESOB, with a regular TESOB obv.)
AVGG Victory with kneeling captive (scarce) Victory and prow (naval Victory) AVGG Victory on globus (obv. dies occasionally supplied to TESOB).
AVGG Rider

Nothing is recorded in the history of the city of Thessalonica itself which would explain the brief appearance of a very unusual gold coinage, some of it struck from particularly elegant dies. The historical context of the issues is therefore a matter for conjecture; but some comments of a general nature are possible.

The strategic importance of Thessalonica to the eastern part of the Empire was enhanced whenever East and West were in conflict. The Via Egnatia, along which it was possible to travel from Rome to Constantinople in 21 days, had allowed speedy movement to Alaric, Stilicho, and Gainas. Theodosius would therefore have had no difficulty in perceiving that the security of Thessalonica was a necessary guarantee of the security of the capital. With the break between the two parts of the Empire in early 407, Stilicho claimed eastern Illyricum for the West and appointed Alaric magister militum per Illyricum. Anthemius, the pretorian prefect of the East, responded by ordering the fortification of the Illyrian towns, in a law of April 407. At the same time Stilicho decreed a naval blockade of the eastern Empire. The death of Alaric and Stilicho's delay turned the course of events; but it seems plausible that the Victory coins, dating from some time after May 408, commemorate the completion of public works.

End Notes

49
Popoviċ, pp. 111-26.
50
Kent (above, n. 3), p. 90.
51
J. Lafaurie, "Solidus d'Honorius, frapp� en Illyricum," BSFN 16-20 (1961-65), pp. 242-43, and "Cruces in vestibus (r�sum�)," BSFN 26-30 (1971-75), pp. 336-40.
52
Hahn (above, n. 13), p. 103.
53
Recent additions in the sale catalogue F. Sternberg 11, Nov. 1981, 985-86.
54
See n. 51, above.
55
Compare Florange and Ciani, 28-29 Apr. 1925, 584 and Dorotheum 285, 3-6 May 1966, 2. The winged figure holds a wreath downwards behind her.
56
Best seen on a coin in Birmingham, 4.32 g, and also on Hess 36, 1968, 593. See also n. 61.
57
Lafaurie (above, n. 51), "Cruces...," p. 336.
58
Mr. Sternberg has kindly written to me about the coins in his auction (above, n. 53), and in reply to my querying the accepted attribution, writes "I personally believe that these coins are from Thessalonica, or from the area which is today Bulgaria. If I am told right, these coins were part of a small hoard, found in western Turkey, but I have no secure confirmation." I am grateful to Mr. Sternberg for this information.
59
The reliability of the formulae AVGG and AVGGG is discussed at length (with reference particularly to western issues) in J. Lafaurie, "Le trésor de Ch�cy (Loiret)" in J. Gricourt et al., Trésors mon�laires et plaques-boucles de la Gaule romaine, Gallia, suppl. 12, (1958), at p. 279.
60
Glendining, 25 Nov. 1953, 206 (mistakenly cited in Lafaurie as 25 Dec).

APPENDIX 2 The Coins: Types and Summary

Theodosius II, 408-60

There are four series of solidi with the mint signature TESOB. They have been published and their dates established by Hahn.63

A. CONCORDIA AVGGG or AVGG, 408-22, Hahn 54a, b, c; 55a, b, c. These were issues of Theodosius and Honorius as joint augusti based on Constantinopolitan prototypes (Plate 15, A and B). Twenty-six of the specimens (1-28) were checked, and were found to be from 22 obverse and 23 reverse dies. The range of styles suggests piecemeal die engraving, with both emperors represented at each stage. More than half the dies (for both emperors) are marked TES.OB·

B. VOT XXX MVLT XXXX, 430 or later. Eight (29-36) specimens all from different dies.

C. VIRTVS EXERC ROM, 440 or later. One specimen (37).

D. GLOR ORVIS TERRAR, 443-50. Among 85 specimens (38-122), there were 5 pairs of die duplicates, plus 3 additional obverse pairs and 4 additional reverse pairs: thus 77 different obverse dies and 76 reverse dies.

For the reign as a whole, we may estimate that around 670 pairs of dies were used. If this total is divided pro rata according to the numbers of specimens in the sample, the four series might be expected to have involved roughly 125, 40, 5, and 470 pairs of dies respectively.

End Notes
61
At first sight running away from Victory, but on closer inspection probably kneeling (on one knee) facing Victory, as on the coins mentioned in n. 56.
62
Hahn (above, n. 13), p. 108.
63
Hahn (above, n. 13), p. 121.

Marcian, 450-57

Two series were minted at Thessalonica.

A. GLOR ORVIS TERRAR. A stylistically compact group, extremely similar to the neater coins of Theodosius of the same type. The mint signature is still TESOB, but the E is not easily distinguished. The 15 specimens (131-45) were from 13 obverse and 13 reverse dies.

B. VICTORIA AVCCC or AVCC. The standing Victory with cross type makes its first appearance at Thessalonica. The cuirass is now in a different style, being divided by vertical lines into three panels. The mint signature has changed to THSOB. The ending CCC has become meaning less, and the significance of the dot (supplementary issue ? second officina ?) is not clear. There are five specimens (146-50), all from different dies.

In the sample the number of duplicates is too small for the estimated total of dies to be firm. The calculation yields a figure of about 60. This is somewhat lower than the result obtained if we assume a similar survival rate to that observed for Theodosius, when the total would be of the order of 100.

Marcian struck silver miliarenses at Thessalonica, also with a military reverse: GLOR ORVIS TERRAR, emperor standing facing, head left, holding spear and shield (see, for example, Münz. u. Med., 19-20 June 1964, 512).

Leo I, 457-74

There are two (or perhaps three) series of solidi, all with reverse VICTORIA AVCCC and the mint signature THSOB.

A. Victory standing left with long cross, which matches the standard Constantinopolitan type: there are two varieties, the first with one star in the left field, the other, more plentiful, with two stars. The first variety has a plain cuirass, the second normally shows the cuirass divided into three panels. The quality of the reverse, in particular the careful treatment of the cross held by the Victory, suggests that the two varieties were not concurrent. AVCCC· occurs as an alternative in both varieties, but is uncommon. One coin, in a style which does not match the rest of the series closely, is of presumed Tunisian provenance. Among 35 specimens (151-85), 31 obverse and 32 reverse dies were found.

image

Fig. 4

B. Consular bust of Leo (Leo was Consul in 458, 462, 466, and 471), reverse with throned figure in consular vestments. This is a specifically Thessalonican type. There are again two varieties, one with one star, the other, less plentiful, with two stars and a broken obverse legend. The folds of drapery over the knees of the seated figure are variously depicted (Fig. 4). The lack of overlap in the styles of the reverses suggests that the two varieties were not concurrent. Hoard evidence would be required to show that they belonged, for example, to Leo's first and second consulships.

The more plentiful variety includes a few coins which seem to be by a different hand: the diadem can be straight instead of curved, and the lettering is larger and less regular. The variability of reverse styles among these coins, e.g. in their treatment of the loros and the drapery over the knees, make them appear to be in some sense derivative. They have accordingly been listed at the end of the sequence, along with one or two even rougher and more problematic coins—a specimen from dies possibly of local workmanship has two stars but an unbroken legend. Could they belong to a different consulship? Again, only hoard evidence could tell. Among 25 specimens (187-211), there are 15 obverse and 20 reverse dies. The Victory and the consular types have, prima facie, had distinctly different survival rates. The former may be estimated to be from ca. 125 obverse and 270 reverse dies, the latter from only ca. 23 obverse and 54 reverse dies. Thus we have one Victory coin from about eight reverse dies originally used, but one consular coin from between two and three. Thessalonican coins of either series are sufficiently scarce that the selective melting down of Victory coins in modern times is rather unlikely. In both issues the reverse: obverse die ratio, equal to or in excess of 2:1, marks a change from the 1:1 ratio of preceding issues; and, among the consular coins at least, the figures are statistically significant.

Leo II, 474, with Zeno; Zeno, 474-75

Hoard evidence would be required to show whether any of the coins given to Leo I belong in fact to Leo II. It is unlikely that any solidi were minted at Thessalonica in the six weeks or so of Zeno's first sole reign (see Zeno, below). There is however, a silver coinage for Zeno which retains the mint signature THSOB (Paris, Sabatier 9).

Basiliscus, 475-

Victory with cross remains the standard reverse type, with meaningless AVCCC. Ten specimens (212-21) from the 20-month reign are from eight obverse and nine reverse dies. This suggests that there may originally have been roughly 20 obverse dies, but possibly a larger number of reverses (see the remarks on die ratio under Leo I).

Zeno, 476-91

The mint mark is now changed to CONOB, and the coins are distinguishable from those struck at the capital by two stars in the reverse field and by the absence of an officina numeral. The reverse type is the standard one. As Basiliscus had used the mark THSOB, it is likely that Zeno's Thessalonican coins are all from his second reign. There are two varieties, the cuirass being divided into either three or two vertical panels.

The two varieties are evidently chronologically parallel. They are linked by a reverse die. The more careful dies and those on which the cross has an X-shaped bar presumably were first. The coarse outer border of 227 is thus kept close in date to the reign of Basiliscus, during which its use was normal.

Various stylistic differences between the Thessalonican and Constantinopolitan coins are sufficient to show that the dies were not cut in the capital and supplied to Thessalonica: the border panels of the cuirass and of the cloak are dotted at Constantinople but have wavy lines here, and the feathering of the Victory's wing occurs in a number of styles at Constantinople, but is uniform at Thessalonica. Lallemand proposed to attribute to Thessalonica a series of solidi of Zeno with the reading AVGGT followed by an officina numeral (CENB 1 [1964], pp. 49-51). This attribution is questionable if only because the series is from a mint with (probably) ten officinae.

The 15 coins examined (222-36) are from 14 or 15 obverse and 14 reverse dies, so no accurate die estimate is possible.

Anastasius, 491-518

There are two issues of the Victory type, reading PERP AVC (MIB 14) and PPAVC (MIB 15) respectively. There is no obvious continuity between them. On MIB 14 the bar of the cross is X-shaped, and there is a horizontal panel with a row of dots at the top of the cuirass. The feathering of the Victory's wing is consistent in style. MIB 15, on which the bar of the cross is straight, includes a group of very elegant obverse dies, with long curled ties to the diadem (247-49), which are presumably the early or best work of a new die engraver.

In all, 31 coins (237-67) are from 30 pairs of dies, implying a large original total, ca. 350; if the next coin to be added to the list made up a second pair, the estimate would fall to ca. 180.

Justin I, 518-27

Thessalonican solidi are of an obsolete reverse type and are further distinguished by the usual two stars. There are two varieties, MIB 6 with diadem ties and MIB 7 without diadem ties.

The obverses are all by the same hand, with the possible exception of 273. The face is broad and heavy, and the shield is broadly triangular, coming to a sharp point. The initial DN of the inscription tends to be out of alignment. An unusual treatment of the cuirass is seen at its most developed on 270, which has an ornamental vertical panel placed centrally. In simplified form it is repeated on 274 and 275.

The reverses are closely interrelated in style, the Victory's wing being a useful diagnostic detail. An exception is reverse 271 which appears to be from an Anastasian reverse die. One coin, 275, is of very low weight, but is authenticated by its reverse style, although the obverse is irregular.

The possible significance of the diadem ties has been discussed by Hahn. There is however no clear distinction of style and, on the relatively elaborate obverse of 273, for example, one must wonder whether the ties may not have been omitted simply by inadvertence.

One obverse die of Justin has been recut and used by Justinian (279). There are nine (270-78) or, with the altered die, ten specimens, all from different dies.

Justinian I 527-65

An obverse die of Justin I (279) was altered by the addition of the letters N I A. The reverse with which it is used may well be an old die too.

Four varieties of solidi were minted at Thessalonica during Justinian's reign, as well as a few tremisses. The distinguishing features of each type are briefly described below, and the dates proposed by Hahn64 for each of them are given.

Solidi (280-363)

MIB 20, 527-37. The reverse type is the old-fashioned Victory left, with two stars in the field. The obverse has no diadem ties.

MIB 21, 537-42. The same, with diadem ties.

MIB 22, ?542-62. The obverse design is brought up to date and now matches the facing portrait used at the Constantinople mint. On the reverse Victory still holds a plain cross instead of a cross with chrismon. The coins are otherwise distinguished only by the absence of an officina numeral and by the lack of a cross on the globus on the reverse.

MIB 23, ca. 562. Transitional issue, as MIB 22, but has cross with chrismon.

MIB, N23, 562-65. The type now corresponds exactly with that used at Constantinople, but there in no officina numeral. The first A in the reverse legend is set vertically on the flan.

Tremisses (364-66)

MIB 24, N24. Indistinguishable except by the correspondence of style of the obverse dies with that on copper coins of Thessalonica. N24 associates an obverse of Justinian with a reverse, without star, probably from the time of Justin II.

Analysis

MIB 21 and 20. In Hahn's scheme MIB 20 is dated to 527-37, and MIB 21 to 537-42. Since MIB 20 is relatively the far more plentiful, this would imply that the minting of gold either ceased very soon after 537, or dwindled to much lower levels. But one may question the sequence, on the evidence both of style and of the Hajducka Vodenica hoard.

The obverses would seem all to be the work of one hand. They are mostly untidy, often with staring eyes, and the cuirass is usually very simply drawn, as in Fig. 5.

image

Fig. 5

The seven versions in the top row seem to be used almost indifferently, and variant 2 (298-308) probably belongs with them. The more elaborate style of variant 1 (280-97) occurs on a die with small, neat lettering, and a compact and well-modelled portrait which one would be inclined to place at the beginning of the sequence: it is of MIB 21. The alternative explanation would be that it reflected a sudden effort to improve the quality of the coinage, in or shortly after 537 (in connection with the reform of 538?).

An interesting stylistic detail is the curl of hair at the left of the emperor's neck. On the better dies it is roughly circular but on the remainder it is horn-shaped. Both styles are known from earlier reigns, and it is not impossible that there was an alternation within the sequence of MIB 20; but the natural arrangement would be to keep the two variants separate, and to see 298 as introducing the horn-shaped curl.

The Hajdučka Vodenica hoard, dated not long after the introduction of the frontal portrait at Constantinople (MIB 7, 542 onward on Hahn's chronology) includes only two specimens of MIB 21, against 10 of MIB 20. One might have expected the Thessalonican issues of 537-42 to be relatively more plentiful in the hoard than that, even if the output in those years was on a restricted scale. Most of the 10 specimens of MIB 20 are dies that stand very close to each other in style, with a horn-shaped lock of hair at the left side of the emperor's neck, and they include four coins from one obverse die, of which two are from a shared reverse die. It would, again, be natural to see those as being among the latest coins in the hoard. But the argument is not conclusive, since one cannot know the particular circumstances in which a single hoard was put together.

There is at present no way in which one can judge the course of mint output during the years 527-42. Could the minting of gold have been on such a small scale before ca. 535 (when administrative changes affecting Thessalonica were made) that three specimens of MIB 21 are all that are known from the quinquennium 527-32? Or could the omission of the diadem ties have been merely the outcome of carelessness or indifference, with no quinquennial significance? The discovery of a hoard concealed ca. 530-35 is needed in order to resolve the uncertainty. As an interim arrangement, however, MIB 21 and 20 have been listed in that order as a single stylistic sequence.

The degree of die linkage among the coins may be exaggerated by the Hajdučka Vodenica hoard. Thirty specimens were found to be from 27 obverse and 29 reverse dies. The best estimate one can offer is 400 or more reverse dies.

MIB 22. The 46 specimens of MIB 22 can be arranged into an approximate stylistic order by reference to the number of curls on the emperor's forehead, the depth of the cuirass, and the treatment of the arm holding the globus cruciger. The coins seem to fall into two sequences, reminiscent of Constantinopolitan MIB 6 and 7 respectively. The early specimens of each sequence are from more careful and elaborate dies, with more curls and a deeper cuirass, but there is no very obvious continuity into the obverses for the next issue, MIB 23 (although see 355). If MIB 22 covered a twenty-year period, there was plenty of time for two separate issues, and perhaps the beginning of a third. The following scheme has been adopted.

Style A:

  • 8 curls and deep cuirass;
  • 7 curls and similar cuirass;
  • 7 curls, line across throat;
  • 6 curls, line across throat;
  • 5 curls, line across throat; and
  • 5 curls.

Style B:

  • 7 curls and deep cuirass;
  • 6 curls sometimes with deep cuirass; and
  • Rougher style, 5 curls.

Style C:

  • 5 curls.

The letters at the end of the reverse legend, CCC, tend to be larger on coins with obverse style B. The arrangement is, however, still very subjective, and the two sequences might have to be amalgamated into one, depending on future hoard evidence from the 540s or the discovery of reverse die links. It may turn out that there are metrological differences, the earlier coins being more exactly controlled in weight, and the later ones showing a wider spread. And it would not be surprising if the survival rate were higher among the late specimens.

The 46 coins are from 41 obverse and 44 reverse dies, suggesting original totals of about 200 obverses and 500 reverses.

MIB23 and N23. On the obverse dies for these coins the emperor wears his hair in a straight fringe instead of five or more curls on the forehead. The two reverse dies for MIB 23 appear to be by the same hand as MIB 22 (314-16) and might even be altered dies cut originally for the earlier coins. The dies for N23 are in direct continuation, again with a fringe, but they introduce something new: five specimens from the same obverse die. More than one interpretation is possible. Either a collapse in output in the final years of Justinian's reign is marked by a higher survival rate, or a failure in the supply of dies led to this one obverse die being used far more than was normal.

Four obverses and nine reverses are recorded among nine specimens. Possibly between 20 and 40 reverse dies were originally used.

MIB 24 and N24. At present only two tremisses have been assigned to the Thessalonican mint. A complete die study of the tremisses might isolate products of Thessalonica now concealed in the "Constantinopolitan" corpus.

End Notes
64
Hahn (above, n. 4), pp. 179-80.

Justin II, 565-78

A solidus type without the traditional star or stars (MIB 16, 367-93) is firmly attributed to Thessalonica on the evidence of the hoard of 1948 and of its stylistic connection with the last coins of Justinian. There is no officina numeral, and the AS in the reverse legend are distinctively aligned.

Similar coins read AVCCCS, AVCCCJ, or AVCCCT (MIB 17, 394-418). Their chronological relationship to the variety without officina numeral is difficult to determine, as the dies are apparently cut by other hands. The reverses, in particular, have normally aligned As and more prominent serifs on the cross and the lettering. The obverses for AVCCCS and AVCCCJ are of Constantinopolitan quality, the fringe of hair is continued vertically past the temples, and the garment above the cuirass at the throat preclude their being part of the same sequence as MIB 16. The obverses associated with AVCCCT are either later or, as Hahn convincingly suggests, locally cut replacements. The heavy representation of these otherwise excessively scarce varieties in the Thessaloniki hoard suggests that they are contemporary with, or later than, the coarser dies of the unnumbered sequence.

Hahn has attributed two varieties of tremisses found in the hoard to the local mint—one on which the star is omitted (MIB N17, 419-22) and another (MIB NN17, 423-33) distinguishable from Constantinopolitan issues only by style (the bust is broader and the head is larger). The heavy die linking among the latter variety is a supporting argument.

MIB 16. Neat dies of thoroughly Constantinopolitan style are followed by others of much coarser execution. On these the face is broad and, on the reverse, the knee is sharply bent. The distribution of the known specimens throws interesting light on the Thessaloniki hoard.

Neat Coarse Total
Thessaloniki hoard 2 15 17
Other sources 7 3 10
9 18 27

Die linkage in the hoard is such that the hoard sample might be expected to yield an underestimate of the number of dies used in the later part of the series. In all, 27 coins are from 13 obverses, with an implied original total of 17 dies. It would be more realistic to increase the estimate to 20 "equivalent dies." The reverses include only three pairs and yield an estimate of 93 dies. The ratio of reverse to obverse dies thus appears to have been surprisingly high, possibly as much as 4:1 or even 5:1.

MIB 17. Twenty-five coins, exclusively from the Thessaloniki hoard, are from eight obverse dies, but from 18 reverse dies, suggesting original totals of ten or a dozen obverses and 45 reverses, a die ratio of about 4:1, not dissimilar to the unmarked series. The hoard may in principle be yielding an underestimate but, as the coins are otherwise unreported and of a period when, because of the invasions, one would expect a high loss rate and therefore a reasonably high survival rate, the tally could in fact be a fair sample.

MIB N 17 and NN 17. Fifteen coins, all from the hoard, are from 11 obverse and 12 reverse dies, implying a conservative estimate (since the hoard may not be a fully random sample) of 30 to 35 pairs of dies, evidently in a 1:1 ratio, quite unlike the ratio employed for solidi. Hoards summarily published in the past need to be looked at again to see whether they included any tremisses which would now be attributed to Thessalonica.

Tiberius II, 578-82

The Thessaloniki hoard contained four solidi (434, 438-40) of Tiberius II attributable to the local mint by a comparison with the style of the bust found on Thessalonican copper coins. Their distinctive feature is the boldly dotted style of the reverse cross. Their reverse legend ends with S (stigma), which has been interpreted as year 6 (579/80).65 Two other specimens (435-36) are known. The six are, apparently, from six different reverse dies, but the implied total of obverse dies is only about a half dozen.

There is also a coin on which the legend ends with H, year 8 (581/2, 441), and the same obverse die is used with a Constantinopolitan reverse (442, MIB 11). As this variety is not found in the Thessaloniki hoard, its deposit may have been no later than ca. 581. The use of a Constantinopolitan reverse die, belonging to the consular issue (MIB 2), makes one wonder whether the mint may not also have obtained the obverse that made the pair, but in default of hoard evidence (and specifically in view of their absence from the hoard of 1948) the attribution of consular coins to Thessalonica still seems highly speculative.

A tremissis in the hoard (MIB N11, 443) can, like the solidi of year 6, be attributed to Thessalonica by comparison with the style of the copper coinage; and another variant (MIB NN11, 444) is probably also from Thessalonica. Whether any of the tremisses in the Barga la hoard are local remains to be seen.

Maurice Tiberius, 582-602

The minting of solidi seems to have been reduced to special issues for the tenth and twentieth anniversaries of Maurice's accession (591/2, 601/2). They are identifiable only by similarities of style with Thessalonican copper coins. The half-dozen dies that could be checked (445-50) were all different, but the rarity of the varieties makes it unlikely that they were minted in large quantities.

There are tremisses (451-57) of three different varieties (MIB 22, 23, N23) which have been attributed to Thessalonica. The very high incidence of reverse die linkage in an admittedly small sample suggests a reverse to obverse ratio of 1:1 as under Justin II.

Phocas, 602-10

No Thessalonican gold issues have been satisfactorily identified, although they may exist; those in DOC are probably Ravennate or in any case western. One specimen (458) which is a candidate for attribution to Thessalonica, was acquired there,66 and has peculiarities of style which cannot readily be matched at other mints. The carefully formed letter G at the end of the obverse legend is the most obvious feature. One may draw attention also to the brooch fastening the cloak, the long and widely spaced pendilia, the reading FOC∧∾, and the cross of the crown (not standing on an arch). On the reverse the cross on the globus is tall. The reverse legend ends with an uncertain numeral which might be I, C, or even ϵ. The general style of the coin would suggest a date quite early in the reign.

Heraclius, 610-41

Thessalonican solidi are distinguished by their style, which can be matched on copper coins of the mint, by details of the design—in particular the break in the reverse legend I AA (cf. IA A at Constantinople)—and by the use of indictional dates.67

The first issue (Heraclius alone) is dated Iϵ (611/2) and possibly Iᒥ (perhaps indiction ᒥ = 614/5, unless the letter in question is merely an ϵ lacking its bar). Coins of the second issue (Heraclius and Heraclius Constantine) have been read as dated ᒥ (614/5) or ϵ (616/7).68 The letter ᒥ is not, however, particularly clear and could well be an unbarred ϵ. This interpretation finds some support in a coin with a particularly small bust of Heraclius Constantine, distinctly smaller than on the coins that have been read as ᒥ. As it was normal for the portrait of the youthful emperor to be increased in size with the passage of time, a decrease would have flouted convention and would have been conspicuous. The coin in question is from neat dies and has been listed first in Appendix 3 on the assumption that the whole issue belongs to the fifth regnal year.

Among ten specimens (459-68) there is one pair of die duplicates. It is not possible to estimate the original total of dies usefully.

End Notes
65
W. Hahn, "More about the Minor Byzantine Gold Mints from Tiberius II to Heraclius," NCirc 87 (1979), pp. 552-55.
66
This coin passed through the hands of Simon Bendall of A. H. Baldwin and Sons, Ltd., and I am much indebted to him for information about it, and for the photograph reproduced in the plates.
67
W. Hahn, "Some Unusual Gold Coins of Heraclius and Their Mint-Attribution," NCirc 85 (1977), pp. 536-39.

APPENDIX 3 The Coins: An Inventory

Theodosius II

A. CONCORDIA AVGGG/AVGG, Theodosius and Honorius as joint augusti; see Plate 15, A and B, for Constantinopolitan prototypes

Hahn 54a, Theodosius: AVGGG

  • 1. Victory and kneeling figure on shield, rev. with elegant prow. Hess-Leu 36, 17-18 Apr. 1968, 593
  • 2. Similar Victory, blunt prow. Dorotheum 387, 29-30 Nov. 1978, 274

Hahn 54b and 54c, Theodosius: AVGG; 54b, TESOB; 54c, TE·SOB·

  • 3. Obv. with dots on stalks above cuirass (i.e. a die cut in the COMOB workshop), naval Victory on shield; rev. in elegant style. Vienna 38, 528, 4.26 g
  • 4. Shield ornament apparently Victory and kneeling figure, slight decline in elaboration and quality, 54a, Birmingham, 4.32 g
  • 5. Elegant naval Victory on shield; rev. with elegant prow. 54b, Florange and Ciani, 28-29 Apr. 1925, 584
  • 6. Dorotheum 285, 3-6 May 1966, 2
  • 7. 54c, Paris
  • 8. Obv. of 7. 54c, Numismatic Fine Arts 2, 26 Mar. 1976, 489, 4.28 g

With horseman on obv. shield

  • 9. Rev. with crude prow, 54c, Peus 269, 26-29 Nov. 1968, 325 = Hirsch 45, 9-10 Dec. 1965, 1907, 4.40 g
  • 10. 54c, Hirsch 138, 28-30 Sept. 1983, 230 = Hirsch 111, 12-15 June 1978, 2274, 4.41 g
  • 11. Rev. with larger, square prow. 54c, Birmingham, 4.43 g
  • 12. 54c, Frey, 15-16 Apr. 1955, 602
  • 13. 54b, Kress 93, 17 Nov. 1952 (ex Balvin coll.), 1553, 4.28 g
  • 14. 54b, Milan
  • 15. 54b, Hirsch 138, 28-30 Sept. 1983, 231, 4.25 g
  • 16. 54c, Peus 296, 2 Nov. 1978, 702, 4.36 g
  • 17. Obv. of 16. 54c, Bank Leu 13, 29-30 Apr. 1975, 534 = Münz. u. Med. 43, 12-13 Nov. 1970, 519, 4.41 g
  • 18. 54c, Fagerlie 283

Hahn 55a, Honorius: AVGGG

  • 19. Victory and kneeling figure on shield; altered rev. die. Glendining, 25 Nov. 1953, 206
  • 20. Budapest [not seen]

Hahn 55b and 55c, Honorius: AVGG; 55b, TESOB; 55c, TES·OB·

  • 21. 55b, naval Victory, obv. close in style to 3. Rev. of 5. Hess-Leu 41, 24-25 Apr. 1969, 712
  • 22. 55c, obv. and rev. close in style to 7-8. Möller 20, 20-21 May 1977, 509 = Winter 31, 22-23 Nov. 1976, 241
  • 23. 55b, Berlin (Klein Tromp hd.), 4.35 g
  • 24. 55b, Stanley Gibbons FPL 2 [1974], 9
  • 25. 55c, Budapest [not seen]
  • 26. 55c, Hirsch 97, 22-25 Mar. 1976, 698
  • 27. 55c, Fagerlie 4
  • 28. 55c, Kress 137, 20 Nov. 1966, 841a = Kress 116, 28 Oct. 1960, 726

Coins 20 and 25 not checked for die duplication.

Obv. links, 7 = 8, 16 = 17.

Rev. links, 5 = 21.

B. VOT XXX MVLT XXXX

  • 29. Berlin (Klein Tromp hd.), 4.43 g
  • 30. London
  • 31. In commerce (Bernardi)
  • 32. Kress 116, 28 Oct. 1960, 1106
  • 33. Kricheldorf 13, 23 Sept. 1963, 356
  • 34. Ratto 179 = Vienna 44, 140, 4.3 g
  • 35. Frankfurter Münzhandlung 90, 2 Mar. 1943, 18
  • 36. Hess-Leu, 24 Mar. 1959, 404, 4.32 g

Die duplicates: none

C. VIRTVS EXERC ROM

  • 37. Obv. die very similar to best of following type, supporting late dating of the GLOR ORVIS TERRAR type. Ars Classica 17, 3 Oct. 1934, 2014, 4.42 g

D. GLOR ORVIS TERRAR

Neat style

  • 38. Berlin (Klein Tromp hd.), 4.26 g
  • 39. Naville 17, 3 Oct. 1934 (Evans), 2011, 4.41 g
  • 40. Kricheldorf 14, 8 July 1964, 391
  • 41. Rev. of 40. Münz. u. Med. 43, 12-13 Nov. 1970, 518, 4.30 g
  • 42. Kunst und Münzen 16, 19-21 Oct. 1976, 642
  • 43. Dorotheum, 8-9 June 1956, 2399
  • 44. Hirsch, 17 Nov. 1955, 625
  • 45. Peus 269, 26-29 Nov. 1968, 324, 4.27 g
  • 46. Numismatic Fine Arts 2, 25-26 Mar. 1976, 488 = Bank Leu 2, 25 Apr. 1972, 466, 4.34 g
  • 47. Kress 133, 6 Sept. 1965, 1151
  • 48. Klenau 88, 2 Mar. 1974, 2018
  • 49. Dies of 48. Kress 174, 8-9 Mar. 1979, 1176 = Peus 294, 15-17 Mar. 1978, 824, 4.15 g
  • 50. Kress 120, 30 Nov. 1961, 945
  • 51. Possibly rev. of 50. Kricheldorf 33, Oct. 1978, 601
  • 52. Kress 109, 24-25 Oct. 1958, 1458a
  • 53. Kress 112, 22 June 1959, 960
  • 54. Birmingham, 4.37 g
  • 55. Kress 125, 17 Apr. 1963, 828
  • 56. Obv. of 55. Kress 127, 23 Oct. 1963, 811
  • 57. Kress 149, 10 Nov. 1969, 956, 4.45 g
  • 58. Berlin (Klein Tromp hd.), 4.37 g
  • 59. Kress 131, 16 Nov. 1964, 942
  • 60. Bernardi FPL, Oct. 1967, 315
  • 61. Oxford, 4.29 g
  • 62. Oxford, 4.38 g
  • 63. Ratto 150 = Malter 21, Summer 1968, 49
  • 64. Rev. of 63. Basel Münzhandlung 6, 18 Mar. 1936, 2112, 4.40 g
  • 65. Naville 3, 16 June 1922, 271
  • 66. Peus 268, 24-26 Apr. 1968, 246, 4.33 g
  • 67. Rev. of 66. Kress 120, 30 Nov. 1961, 946
  • 68. Münz. u. Med. 52, 19-20 June 1975, 4.33 g
  • 69. Kunst und Münzen AG 15, 8-10 Dec. 1975, 877

Standing figure is broader

  • 70. Peus 263, 14-16 May 1962, 18
  • 71. Paris
  • 72. In trade
  • 73. Dies of 72. Münz. u. Med. FPL 168, May 1957, 42
  • 74. Rauch 26, 23-25 June 1980, 412
  • 75. Hess-Leu 41, 24-25 Apr. 1969, 727, 4.46 g
  • 76. Kress 131, 16 Nov. 1964, 939
  • 77. Dorotheum 287, 30 Aug. 1966, 3
  • 78. Kress 135, 15 Mar. 1966, 531 = Kress 130, 30 June 1964, 965
  • 79. Paris
  • 80. Sternberg, 29-30 Nov. 1974, 670, 4.47 g
  • 81. Same coin as 80? Kricheldorf 30, 16 Apr. 1976, 376
  • 82. Kricheldorf 13, 23 Sept. 1963, 352
  • 83. Möller 15, 19-20 Sept. 1975, 4.35 g
  • 84. Kress 123, 4 Oct. 1962, 689a
  • 85. Kress 116, 28 Oct. 1960, 1102a
  • 86. Sotheby, 6 June 1973, 50, 4.42 g
  • 87. Glendining, 27 Nov. 1958, 57
  • 88. Bonhams 2, 23-24 Sept. 1980, 376 = Möller 17, 23-24 Apr. 1976, 495, 4.40 or 4.42 g
  • 89. Hess-Leu 49, 27-28 Apr. 1971, 490
  • 90. Dies of 89. Kress 108, 10-11 June 1958, 595
  • 91. Elsen FPL 22, May 1980, 189
  • 92. Möller 14, 31 Jan.-1 Feb. 1975, 472, 4.28 g
  • 93. Myers 6, 6 Dec. 1973, 415, 4.35 g
  • 94. Kress 144, 22 July 1968, 848
  • 95. Carthage, 4.30 g

Distinctive X in circle at midriff

  • 96. Kress 131, 16 Nov. 1964, 941
  • 97. Dies of 96. Superior Stamp and Coin, 19-23 Aug. 1975, 3087
  • 98. Obv. of 96. Dorotheum 203, 3-6 July 1957, 60
  • 99. Obv. of 96, rev. of 98. Hirsch 22, 2-5 Dec. 1959, 677
  • 100. Numismatic Fine Arts 12, 23-24 Mar. 1983, 496 = Bonhams 1, 21-22 May 1980, 582 = Sternberg, 28-29 Nov. 1975, 530, 4.37 or 4.39 g
  • 101. Auctiones AG Basel 4, 26-27 Nov. 1974, 577, 4.38 g
  • 102. Paris
  • 103. Dorotheum 321, 29 Sept.-2 Oct. 1970, 3
  • 104. Milan
  • 105. Hirsch 31, 28-30 May 1962, 657
  • 106. Kress 139, 19 June 1967, 2058 = Kress 123, 4 Oct. 1962, 696
  • 107. London
  • 108. Kress 152, 5 July 1971, 793
  • 109. Kricheldorf 35, 15 Oct. 1981, 577
  • 110. Basel Münzhandlung 10, 15 Mar. 1938, 805
  • 111. Paris
  • 112. Rev. of 111. Ratto 151

Standing figure with long legs

  • 113. London
  • 114. Tellmann FPL 116, July 1967, 338 = Tellmann FPL 89, Mar. 1965, 147 = Tellmann FPL 79, May 1964, 3
  • 115. Peus 284, 9-10 Dec. 1974, 1216
  • 116. Rauch 12, 18-19 May 1973, 595
  • 117. Kricheldorf, 15 Oct. 1955, 1

Cruder style

  • 118. Hirsch 68, 1-3 July 1970, 891
  • 119. Kress 131, 16 Nov. 1964, 940
  • 120. Münz. u. Med. 38, 6-7 Dec. 1968, 671
  • 121. Ex Balvin coll., 4.35 g
  • 122. Dorotheum 215, 17-19 Sept. 1958, 132

Contemporary counterfeit?

Imitation from the northern lands?

  • 124. Kress 109, 24-25 Oct. 1958, 443, 4.3 g

Addenda

  • 125. Kunst und Münzen AG, 2-4 June 1975, 695
  • 126. Cambridge, 4.30 g
  • 127. Hirsch 138, 28-30 Sept. 1983, 232
  • 128. Vienna 30.100, 4.42 g
  • 129. Vienna 30.091, 4.4 g
  • 130. M. Amandry et al., Catalogue des monnaies d'or (Musee de Saint-Omer, Arras, 1983), 155, 4.34 g

Die duplicates: 48 = 49, 72 = 73, 89 = 90, 96 = 97, 98 = 99 (80 and 81 may well be the same coin, or alternatively another die identity; for purpose of estimation they have been treated as one coin).

Obv. links: 55 = 56, 96 = 98.

Rev. links: 40 = 41, 63 = 64, 66 = 67, 111 = 112.

End Notes
68
Hahn (above, n. 68), p. 536.

Marcian

A. GLOR ORVIS TERRAR

  • 131. Unusual treatment of cuirass. Spink 22, 15-16 June 1982, 606
  • 132. Rev. of 131. Schulman 236, 1-4 May 1962, 165 = Birmingham, 4.42 g
  • 133. Hirsch 24, 10 May 1909 (Weber), 2959 = Tolstoi 1, 4.28 or 4.45 g
  • 134. Peus FPL 15, May 1970, 91
  • 135. Kress 139, 19 June 1967, 2061 = Kress 137, 21 Nov. 1966, 860
  • 136. Dies of 135. Lanz (Munich) 22, 10 May 1982, 956, 4.49 g
  • 137. The Hague
  • 138. Turin = Münz. u. Med. 12, 11-13 June 1953, 899
  • 139. Fagerlie 375, 4.39 g
  • 140. Numismatica 13, 9-11 Nov. 1976, 1938, 4.45g (Hirsch 111, 12-15 June 1978, 2275, is apparently the same coin)
  • 141. Obv. of 140. Copenhagen
  • 142. Milan
  • 143. New York City = Glendining, 7-8 March 1957, 475
  • 144. London
  • 145. Vienna 35.284, 4.35 g

B. VICTORIA AVCCC or AVCCC.

  • 146. New York City, 3.83 g, plain
  • 147. Birmingham, 4.43 g, dot
  • 148. Tolstoi 15, 4.55 g, dot
  • 149. Münz. u. Med. FPL 417, Nov.-Dec. 1979, 38, 4.24g, plain
  • 150. Malter 14, 8 June 1980, 183, 4.45 g

Die duplicates: 135 = 136.

Obv. links: 140 = 141.

Rev. links: 131 = 132.

Leo I

A. Standing Victory

1a. Plain cuirass, one star, AVCCC

  • 151. Private coll., London, 4.31 g
  • 152. Birmingham, 4.43 g
  • 153. Kress 112, 22 June 1959, 963
  • 154. In commerce, 1981, 4.28 g

lb. Plain cuirass, one star, AVCCC·

  • 155. Copenhagen = Fagerlie 524, 4.35 g
  • 156. Turin = Münz. u. Med. 11, 23-24 Jan. 1953, 210
  • 157. Stanley Gibbons FPL 2, [1974], 13
  • 158. Dies of 157. Bendall coll.
  • 159. Ratto 251
  • 160. Fagerlie 522, 4.44 g

2a. Plain cuirass, two stars AVCCC

161. New York City, 4.36 g

2b. Cuirass in three panels, two stars, AVCCC

  • 162. Tolstoi 14
  • 163. Kress 135, 15 Mar. 1966, 569 = Kress 123, 4 Oct. 1962, 701 = Kress 120, 30 Nov. 1961, 964
  • 164. Fagerlie 525, 4.46 g
  • 165. Münz. u. Med. 13, 17-19 June 1954, 808
  • 166. Rev. of 165. Milan
  • 167. Obv. of 166. Numismatic Fine Arts 12, 23-24 Mar. 1983, 508, 4.29 g = Kress 147, 5 May 1969, 818, 4.4 g
  • 168. Kress 91, 26 Nov. 1951, 270a = Balvin coll., 4.41 g
  • 169. Kress 128, 23 Nov. 1963, 872 = Kress 123, 4 Oct. 1962, 697
  • 170. Kress 130, 30 June 1964, 975
  • 171. F�eg coll.
  • 172. Kress 109, 24-25 Oct. 1958, 1464, 4.3 g
  • 173. Rauch 7, 4-5 June 1971, 333
  • 174. New York City, 4.47 g
  • 175. Obv. of 174. London
  • 176. Vienna 38.752, 4.47 g
  • 177. Four dots in horizontal panel of cuirass. Hess-Leu 28, 5-6 May 1965, 558, 4.28 g
  • 178. Obv. of 177, Peus 269, 26-29 Nov. 1968, 332, 4.30 g
  • 179. Fagerlie 529, 4.43 g
  • 180. Münz. u. Med. 43, 12-13 Nov. 1970, 529, 4.41 g
  • 181. Milan
  • 182. Fagerlie 530, 4.47 g

2c. Cuirass in three panels, two stars, AVCCC·

  • 183. Vinchon, 14 May 1982, 83, 4.48 g = Loudmer and Poulain, 15-16 June 1976, 296 = Ratto FPL 1, 1968, 19
  • 184. Rome, Capitol Mus.
  • 185. Obv. distinctive cuirass, shield, lettering, etc.; rev. short, curved wing, one star, lettering as on obv. Carthage, 4.45 g

2d. Contemporary counterfeit

  • 186. Rev. one star, dotted outline of helmet. Copenhagen

B. Consular type

1. Obv. has taller bust with long neck, broken legend; rev. has two stars

  • 187. End of loros hangs down between knees. Glendining, 26 May 1959, 224
  • 188. No loros between knees. Tolstoi 16
  • 189. Obv. of 188. Cahn 47, 17 May 1922, 1246
  • 190. Hirsch 53, 19-22 June 1968, 3608

2. Obv. with underbroken legend; rev. has one star

  • 191. Numismatic Fine Arts 12, 23-24 Mar. 1983, 507, 4.35 g = Bank Leu 13, 29-30 Apr. 1975, 553 = Kress 112, 22 June 1959, 964
  • 192. Münz. u. Med. 25, 17 Nov. 1962, 683
  • 193. New York City = Münz. u. Med. 13, 17-19 June 1954, 809, 4.41 g
  • 194. Tolstoi 15
  • 195. Ratto 252 = Naville 3, 16 June 1922, 286
  • 196. Dies of 195. Basel Münzhandlung 10, 15 Mar. 1938, 820
  • 197. Almost certainly obv. of 195, definitely rev. of 195. Hirsch 13, 9-10 July 1957, 958 = Balvin coll., 4.27 g
  • 198. New York City = Glendining, 27 Nov. 1958, 66, 4.27 g
  • 199. Obv. of 198; rev. in inferior style. Turin
  • 200. Birmingham = Glendining, 25 Nov. 1953, 220, 4.44 g
  • 201. Obv. of 200. Boudin & Bourgey, 14-15 Dec. 1911, 557
  • 202. Hirsch 31, 6 May 1912, 2064
  • 203. Dies of 202, same coin? Berlin 617/1912
  • 204. New York City = Sotheby, 2 Dec. 1924, 307 = Naville 3, 16 June 1922, 285, 4.47 g

3. As 2, but obv. with large lettering, and straight diadem; rev. with characteristic drapery over knees

  • 205. Rollin and Feuardent, 2-11 May 1898 (Hoffmann), 2239
  • 206. Dies of 205. Berlin 28779, 4.49 g
  • 207. Obv. of 205. Oxford, 4.50 g
  • 208. Obv. of 205. Vienna 30.179, 4.5 g
  • 209. Obv. with curved diadem but probably by the same hand as 208; sketchy rev. Schweizer. Kredit. Monetarium 30, Winter 1979/80, 171

4. Irregular coins

  • 210. Rev. with 2 stars in inferior style. Sangiorgi, 15 April 1907 (Strozzi), 2016
  • 211. Obv. with straight diadem; rev. with large, erratically aligned lettering, crude loros, etc. London

Die duplicates: 157 = 158, 195 = 196 = ? 197, ? 202 = 203, 205 = 206.

Obv. links: 166 = 167, 174 = 175, 177 = 178, 188 = 189, 198 = 199, 200 = 201, 205 = 207 = 208.

Rev. links: 165 = 166.

Basiliscus

  • 212. Dreer coll., Klagenfurt
  • 213. New York City, 4.44 g
  • 214. Obv. of 213. Lindpaintner coll. = Hirsch 137, 29-30 June 1983, 389, 4.47 or 4.45 g
  • 215. Münz. u. Med. 4, 5 Nov. 1945, 440, 4.4 g = Münzhandlung Basel 10, 15 Mar. 1938, 827
  • 216. Lanz (Munich) 18, 13 May 1980, 748, 4.39 g
  • 217. Vienna 30.255, 4.45 g
  • 218. Rev. of 217. Sternberg 11, 20-21 Nov. 1981, 992, 4.43 g
  • 219. Obv. of 218. Birmingham, 4.28 g (worn, has been mounted)
  • 220. Tolstoi 80, 4.35 g
  • 221. Schulman 258, 10-13 June 1974, 2086

Obv. links: 213 = 214, 218 = 219.

Rev. links: 217 = 218.

Zeno

1. Cuirass dividend into three panels

a. Cross with X-shaped bar

  • 222. Boldly dotted borders. Brussels = H. M. F. Schulman, 5-9 Apr. 1965, 1892
  • 223. Loudmer and Poulain, 15-16 June 1976, 300

b. Plain cross

  • 224. DOC 5
  • 225. Dorotheum 282, 11-13 Jan. 1966, 2
  • 226. Balvin coll. = Erste Österreichische Spar-Casse, Vienna, 4.33 g

b ? Possibly imitative

  • 227. The style of the obv. arouses some misgivings: the fully frontal portrait, the flat bold lettering with inconspicuous serifs, the middle panel of the cuirass, and the ear are slightly irregular. Stockholm

2. Cuirass divided into two panels

a. Cross, with X-shaped bar.

  • 228. Boldly dotted borders. DOC 6
  • 229. Birmingham = Ratto 285
  • 230. New York City, 4.45 g
  • 231. In commerce, 4.44 g

b. Plain cross

  • 232. Rev. of 224, one star removed. Hahn coll. = Rauch 22, 9-10 Mar. 1978, 922, 4.39 g
  • 233. Possibly obv. of 232. Fragerlie 601, 4.38 g
  • 234. Vatican
  • 235. Tolstoi 42, 4.35 g
  • 236. Copenhagen

Obv. links: ? 232 = 234.

Rev. links: 224 = 232.

Anastasius

MIB 14, TAS IVS

  • 237. Stockholm, 4.43 g, see ANSMN 13 (1967), Pl. 29, 9 = Fagerlie 676
  • 238. New York City, see Annual Report of the American Numismatic Society 1970, P1. 2, 2, 4.36 g
  • 239. Stockholm
  • 240. Numismatica 4, 22-23 Apr. 1974, 622
  • 241. Protonotarios coll., 4.43 g
  • 242. One star removed from coin, wreath ties very irregular. Bendall coll., 4.27 g

MIB 14, TA SIVS

  • 243. Sternberg, 28-29 Nov. 1975, 560, 4.44 g
  • 244. Kress 89, 17 Apr. 1944, 2425, 4.22 g
  • 245. BM [not seen]
  • 246. Münz. u. Med. FPL 13, Nov. 1948, 56 [not seen]
  • MIB 15, with wavy ornament on paludamentum, and cuirass, or on cuirass only
  • 247. Right star partially obliterated? Hajdučka Vodenica 4, 4.31 g
  • 248. Protonotarios coll., 4.38 g = Hirsch 79, 27-29 June 1972, 1368
  • 249. Münz. u. Med. 25, 17 Nov. 1962, 687, 4.36 g

MIB 15, with wavy ornament on paludamentum, Victory's wing in different style

  • 250. Belgrade National Museum
  • 251. Rollin and Feuardent, 20 Apr. 1896 (Montagu), 1085
  • 252. Dies of 251, right star obliterated. Coins 251-52 bear a marked stylistic resemblance to 269 which was struck from different dies, seen in London in March 1982, and which came from Syria or Lebanon. J. Elsen FPL 39, Dec. 1981, 74, 4.21 g = Bonhams 6, 14 Sept. 1981, 608

MIB 15, wing with horizontal shading

  • 253. Istanbul Arch. Mus.
  • 254. Tolstoi 80, 4.35 g
  • 255. Peus 284, 9-10 Dec. 1974, 1228, 4.36 g
  • 256. New York City, 4.34 g
  • 257. RIN 62 (1960), pl. 20, A, 4.33 g
  • 258. Tolstoi 78, 3.45 g [sic]
  • 259. Bonhams 3, 3 Dec. 1980, 9

MIB 15, simpler style of wing

  • 260. Ljubljana
  • 261. DOC 27, 4.36 g
  • 262. A. Portner coll., 4.17 g
  • 263. Left star obliterated. Kress 135, 15 Mar. 1966, 584
  • 264. Left star obliterated. Hajdučka Vodenica 3, 4.37 g
  • 265. Glasgow
  • 266. Right star removed. H. M. F. Schulman, 26 Feb. 1973, 25 = H. M. F. Schulman, 18-20 Nov. 1965, 28
  • 267. Tolstoi 79, 4.45 g
  • 268. BM [not seen]

Uncertain

  • 269. In commerce, 4.33 g

Die duplicates: 251 = 252

Justin I

MIB 6

  • 270. NS (for NVS), cuirass in three panels, central one with vertical ornament. Protonotarios coll., 4.4 g
  • 271. Cuirass in three panels. Berk coll., 4.51 g
  • 272. Tolstoi 102, 4.35 g
  • 273. Cuirass in three panels, right star obliterated. Bank Leu 13, 29-30 Apr. 1975, 571, 4.25 g
  • 274. Cuirass undivided. Belgrade National Museum
  • 275. Cuirass undivided. Bonhams 3, 3 Dec. 1980, 18, 3.44 g (clipped ?)

MIB 7

  • 276. Cuirass in three panels, upper border ornamented with wavy line. Paris, 4.30 g
  • 277. Cuirass in three panels. Hajdučka Vodenica 8, 4.35 g
  • 278. Cuirass undivided. Hirsch 79, 27-29 June, 1381 = Bonhams 3, 3 Dec. 1980, 17, 4.35 g

See also Justinian, 279. Die duplicaties: none.

Justinian I

  • 279. Obv. die of Justin I, MIB 7, NI A added. O'Hara coll., 4.35 g

1. Curls represented by dot to left and right of neck

MIB 21

  • 280. Cuirass undivided, three dots in panel, and wavy horizontal lines below. Hajdučka Vodenica 29, 4.36 g
  • 281. Wavy line in panel. Vienna (= NZ 69, 1936, pl. 1, 1)
  • 282. One wreath-tie, wavy line in panel. Hajdučka Vodenica 24, 4.44 g
  • MIB 20
  • 283. Wavy line in panel, three dots in vertical section below. Hess 249, 13 Nov. 1979, 460, 4.32 g
  • 284. Peus 299, May 1980, 797 = Hirsch 115, 6-10 Apr. 1979, 54 = Hirsch 79, 27-29 June 1972, 1396, 4.45
  • 285. Wavy line in panel. F�eg coll.
  • 286. Schulman, 8 June 1937, 319
  • 287. Ars Classica 15, 1930, 2166, 4.33 g
  • 288. Hajdučka Vodenica
  • 289. Right star on rev. obliterated. Schulman, 8 June 1937, 320
  • 290. Three tadpoles in vertical sections below. Bonhams 3, 3 Dec. 1980, 28, 4.36 g
  • 291. Vertical wavy lines on cuirass; obv. die somewhat distinctive. Mangakis Coll.
  • 292. Tolstoi 528 = Rollin and Feuardent, 20 Apr. 1896 (Montagu), 1097, 4.3 g
  • 293. Hirsch 79, 27-29 June 1972, 1395
  • 294. Taller helmet; rev. with carefully drawn cross. Kricheldorf 19, 20-29 June 1968, 54
  • 295. Single cross on rev., A of VICTORIA resembles star. N Circ. 86 (1978), p. 579. Sternberg 8, 16-17 Nov. 1978, 771, 4.21 g
  • 296. Single cross on rev., rough style. Hajdučka Vodenica 18, 4.38 g
  • 297. Bonhams 3, 3 Dec. 1980, 29, 4.36 g. Single cross on rev., cf. no. 290, a better photograph in N Circ. 86, 1978, 579

2. Horn-shaped curl to left of neck

MIB 20

  • 298. Hajdučka Vodenica 25, 4.26 g
  • 299. Protonotarios coll., 4.47 g
  • 300. Münz. u. Med. FPL 199, Apr. 1960, 46 = Dorotheum 228, 26-29 Jan. 1960, 21
  • 301. Hajdučka Vodenica 28, 4.31 g
  • 302. Stuttgart
  • 303. Hajdučka Vodenica 27, 4.28 g
  • 304. Hajdučka Vodenica 26, 4.36 g
  • 305. Hajdučka Vodenica 21, 4.40 g
  • 306. Obv. of 305. Hajdučka Vodenica 20, 4.35 g
  • 307. Obv. of 305. Hajdučka Vodenica 23, 4.37 g
  • 308. Dies of 307. Hajdučka Vodenica 22, 4.37 g

MIB 22

A. Style A

1. 8 curls and deep cuirass

2. 7 curls and similar cuirass

  • 310. Hess 247, 29 June 1978, 451, 4.37 g

3. 7 curls, line across throat

  • 311. Leu 13, 29-30 Apr. 1975, 576
  • 312. Birmingham = Glendining, 21-22 Sept. 1960, 913

4. 6 curls, line across throat

  • 313. Elsen FPL 32, May 1981, 102 = Elsen FPL 22, May 1980, 192, 4.42 g
  • 314. Athens
  • 315. Kricheldorf 14, 7-8 July 1964, 395 = Münz u. Med. FPL 247, Sept. 1964, 37
  • 316. Graz (Joanneum), pierced
  • 317. Paris, 4/Cp/Ꜹ/9, 4.42 g
  • 318. Münz. u. Med. 35, 16-17 June 1967, 216, 4.39 g

5. 5 curls, line across throat

  • 319. Münz. u. Med. FPL 156, Apr. 1956, 17
  • 320. Münzzentrum Köln 36, 7-9 Nov. 1979, 757, 4.28 g

6. 5 curls

  • 321. Kress 112, 22 June 1959, 968
  • 322. Tolstoi 17, 4.35 g
  • 323. Tolstoi 18, 4.4 g
  • 324. New York City = Kress 135, 15 Mar. 1966, 635, 3.77 g
  • 325. Vienna = Restle, pl. 3
  • 326. Obv. of 325. Brussels
  • 327. Belgrade National Museum
  • 328. DOC 7.4, 4.20 g
  • 329. Münz. u. Med. 52, 19-20 June 1975, 826
  • 330. Vinchon, 14 May 1982, 87, 3.58 g
  • 331. Protonotarios coll., 4.43 g
  • 332. Obv. of 331. Peck coll., 4.37 g
  • 333. Glendining, 7-8 Mar. 1957, 524

B. Style B

1. Other style of arm, 7 curls, deep cuirass

  • 334. Tellmann 179, Oct. 1972, 426 = Tellmann 147, Feb. 1970, 13 = Tellmann 131, Oct. 1968, 16

2. 6 curls, sometimes deep cuirass

  • 335. Zagreb
  • 336. DOC 7.2, 4.39 g
  • 337. Dies of 336. Stacks, 20-22 Jan. 1938, 1716
  • 338. Baldwins, 4.17 g
  • 339. Belgrade National Museum
  • 340. Kress 163, 8-11 July 1975, 1410

3. 5 curls, rougher style

  • 341. Karlsruhe (pierced)
  • 342. Ratto 449
  • 343. Dies of 342. Budapest
  • 344. Obv. of 342. Istanbul, inv. 329
  • 345. Tolstoi 19, 4.4 g
  • 346. New York City, 4.33 g
  • 347. Bucharest (Academy)
  • 348. Munich (Restle, pl. 3)
  • 349. DOC 7.1, 4.34 g
  • 350. Thessaloniki hd. 22

C. Style C

1. By a different die engraver?

  • 351. DOC 7.3, 4.46 g
  • 352. Münz. u. Med. FPL 394, Oct. 1977, 15 = Hahn Coll., 4.22 g
  • 353. Athens

2. Modem forgery?

  • 354. Münz. u. Med. FPL 390, June 1977, 22

MIB 23

  • 355. Rev. die in earlier style, could be altered MIB 22 die. In commerce, 1981 (Monte Judica hd. 58, 4.46 g)
  • 356. Obv. of 355. Thessaloniki hd. 23, 4.40 g
  • 357. Tolstoi 41, 4.2 g

MIB N23

  • 358. Bonhams 3, 3 Dec. 1980, 30, 4.46 g = Auctiones AG 8, 27-28 June 1978, 973
  • 359. Budapest
  • 360. Obv. of 359. Ljubljana, 4.00 g
  • 361. Obv. of 359, die flaws above cuirass through final C. Hahn coll. = Rauch 26, 23-25 June 1980, 420
  • 362. Obv. of 359, die flaws of 361. Birmingham
  • 363. Obv. of 359, die flaws of 361. Thessaloniki hd. 24, 4.32 g

MIB 24 and N24

  • 364. Helbing 63, 29 Apr. 1931, 1333
  • 365. Peus 294, 15-17 March 1978, 836, 1.45 g
  • 366. Sternberg 17, 9-10 May 1986, 717 = Münz. u. Med. FPL 461, Nov./Dec. 1983, 35, 1.44 g

See also Justin II, 419.

Die duplicates: 307 = 308, 336 = 337, 342 = 343.

Obv. links: 305 = 306, 325 = 326, 331 = 332, 342 = 344, 355 = 356, 359 = 360 = 361 = 362 = 363.

Justin II

MIB 16

1. Neat dies

  • 367. Experimental rev. die indicates breasts of seated figure. Birmingham
  • 368. Obv. of 367, rev. similar. RIN 62 (1960), pl. 20, C, 4.34 g
  • 369. Obv. of 367. Thessaloniki hd. 61, 4.31 g
  • 370. Protonotarios coll., 4.32 g
  • 371. Berlin
  • 372. Obv. of 371. Bucharest (Academy)
  • 373. Thessaloniki hd. 58, 4.34 g
  • 374. Bridge coll.
  • 375. Münz. u. Med. FPL 434, June 1981, 21, 4.48 g (Monte Judica hd. 82, 4.48 g)

2. Coarser dies with sharply bent knee

  • 376. Thessaloniki hd. 45, 4.37 g
  • 377. Obv. of 376. Thessaloniki hd. 46, 4.17 g
  • 378. Obv. of 376. Thessaloniki hd. 47, 4.32 g
  • 379. Thessaloniki hd. 48, 4.40 g
  • 380. Obv. of 379. Thessaloniki hd. 49, 4.35 g
  • 381. Thessaloniki hd. 54, 4.37 g
  • 382. Dies of 381. Thessaloniki hd. 55, 4.33 g
  • 383. Thessaloniki hd. 56
  • 384. Rev. of 383. Thessaloniki hd.57
  • 385. Dies of 384. Budapest (pierced)
  • 386. Thessaloniki hd. 59, 4.34 g
  • 387. Thessaloniki hd. 60, 4.37 g
  • 388. Thessaloniki hd. 50, 4.37 g
  • 389. Dies of 388. Thessaloniki hd. 51, 4.38 g
  • 390. Obv. of 388. Thessaloniki hd. 52, 4.37 g
  • 391. Obv. of 388. Thessaloniki hd. 53, 4.32 g
  • 392. Obv. of 388. New York City, 4.42 g
  • 393. Obv. of 388. Zagreb

MIB 17

1. S

  • 394. Thessaloniki hd. 68, 4.43 g
  • 395. Thessaloniki hd. 69, 4.33 g
  • 396. Thessaloniki hd. 70, 4.41 g
  • 397. Thessaloniki hd. 71, 4.31 g
  • 398. Dies of 397. Thessaloniki hd. 72, 4.35 g
  • 399. Dies of 397. Thessaloniki hd. 73, 4.38 g
  • 400. Obv. of 397. Thessaloniki hd. 74, 4.34 g
  • 401. Obv. of 397. Thessaloniki hd. 75, 4.32 g
  • 402. Obv. of 397. Thessaloniki hd. 76, 4.32 g

2. TS

  • 403. Obv. of 397. Thessaloniki hd. 77, 4.34 g
  • 404. Obv. of 397. Thessaloniki hd. 78, 4.30 g
  • 405. Dies of 404. Thessaloniki hd. 79, 4.32 g
  • 406. Obv. of 397. Thessaloniki hd. 80, 4.30 g
  • 407. Thessaloniki hd. 81, 4.31 g
  • 408. Dies of 407. Thessaloniki hd. 82, 4.42 g
  • 409. Dies of 407. Thessaloniki hd. 83, 4.34 g
  • 410. Obv. of 407. Thessaloniki hd. 84, 4.34 g
  • 411. Dies of 410. Thessaloniki hd. 85, 4.43 g
  • 412. Dies of 410. Thessaloniki hd. 86, 4.30 g
  • 413. Thessaloniki hd. 62, 4.40 g
  • 414. Obv. of 413. Thessaloniki hd. 63, 4.44 g
  • 415. Obv. of 413. Thessaloniki hd. 64, 4.40 g
  • 416. Obv. of 413. Thessaloniki hd. 65, 4.39 g
  • 417. Thessaloniki hd. 66, 4.41 g
  • 418. Obv. of 417. Thessaloniki hd. 67, 4.33.g

MIB N17

  • 419. Obv. die of Justinian. Thessaloniki hd. 25, 1.43 g
  • 420. Thessaloniki hd. 87,1.42 g
  • 421. Thessaloniki hd. 88,1.45 g
  • 422. Thessaloniki hd. 89,1.46 g

MIB N17

  • 423. Thessaloniki hd. 107, 1.45 g
  • 424. Thessaloniki hd. 108, 1.49 g
  • 425. Thessaloniki hd. 90, 1.47 g
  • 426. Rev. of 425. Thessaloniki hd. 91, 1.46 g
  • 427. Obv. of 426. Thessaloniki hd. 92, 1.45 g
  • 428. Obv. of 426. Thessaloniki hd. 93, 1.47 g
  • 429. Thessaloniki hd. 94, 1.44 g
  • 430. Dies of 429. Thessaloniki hd. 1.45 g
  • 431. Thessaloniki hd. 96, 1.47 g
  • 432. Same rev. die as 431. Thessaloniki hd. 97, 1.44 g
  • 433. Thessaloniki hd. 98, 1.46 g

Die duplicates: 381 = 382, 384 = 385, 388 = 389, 397 = 399, 404 = 405, 407 = 409, 410 = 412, 429 =430.

Obv. links: 367 = 369, 371 = 372, 376 = 378, 379 = 380, 390 = 393, 400 = 403, 413 = 416, 417 = 418, 426 = 428.

Rev. links: 383 with die duplicates 384 = 385, 425 = 426, 431 = 432.

Tiberius II

MIB 10

1. Year 6

  • 434. Rev. with boldly dotted cross. Thessaloniki hd. 112, 4.35 g
  • 435. Rev. with boldly dotted cross. Athens, old coll.
  • 436. Obv. of 435, rev. plain cross. The Hague (pierced)
  • 437. Obv. of 435, rev. plain cross. Berk coll., 4.35 g
  • 438. Thessaloniki hd. 111, 4.39g
  • 439. Obv. by different die engraver. Thessaloniki hd. 115, 4.39 g
  • 440. Obv. of 439; rev. not illustrated. Thessaloniki hd. 114, 4.36 g

2. Year 8

  • 441. Birmingham
  • 442. Obv. of 441. BMC 10, 4.38 g

MIB N11

  • 443. Thessaloniki hd. 113, 1.45 g

MIB NN11

  • 444. BMC 16, 1.47 g

Obv. links: 435 = 437, 439 = 440, 441 = 442.

Maurice

MIB V21, for the decennalia

  • 445. Bonhams 3, 3 Dec. 1980, 80, 4.29 g = Schulman 264, 26 Apr. 1976, 5598

MIB 21, for the vicennalia

  • 446. Tolstoi 33
  • 447. Copenhagen (Thomsen 330)
  • 448. Karlsruhe
  • 449. Kress 125, 17 Apr. 1963, 867 = Birmingham
  • 450. Birmingham

MIB 22, Tiberius Mauricius

  • 451. Tolstoi 47.
  • 452. Rev. of 451. Münz. u. Med. FPL 299, Apr. 1969, 35
  • 453. Probably obv. of 451. Istanbul = "More about the Minor... Mints," p. 553, 5
  • 454. Rev. of 453. Zagreb = "More about the Minor... Mints," p. 553, 6

MIB 23, Mauricius Tiberius

  • 455. Tolstoi 43

MIB N23, Tiberius on obv., Mauricius on rev.

  • 456. BM "More about the Minor ... Mints," p. 553, 7
  • 457. Oxford (Goodacre) = "More about the Minor ... Mints," p. 553, 8, 1.30 g

Obv. links: 451 and 453?

Rev. links: 451 = 452, 453 = 454.

Phocas

MIB—.

  • 458. In commerce, 1982. 4.38 g

Heraclius

MIB 81

  • 459. Paris 10/Cp/Ꜹ/8, Iϵ 4.33 g
  • 460. Dies of 459. Paris 10/Cp/Ꜹ/9 = "Unusual Gold Coins ...," p. 536, 1, 4.24 g (pierced)
  • 461. Berlin Iϵ = "Unusual Gold Coins ...," p. 536, 2
  • 462. Birmingham IГ? = "Unusual Gold Coins ...," p. 536, 3

MIB 82a, 82b

  • 463. G. Hirsch 35, 25-28 June 1963, 1276
  • 464. Ratto 1378 = "Unusual Gold Coins ...," p. 536, 4
  • 465. Münz. u. Med. 19, 5-6 June 1959, 284 = "Unusual Gold Coins ...," p. 536, 4, 4.39 g
  • 466. DOC 188b = "Unusual Gold Coins ...," p. 536, 6, 4.17 g
  • 467. Beckenbauer 2, 9 Mar. 1962, 372 = "Unusual Gold Coins ...," p. 536, 7
  • 468. Star between heads. Kress 131, 16 Nov. 1964, 951

Die duplicates: 459 = 460.

APPENDIX 4 Hoards and Finds of Fifth and Sixth Century Gold Coins from Illyricum

1. Hoards

The list below is arranged in the approximate order of the closing of the hoards. Date of discovery (when known), description, and references to published materials follow.69

Donji Lapac , 425. Before 1837, about 560 coins, of which 5 survive in Vienna; Constantine I, Theodosius II, Valentinian III. Mirnik 324; K. Patsch, Die Lika in der römischen Zeit (Vienna, 1900).

Glogović , 425. 1926, 22 solidi plus jewelry: 2 of Honorius, 1 of Valentinian III. Mirnik 325; Sergejevski, Glasnik zemaljskog muzeja u Bosni i Hercegovini 42, 2 (1930), pp. 119-32.

Kamnik , 425. 1842, about 300 coins: Honoria. Mirnik 327.

Mlakvanska Greda, 425. 2 (? + ) of Valentinian III. Mirnik 329.

Po�arevac , 443. Ca. 1901, 6 ( + ) solidi of Theodosius II: CONCORDIA AVGG (1: A); VOT XX MVLT XXX (3: Г, ϵ, H); VOT XXX MVLT XXXX (1: ?Z); GLOR ORVIS TERRAR (1: TESOB). Mirnik 331.

Gotse Delchev , fifth century. Ca. 1934-36, 30 coins listed: Honorius (1), Theodosius II (24), Pulcheria (1), Valentinian (2), Marcian (2). T. Gerasimov, "Trésors de monnaies trouv�s en Bulgarie pendant 1934-1936," BIAB 11 (1937), p. 322, s.v. Nevrokop.

Novi , fifth century. Mirnik 330.

Kyustendil region (Bŭrzovitsa) , 491. 1925-26, 5 coins listed: Leo I (1), Zeno (3), Anastasius (1). N. A. Mouchmov, "Trésors de monnaies trouv�s en Bulgarie pendant 1925-26," BIAB 4 (1926-27), p. 323.

Vakarel (Sofia), 491. One gold coin of Anastasius and an uncertain number of sixth-century copper coins. I. Welkov, "Funde aus verschiedenen Orten in Bulgarien," BIAB 14 (1940-42), p. 279.

Kaprije , 527. 1901, 2 (? + ): Anastasius (1); Justinian, trem. (1). Mirnik 339.

Sisak , 542? Ca. 1864, these 6 gold coins in the Zagreb museum presumably reflect a hoard or hoards: Theodosius II, GLOR ORVIS TERRAR, TESOB; Theodoric (in name of Anastasius), trem., cf. BMC 5 and 11 (star to right), and another, cf. BMC 67-70 (stars right and left); Justinian, solidus, full faced, cf. BMC 15, off. Z; imitative solidus, ߘONO;; imitative trem., star r. Mirnik 346.

Hajdučka Vodenica , 542. V. Kondić, "Le Trésor de monnaies d'or de Hajdučka Vodenica (limes danubien)," Caričin Grad 1 (Belgrad, 1984), pp. 179-88.

Peloponnese , 565? Ca. 1948, 5 solidi of Justinian, including one with inscription I⊖ⴱ, found in the center of the Peloponnese; subsequently 3 more were recovered, of Justinian and Justin II, perhaps part of a dispersed hoard of about a hundred coins. J. Tr�heaux, "Chronique des fouilles...en 1948," BCH 73 (1949), p. 520; H. Gallet de Santerre, "Chronique des fouilles...en 1949," BCH 74 (1950), p. 293.

Stobi , 565. 1932, found in the excavations of the Basilica, 5 solidi: 2 of Justinian and 3 of Justin II, Constantinople mint. "Koutrigours," p. 620; "Illyricum," pp. 459-60.

Grabovnik , 567. Of 23 solidi 19 are carefully described with weights: 4 are of Justinian (post-542), Constantinople mint, off. B, Δ, S, and ⊖ over H; 15 are of Justin II, and of these 1, without off. numeral, is presumably Thessalonican; 1 has С in the field (MIB 3a, off. Г); 2, from off. ⊖, are from the same die; the rest—A, Г, Δ (4), ϵ (2), S, I (2)— are all from different dies; average weight, 4.47, is high. Mirnik 338; "Koutrigours," p. 615 (concealment perhaps in 569).

Sekulica , 540? A very large hoard of gold ending with coins of Justinian. Glasnik Inst. za Nat. Istorila 26, 1 (1982), pp. 241-47.

Solin (Salonae) , 578. 5 solidi, of Justinian (1), Justin II (1), and Tiberius II (3). "Illyricum," p. 466.

Goren Kozjak (Bargala), 578. 1968, includes 13 gold coins together with copper coins: Justin II, solidi (2), trem. (5); Tiberius II, trem. (6). Blaga Aleksova and Cyril Mango, "Bargala: A Preliminary Report," DOP 25 (1971), pp. 273-75, and information kindly supplied by Prof. Mango.

Guberevac (Kosmaj, western Serbia), 578. 2 solidi of Justin II and Tiberius II. "Illyricum," p. 466.

Thessaloniki , 581 ? 1948, 83 solidi and 32 trem. (17 + 8 of Justinian I, 62 + 23 of Justin II, 4 + 1 of Tiberius I.). M. Oeconomides and J. Touratsoglou, "The 1948 Thessaloniki Hoard of 6th Century Byzantine Coins: A Contribution to the Study of the Mint of Thessaloniki," NumAnt-Class 8 (1979), pp. 289-312; Hahn (above, n.4); for the composition of the hoard by mints, see above, p. 73.

Narona (Vid), 582. 65 coins and jewelry. Justinian: 24 solidi, with facing bust (542 and later), and with a range of off. numerals. Justin II: 24 solidi, with a range of off. numerals; on 2 specimens the numeral is lacking, and these are presumably Thessalonican; there are 4 specimens from off. S, and some or all of these could presumably be Thessalonican. 6 trem. (originally attributed to Justin I, but the age structure of the hoard makes this unlikely) which could in principle include Thessalonican issues. Tiberius II: 5 solidi, off. B, Δ (2), ⊖, and S (this last could in principle be Thessalonican) and 5 trem. (MIB 9a). Maurice: 1 trem. (MIB 18b). F. Bulić, VAHD 25 (1902), pp. 197-212; Mimik 352.

Athens (Osteotheke), 583/4. 7 gold coins. Tiberius II, solidi (5), MIB 4, off. Δ, Z, ⊖, I; sketchy style, off. I. Maurice, solidus, MIB 6, off. I; trem., MIB 20. D. M. Metcalf, "The Slavonic Threat to Greece circa 580...," Hesperia 31 (1962), p. 157, 435-38, and further details from casts in Ashmolean Museum, Oxford.

Sadovets , 582/3. Three hoards have recently been carefully listed with details of types and officinae by Mme I. Yurokova. They are to be published in full in a volume on Sadovets by the Bavarian Academy of Sciences. It appears that there is at least one earlier hoard, namely that published in 1934.

Hoards B-D are summarized above, pp. 70-71. In hoard B the 5 solidi of Justinian are all of MIB 7, and the 3 solidi of Maurice are as MIB 4B, 5D, and 6E. In hoard C the 3 solidi of Justinian are MIB 6. Three of the solidi of Maurice are MIB 6E and the fourth is a light weight solidus, MIB 11E. In hoard D, 2 of the solidi of Justinian are MIB 5 (before 538) and there are thus 3 old solidi in the hoard. The other 2 solidi of Justinian are apparently all MIB 7. The 20 solidi of Maurice are MIB 4B and 6E. The hoard of 1934 and what appears to be yet another hoard, briefly described by Kovacević, may be summarized as follows:

Hoard of 1934 ? Further Hoard
Sol. Sem. Trem. Sol. Trem.
Justinian 8 - 2 3 -
Justin II 17 - 5 7 -
Tiberius II 7 1 8 4 1
Maurice 5 - 1 3 -
37 1 16 17 1

"Illyricum," p. 469, n. 8 = Hoard C; p. 470, n. 3 (cf. further hoard?), and the hoard of 1934 (to which the 50 bronze belong, according to Mosser).

Kupusina , 602. Solidi, from Zeno to Phocas. "Avari i zlato."

Padania , 602. 5 solidi, of Maurice (2) and Phocas (3). F. Chamoux, "Chronique des fouilles... en 1946," BCH 71-72 (1947-48), p. 393.

Gornja or Donja Vre�ina , 610. Three light weight solidi of Heraclius alone, MIB 54, die linked. A fourth light weight solidus, of Heraclius with Heraclius Constantine, likewise in the Ni� museum, might be from the same presumed hoard or, if not, then from the same historical context of the capture of Naissus by the Avars, at a date between 611 and 616. (There is just 1 weight solidus in the 3 listed Sadovets hoards.) "Koutrigours," pp. 628-29; "Illyricum," pp. 493-96.

End Notes
69
Abbreviations used in Appendix 4 are the following:
BIAB: Bulletin de l'institut archéologique bulgare.
BSAB : Bulletin de la société archéologique bulgare.
"Avari i zlato" : J. Kovačević, "Avari i zlato," Starinar 1963, pp. 125-35.
Simplozijum : J. Kovačević, "Archeolo�ki prilog preciziranju hronologije slovenskog naseljavanja Balkana," Simpozijum predslavenski etnički elementi na Balkanu i etnogenezi ju�nik Slovena 12 (Posebna Izdanja, Sarajevo 1969), pp. 57-83.
Mirnik : I. A. Mirnik, Coin Hoards in Yugoslavia, BAR International Series 95 (Oxford, 1981).
"Illyricum" V. Popović, "Les t�moins archéologiques des invasions Avaro-Slaves dans l'llyricum byzantin," M�langes de l'école française de Rome. Antiquité 87 (1975), pp. 445-504.
"Koutrigours" : V. Popović, "La descente des Koutrigours, des Slaves et des Avares vers la Mersg�e: le t�moignage de l'arch�ologie," Académie des Inscriptions et Belles-Lettres, Comptes Rendus 1978, pl. 596-648.
VAHD Vjesnik za arkheologiju i historiju dalmatinsku.

2. Finds

Sarajevo Museum, 55 gold coins, without exact provenance except as stated, and provisionally identified. Arcadius: solidi of Constantinople (1) and Milan (1); trem. found at Izačić. Theodosius II: solidi of Constantinople, VOT XX MVLT XXX (2), VOT XXX MVLT XXXX (1). Marcian: solidi of Constantinople (1) and Thessalonica (1), GLOR ORVIS TERRAR. Pulcheria: imitative trem., bought in Kladanj. Leo I: 6 Constantinopolitan solidi from various sources, including 1 excavated in the Vranplanina; 3 trem. Zeno: 4 Constantinopolitan solidi; 3 trem. of which at least 1, from Skender Vakuf, is Ostrogothic (Odovacer). Anastasius: 8 Constantinopolitan solidi, including specimens from the western Hercegovina (Hum), Knin, and Donji Lapac (not from the hoard?); 4 trem. of which 2 are Ostrogothic, from the Morača valley and the �panjac region respectively (1 of the imperial trem. is from Hum). Justin I: probably Justin II. Justinian: Constantinopolitan solidi, 527-42 (2), and 542-65 (1); Ostrogothic solidi, AVCCCA, COMOB (3); 8 trem. including specimens from Glamoč, the Petrovac region, southern Albania, and Albania; Ostrogothic (1); (?) a gilt half-siliqua. Justin II: 2 trem., from Stolec and �rmanjci (the latter apparently antedates the nearby Vid hoard). K. Patsch, "Nahogjaji novaca," Glasnik zemaljskog muzeja u Bosni i Hercegovini 12 (1900), pp. 543-73.

Leskovac , 10 coins, found in the neighborhood of Leskovac except where specified more exactly. Theodosius II: GLOR ORVIS TERRAR, TESOB. Eudoxia: semissis, cross in wreath, CONOB*, from Gornje Lisarje, Ora�ec. Zeno: solidus, Constantinople. Anastasius: solidus, MIB 7?; semissis. Justinian: solidus, MIB 7; 3 trem. Phocas: solidus, Constantinople. Epirus , the Troianski collection is believed to have been built up from local finds made in Epirus. Whether all the following coins are from Epirus is perhaps uncertain. 61-63, Theodosius II: GLOR ORVIS TERRAR, CONOB (2), TESOB (1). 66-68, Marcian, Pulcheria: Constantinople. 69-72, Leo: Constantinople. 73-75, Zeno: Constantinople (2), COMOB (1). 76-78, Basiliscus: Constantinople. 79-82, Anastasius: Constantinople. 87-92, Justin I: 88 might be from Thessalonica, the others from Constantinople. 98-101, Justinian: 98-99 are mentioned by Oeconomides and Touratsoglou in their discussion of the Thessaloniki hoard, the others from Constantinople. 132-33, Justin II: Constantinople. J. N. Svoronos, "Ethnikon Noumismatikon Mouseion...," JIAN 11 (1908), pp. 274-86.

Beljarica . Anastasius: solidus, Constantinople. "Avari i zlato."

Peru�ić , contemporary forgery of a gold coin of Anastasius, with gold shell and bronze core. VAHD 4 (1881), March, cover.

Corinth . Justinian: solidus, Constantinople. G. Daux, "Chronique des fouilles...en 1962," BCH 87 (1963), p. 730, fig. 15.

Goranovtsi . Justinian: one gold coin. B. Filow, "D�couvertes archeologiques en Bulgarie pendant 1914," BSAB 4 (1914), p. 282.

Sisak . Justinian: solidus, Constantinople. "Avari i zlato" (? = the coin in the hoard listing, above).

Mitrovica. Justin II: semissis, Ravenna? D. M. Metcalf, "The Currency of Byzantine Coins in Syrmia and Slavonia," Hamburger Beitr�ge zur Numismatik 14 (1960), no. 4. Beograd district. Maurice: 2 solidi, Constantinople (Belgrade museum, exact find spots unknown). Simpozijum, p. 65.

Po�arevac (Viminacium) . Heraclius alone: solidus. "Illyricum," p. 496.

Ni� . Heraclius with Heraclius Constantine: solidus. See hoard list, Gornja/Donja Vre�ina.

Skopje district. Heraclius. Grbić, Glasnik Skopskog naučnog dru�tva 5 (1929), p. 51.

Pristina (Ulpiana) . Heraclius: solidus. Simposijum, pp. 73-74.

End Notes

1
RIC 9, pp. xxvi-xxvii.
2
F. Baratte, "Lingots d'or et d'argent en rapport avec l'atelier de Sirmium," Sirmium 8 (1978), pp. 99-109, for a recent survey; also O. Iliescu, "Un nou lingou roman de aur, descoperit �n Transilvania," Revista Muzelor 2 (1965), pp. 6-11 (Thessalonican, from Feldiora). The ingots are a phenomenon associated especially with the last quarter of the fourth century.
3
J. P. C. Kent, "Gold Coinage in the Late Boman Empire," pp. 90-204.
4
W. Hahn, "New Light on the Thessalonican moneta auri in the Second Half of the Sixth Century," NC 1981, pp. 178-82.
5
Appendix 1, pp. 80-83.
6
DOC 1, p. 26.
7
Fagerlie, p. 66.
8
M. Oeconomides-Caramessini and J. Touratsoglou, 'The 1948 Thessaloniki Hoard of 6th Century Byzantine Gold Coins: A Contribution to the Study of the Mint of Thessaloniki," NumAntClass 8 (1979), pp. 289-312.
9
For the closing date, MIB 3, pp. 90-91.
10
Hendy, pp. 129-54.
11
P. Lemerle, "Invasions et migrations dans les Balkans depuis la fin de l'�poque romaine jusqu'au VIIIc siècle," Revue historique 211 (1954), pp. 265-73, and Philippes et la Macedonie orientate, 1945, p. 82. The Synekdemos of Hierocles describes the prefecture of Illyricum ca. 527 as consisting of the following 13 provinces with their capitals: Macedonia I (Thessalonica), Macedonia II (Stobi), Thessaly (Larissa), Hellas (Corinth), Crete (Gortyna), Epirus vetus (Nicopolis), Epirus novus (Dyrrachium), Dacia interior (Serdica), Dacia ripensis (Ratiaria), Dardania (Scupi), Prevalitana (Doclea), Moesia I (Viminacium), and Pannonia (Sirmium).
12
Hendy, pp. 129-54.
13
Summary in W. Hahn, "Die �stliche Gold- und Silberpr�gung unter Theodosius II," LNV (Vienna, 1980), pp. 103-28, at p. 122.
14
Fagerlie, passim.
15
Fagerlie, passim.
16
K. Kovacević, "Avari i zlato," Starinar 13, 14 (1962-63), pp. 125-35.
17
A. Jeločnik, "Les multiples d'or et de Magnence d�couverts × Emona," RN 1967, pp. 209-35.
18
M. Vasić and V. Popović, "Un trésor de monnaies d'or de Sirmium," MEFRA 87, 1 (1975), pp. 425-43.
19
M. Vasić and V. Popović (above, n. 18), p. 425, revising the earlier publication of the Borca hoard.
20
V. Kondić, "Nalaz rimskih zlatn�ka kod U�ća blizu Obrenovca," Starinar 18 (1967), pp. 207-12; CH 1975, no. 211.
21
Monatsblatt Num. Gesell, in Wien 10 (1910), pp. 59-60; Jeločnik (above, n. 17), p. 209.
22
V. Popović, "Les t�moins archéologiques des invasions Avaro-slaves dans l'Illyricum byzantin," MEFRA 87, 1 (1975), pp. 445-504, at pp. 460-61.
23
E. Stein, "L'administration de la Dalmatic byzantine," in Histoire du Bas-Empire, 2 (Paris, 1949), pp. 801-2, observes that in Menander, A.D. 579, Dalmatia is clearly distinguished from Illyricum whereas by March 592 it is part of the prefecture. He does not doubt that the change was made at the same date as the exarchates of Italy and Africa were instituted, i.e. before 584.
24
See Appendix 4, below.
25
The numismatic and other archaeological evidence is more fully discussed in Popović (above, n. 22), pp. 469-70.
26
Popović (above, n. 22), pp. 464-65.
27
See Appendix 4, below.
28
See Appendix 4, below.
29
See Appendix 4, below.
30
K. Patsch, "Nahogjaji novaca," Glasnik Zemaljskog Muzeja u Bosni i Hercegovim 12 (1900), pp. 543-73.
31
Ż. Demo, "Novae germanskih vladara druge pol. 5. do u drugu pol. 6. st. u. Numizmatičkoj zbirci Arheolo�kog muzeja u Zagrebu," Arheolo�ki Vestn�k 32 (1981), pp. 454-81.
32
S. Du�anić, "Iz numizmatick� zbirke Narodnog muzeja u Leskovcu," Leskovački Zborn�k 1 (1961), pp. 93-100.
33
I. Marović, "Stanje i problemi numizmatickou kabineta arheolo�kog muzeja u Splitu," NV Year 13, no. 24 (1966), pp. 36-42. (The numbers of coins of Theodosius II and of Marcian are not stated, but are pre sumably smaller.)
34
Above, n. 8; some of the attributions are revised by Hahn (above, n. 4).
35
Popović (above, n. 22), pp. 449-50.
36
D. M. Metcalf, "Coinage and Coin Finds Associated with a Military Presence in the Medieval Balkans,' Kovanje i kovnice antičkog i srednjouekovnog novca, ed. V. Kondi� (Belgrade, 1976), pp. 89-97.
37
Originally commended to numismatists in 1970 by C. S. S. Lyon, in H. R. Mossop, The Lincoln Mint, c. 890-1279 (Newcastle-upon-Tyne, 1970), pp. 16-17 and table 4, the formula and other similar ones are discussed at some length in Statistics and Numismatics, PACT 5 (1981).
38
D. M. Metcalf, Coins of Henry VII, SCBI 23 (1976), pp. xxvi-xxvii.
39
Hahn (above, n. 13), pp. 112-13, but more recently in "Die �stliche Gold- und Silberpr�gung unter Theodosius II," LNV 1 (1979), pp. 101-6, he prefers to stretch the Thessalonican GLOR ORVIS TERRAR solidi of Theodosius II over the whole period 424-50 with only short interruptions for the VOT XXX MVLT XXXX and the VIRTUS ЄXЄRC ROM issues.
40
J. P. C. Kent, "Auream monetam cum signo crucis," NC 1960, pp. 129-32.
41
M. Vickers, "Fifth-Century Brickstamps at Thessaloniki," Annual of the British School at Athens 68 (1973), pp. 285-94; A. H. M. Jones et al., Prosopography of the Later Roman Empire, vol. 2, s.v. Hormisdas. I am indebted to my colleague Mr. Vickers, who discussed this problem with me and permitted me to consult his unpublished notes. But see most recently J. M. Spieser, "La ville en Gr�ce du iii e au vii e siècle," in Villes et peuplement dans ľIllyricum protobyzantin, Collection de l'école française de Rome 77 (Borne, 1984), pp.318–19, and particularly the comment on p. 129.
42
G. Downey, "Earthquakes at Constantinople and Vicinity, A.D. 342-1454," Speculum 30 (1955), pp. 596–600. There seems, unfortunately, to be some confusion in the sources between the earthquakes of 437 and 447, both of which were severe, and in one of which the walls with many of their towers collapsed at Constantinople.
43
P. Lemerle, Philippes et la Mac�doine orientale (Paris, 1945), p. 82, n. 3; E. Demougeot, De l'Unit� × la division de l'Empire romain, 395-410. Essai sur le gouvernement impérial (Paris, 1951), p. 502, n. 39, offering an argument in favor of 437. J. W. E. Pearce, "Gold Coinage of the Beign of Theodosius I," NC 1938, p. 242, suggests that the moneta auri at Sirmium migrated to Thessalonica in 395.
44
The problem has connections with that of the cult of St. Demetrius.
45
V. Kondić, "Le Trésor de monnaies d'or de Hajdučka Vodenica (limes danubien)," Caričin Grad 1 (Belgrade, 1984), pp. 179-88.
46
V. Popović, "La Descente des Koutrigours, des Slaves et des Avares vers la Mersg�e: le t�moignage de l'arch�ologie," Academie des Inscriptions et Belles-lettres 1978, pp. 596-648.
47
Sec Appendix 3, below.
48
V. Popović, "Une Invasion Slave sous Justin II inconnue des sources �crites," Numizmatičar 4 (1981), pp.111-26.

SEVENTH-CENTURY BYZANTINE COINS IN SOUTHERN RUSSIA AND THE PROBLEM OF LIGHT WEIGHT SOLIDI

John Smedley

Adelson's monograph on the Byzantine light weight solidi, though the first systematic study of that coinage, failed to provide satisfactory solutions to the problems surrounding the coins. Subsequent work, while pointing out the deficiencies of his proposals and putting forward alternatives, in its turn has reached no conclusive answers.1 Futhermore, attention has focused largely upon the period from Justinian I to the first decades of the seventh century, while later coins have been relatively neglected. This is due in part to their rarity—of some types only a single specimen is known—and in part to the difficulty of access to the material, for most of it is in the U.S.S.R. and much is unpublished. Soviet scholars have, to some extent, analyzed the distribution of seventh-century Byzantine coins in the U.S.S.R.,2 but have not, to my knowledge, recognized the peculiarity of the light weight solidi. Again, this situation may be largely due to the lack of information, for it is usually necessary to rely on descriptions of varying detail, which may or may not be accurate. This study focuses on the distribution of Byzantine gold coins in the steppe zone of southern Russia, in particular the 20-carat solidi, for this distribution does reveal suggestive patterns and can thus contribute to our understanding of the use and purpose of light weight solidi, as well as to our knowledge of the history of southern Russia.

The list of finds given below relies principally upon that established in Kropotkin's invaluable book on the hoards of Byzantine coin in the territory of the U.S.S.R.3 and concentrates on gold coins from the reign of Heraclius to that of Constantine IV, with whom the production of 20-carat solidi appears to have come to an end. It also largely excludes finds from areas that are not relevant to the present topic, in other words outside the steppe zone of southern Russia and the regions immediately adjacent to it. Where possible, cross-references to MIB 3 have been given, and Hahn's classification and dating have been followed. It is assumed that none of the coins discussed represent new types, as there are no new types among the coins for which detailed information is available. The first 11 finds are from the steppe and are listed, roughly, from west to east.

image

Finds and sites listed:

  • Pechenaya
  • Zhabotin
  • Maistrov
  • Kelegeiskie khutora
  • Novye Sendzhary
  • Maloe Pereshchepino
  • Makukhovka
  • Kupyansk district ( ?)
  • Serpovoe
  • Serpovoe
  • Romanovskaya stanitsa
  • Cherson
  • Pampuk-kai
  • Uyutnoe
  • Kerch'
  • Kerch'-Taman' area
  • Chemburko khutor
  • Sukko
  • Krymskaya
  • Starodzherelievskaya stanitsa
  • Amgata, Upper Kuban'
  • Kislovodsk-N. Osetia group
  • Verkhnyi Chiryurt
  • Alechdar
  • Dnepropetrovsk
  • Poltava (Museum)
  • Zolotonosha
  • Poltava
  • Nekhvoroshcha
  • Khortitsa
  • Dniepr estuary
  • Nizhnaya Duvanka

Other finds mentioned:

  • Sennaya (Phanagoria)
  • Chufut-kale
  • Beloyarovka
  • Nokalakevi
  • Finds of 7th-century coins in Georgia and Abkhazia

1. Pechenaya : near Rovnoe, between the southern Bug and the Ingul; a burial with a solidus of Heraclius and Heraclius Constantine, probably of ca. 625-29.4

2. Zhabotinq : near Kamenka, south of the confluence of the Ros' and the Dniepr; a 20-carat coin of Heraclius and his sons, dating to the years 639-41, was bought in the village.5

3. Maistrov island : near Solenoe, a little below the Dniepr rapids; a clay pot with several hundred Byzantine gold coins. The only one to have survived and be identified is of Heraclius and his sons, of full weight and, from the description, almost certainly of the years 639-41.6

4. Kelegeiskie khutora: on the southern side of the Dniepr estuary; a burial with six gold coins of Heraclius and one of Constans II. Kropotkin gives no description of these. Bauer states that, according to the information he received, all the Heraclian coins of the three-emperor type had the mint mark BOXX in the exergue, indicating that they were of 20 carats and should date to 637/8 or 638/9.7 Unfortunately, it is not known how many of the Heraclian coins were of this type, nor what the Constans coin was. With these, possibly, was a light weight solidus of Justinian I.8 Most of the coins were adapted for use in a necklace.

5. Novye Sendzhary: also known as Zachepilovka, just south of Poltava; a burial with one solidus of Phocas, four of Heraclius with his two sons of the period 632-41, and two 20-carat coins of Constans II from the years 642-46. Kropotkin indicates that all the Heraclian coins had the mark CONOB in the exergue, showing they were full weight.9

6. Maloe Pereshchepino: near 5, above; an extremely rich find of gold, silver and other objects (over 21 kilograms of gold objects were discovered), including 69 Byzantine gold coins of the period from Maurice to Constans II. Sixty-one is the total usually given, but to this must be added eight coins originally in the Khanenko collection and transferred in 1926 from the State Academy of the History of Material Culture (GAIMK) to the Hermitage. This find is of the greatest importance, for not only is it the largest, but illustrations of the group of 61 coins have been published.10 Almost all, if not in fact all, the coins are pierced or otherwise transformed into jewelry with the addition of loops or mounts for stones. This makes any assessment of their original weight very difficult, and the mounts occasionally obscure the coins. Nevertheless, in most cases sufficient detail can be made out from the plates published by Kropotkin to enable a positive identification to be made;11 in particular, the distinctive marks in the exergue of the 20-carat coins BOXX, BOXX + or OBXX are generally clear.

Chronologically, there is a solidus of Maurice, followed by two of Phocas,12 then six coins of Heraclius and Heraclius Constantine, datable from 629 to January 632.13 Four of these appear to be normal weight solidi, but two, although struck with ordinary dies, are much heavier— 11.19 and 11.13 g, perhaps corresponding to a 21 solidus multiple. So far as is known, these two coins are unique and must represent an exceptional issue; the reverse dies may be identical.14 The next group of Heraclian coins published by Kropotkin, 35 in all (over half the total), is of the year 637/8 and consists of two varieties: one with BOXX in the exergue (27 specimens),15 the other with BOXX 4- (eight specimens).16 In both cases these are light weight solidi of 20 carats. They are followed by three more 20-carat solidi of either 637/8 or 638/9,17 two of 639-41,18 and 19 of Constans II from the period 642-46.19 In addition, there is one full weight solidus of Heraclius and his sons, all crowned, so minted between 636 and 641.20

7. Makukhovka : close to Poltava; a find of gold objects with one coin of Heraclius and his sons, 632-41.21

8. Kupyansk uezd ( ?): in the oblasť of Khar'kov; a solidus of Constans II and Constantine IV, 654-59.22

9. Serpovoe : far to the northeast near Morshchansk, by the river Tsna; a burial with twelve solidi of Heraclius and his sons and of Constantine IV. As Kropotkin quotes the opinion of Oreshnikov, dating the Heraclian coins to 638-41,23 presumably all the figures were crowned; according to Hahn they would then fall into the period 636-41.

10. Serpovoe: from the bank of the river Tsna; two stray full weight solidi, both pierced, of Constantine IV and his two brothers, ca. 674-81.24

11. Romanovskaya stanitsa: on the lower Don; a burial with a full weight solidus of Constantine IV alone, 681-85, and another of Leontius, 695-98; Kropotkin assigns the latter to Leo III, but the plump face makes it clear it is of Leontius.25

The next nine finds are from the coast.

12. Gherson: near Sevastopol'; a gold coin of Phocas.26

13. Pampuk-kai: in the hills of the southwestern Crimea; a burial in the church. In the mouth of the skeleton was a full weight coin of Heraclius and his sons, all crowned, of the years 638-early 641.27

14. Uyutnoe: from the seashore near Sudak; a full weight solidus of Constantine IV, ca. 674-81.28

15. Kerch': from the "Hill of Mithridates"; a full weight solidus df Constantine IV and his brothers, probably 669-ca. 674.29

16. Kerch' or Tainan' region: a solidus of Heraclius and his sons, 632-41.30

17. Chemburko khutor: across the Kerch' strait, near Anapa; a 20-carat coin of Constans II with his three sons, ca. 662-67.31

18. Sukko: near Anapa; a find of 14 seventh-century Byzantine gold coins and, so it is said, four silver ones; Kropotkin wonders whether the latter were not a separate find. Of these, five gold coins have been preserved, all of full weight and in exceptional condition: three of Constans II with his three sons, ca. 662-67, and two of Constantine IV and his brothers of 669-ca. 674. The two identifiable silver coins are hexagrams of Constantine IV of the same period.32

19. Krymskaya: south of the lower Kuban'; a full weight solidus of Heraclius and Heraclius Constantine of 613-29.33

20. Starodzherelievskaya stanitsa: near Krasnoarmeisk, north of the lower Kuban'; a burial with a full weight solidus of Phocas, pierced.34

In the northern Caucasus, finds are grouped in three areas.

21. Amgata: on the upper reaches of the Kuban' in Karachaevo-Cherkessia; a burial with a pierced, full weight solidus of Constantine IV with his brothers.35

22. From Pyatigorsk and Kislovodsk: to the southeast, principally in the Kabardino-Balkar A.S.S.R. and the North Osetian A.S.S.R.; six solidi of Phocas, four of Heraclius and Heraclius Constantine, two of Heraclius with his two sons, and one of Constans II with Constantine IV. In addition, there are many imitations of these coins, impressions made in thin gold or silver gilt sheet; these are most common for the coins of Heraclius and Heraclius Constantine, and include a few imitations of those of Constans II with Constantine IV, and of Constantine IV.36 One coin, from a burial at Galiat, appears to be of 23 carats—a coin of Heraclius and Heraclius Constantine;37 whether others are light weight, it is impossible to tell, but the descriptions, where given, imply that they are at least not of 20 carats. Either the mark CONOB is mentioned, or the coin is assigned to the mint of Constantinople which, as used by Kropotkin, means the coins had the mark CONOB. Since the majority of the imitations copy only coin obverses, there is no way of knowing what the originals were.

23. Upper Chiryurt: in northern Daghestan; its site and cemeteries contained imitations of solidi of Heraclius and Heraclius Constantine.38

In addition, there are finds in the steppe zone which cannot either be identified or precisely located, and a few finds of coins or imitations in metals other than gold.

24. Alechdar: near Rezina in Moldavia; from the excavations of a settlement there, a follis of Heraclius of 612/3, overstruck on one of Phocas.39

25. Dnepropetrovsk: just west of the Dniepr, near the town; a burial with 72 cast, silver gilt imitations of the 20-carat solidi of Constans II from the years 642-46. All were pierced for use in a necklace, and all are so similar as to suggest to Golenko, who published four of them, that they were copied from just one original, although several molds were used.40

26. Poltava Museum: from this collection Adelson published two pairs of 20-carat solidi, the first being coins of Heraclius and his sons from 637/8,41 the second of Constans II from the period 642-46.42 Unfortunately, the records relating to these coins were destroyed in the Second World War,43 but Adelson wondered if they might not have come from the finds at Maloe Pereshchepino or Novye Sendzhary. This is less likely than he believed, for the 61 coins commonly referred to as constituting the "hoard" of Maloe Pereshchepino are all in the Hermitage, while the additional eight from the Khanenko collection were transferred there from the GAIMK in 1926. Also, all 16 of the coins of Constans II published by Kropotkin (i.e. excluding the three from the Khanenko collection) have the mark of officina IO,44 but Adelson's two have that of the ninth officina. The coins from the burial at Novye Sendzhary certainly included two 20-carat pieces of Constans II and were housed in the Poltava Museum—up to 1941, according to Kropotkin—but those of Heraclius, he states, had the mint mark CONOB,45 not the BOXX of the 20-carat coins published by Adelson. Unless Kropotkin is wrong, therefore, Adelson's coins are unlikely to have come from that burial; and, given the discrepancies already noted, it is not very likely that they came from Maloe Pereshchepino. On the other hand, it is certainly possible, if not rather more probable, that they came from one of the other finds made locally, such as that at Makukhovka, 7, or the two listed below, 28 and 29.

27. Zolotonosha: opposite the confluence of the Ros' with the Dniepr; in 1957 M. F. Ponomarenko obtained six Byzantine solidi, found locally, including one of the emperor Maurice.

28. Poltava : near the town; a find of 20 Byzantine gold coins.

29. Nekhvoroshcha : south of Poltava; one seventh-century Byzantine gold coin.

30. Khortitsa island in the Dniepr, just below the rapids; a find of some Byzantine gold coins.

31. Dniepr estuary: on the southern shore; a hoard of Byzantine gold coins of which 33 went into the collection of E. E. Lyutsenko.

Of these last five finds,46 that at Nekhvoroshcha, 29, is almost certainly relevant to the present discussion. Also relevant, perhaps, are the finds at Zolotonosha, as suggested by the coin of Maurice (one of his coins was found at Maloe Pereshchepino), and the finds at Poltava, 28. Such finds of Byzantine gold, especially the larger ones, occur principally in the seventh century and in the Kievan period, after the ninth century.47 Since the Poltava region never formed part of the early Kievan state, it is more than likely that 28, as 29, is of the seventh century and that both are contemporary with the coins from Maloe Pereshchepino and Novye Sendzhary; certainly there are no Byzantine coins of any other date known from the region. The finds at Khortitsa and by the Dniepr estuary, however, may well date to a later period, after the tenth century, when the river became an important artery of communication and trade between Byzantium and Kiev.

32. Nizhnaya Duvanka: west of Kupyansk; a copper coin of Constantine IV and his brothers,48 which is the only coin in a metal other than gold to be found in the European Bussian steppe.

In addition to the odd finds of silver and copper near Anapa and in the northern Caucasus, seventh century copper coins occur quite often in the Crimea at Cherson (where a Byzantine mint functioned under Heraclius and, most probably, under Constans II),49 in the Kerch' peninsula,50 and there is one from the site at Sennaya (the ancient Phanagoria) in the Tainan' peninsula.51 A single silver coin of Heraclius from the years 619-25 was found in a burial at Chufut-kale, in the hills of the southwestern Crimea.52

Certain distribution patterns emerge from this list. Finds of copper coin are concentrated (with the sole exception of the coin from Nizhnaya Duvanka, 32, on the fringes of the steppe in areas under Byzantine rule, as in the Crimea, or adjacent to spheres of regular Byzantine influence, as in the northern Caucasus;53 both are also areas where settled populations existed. The coin from the sixth- to seventh-century settlement at Alechdar in Moldavia, 24, should probably be linked with similar finds in Romania.54 The distribution of silver coin is similar, and examples north of the Caucasus may be due to contacts with eastern Georgia and Armenia, where such coins predominate.55 Byzantine silver also occurs in the Balkans and Central Europe 56 and, far to the northeast, in the Kama-Vyatka region of Russia.57 In contrast, gold dominates the Black Sea steppe and, to a lesser extent, the lands under Avar control in the middle Danube and in Transylvania 58 and is usual in the northern Caucasus—all areas of nomadic, or semi-nomadic populations.59 It is also predominant in Abkhazia and in Georgia west of the Likhian mountains (the ancient Lazica),60 but these regions are not otherwise connected with the steppe zone under discussion and so may be left aside. It is interesting to note, however, that a find of 23-carat coins, of the emperor Maurice, comes from the old capital of Lazica, a hoard from Nokalakevi.61 If the coin from the burial at Galiat, 22, is indeed of 23 carats, this could provide another link to suggest that the coin in the northern Caucasus reached there from the lands across the mountains to the south.

Looking in detail at the finds in the Russian steppe, one feature becomes immediately apparent: there is a concentration of finds with coins of the 630s and 640s around the lower Dniepr, from the Ros' to the Black Sea, and, most particularly, near the Vorskla in the region of Poltava. Most of the coins, where there is sufficient information to permit precise identification, fall into a single decade, from 636 to 646 (following Hahn's chronology). The finds are those at Zhabotin, Maistrov, Kelegeiskie khutora, Novye Sendzhary, Maloe Pereshchepino, and Maku-khovka, 2 through 7, with the cast imitations of the Constans coins from Dnepropetrovsk and the coins in the Poltava Museum, 25 and 26, and, quite possibly, the three finds at Zolotonosha, near Poltava and at Nekhvoroshcha, 27 through 29.

Coins earlier than the 630s are rare in comparison: there is just one from Pechenaya, 1, and ten from the finds at Novye Sendzhary and Maloe Pereshshepino—although these occur with later coins, as possibly may be the case at Zolotonosha. At the other end of the period, from no site in this entire region, stretching from Poltava across the Dniepr to the southern Bug and south from the Ros' to the Dniepr estuary, are there known coins later than the first years of Constans II, 642-46. It may be that the coin from Kelegeiskie khutora, 4, is later, as may be some of the unidentified coins from other finds, but the absence of any coin known to be later than 646 is striking. Such later coins as are known (and they are far fewer in number) are all from further east: at Serpovoe on the river Tsna in the north, near Kupyansk east of the Oskol, and, principally, in a band in the south from the Crimea to the northern Caucasus.

The singularity of this concentration is reinforced in other parts of the steppe by the absence of specimens from the last years of Heraclius' reign and the first years of Constans II's and by their comparative rarity in adjoining regions. To the west, the Avar finds of the middle Danube have yielded coins of Phocas, Heraclius and Heraclius Constantine, Constans II with Constantine IV, and of Constantine IV but, apparently, none of the later 630s or early 640s.62 Eastward, in the steppe, there are only later coins from near Kupyansk, Romanovskaya stanitsa, 8 and 11, and a copper one from Nizhnaya Duvanka, 32, while by the Kuban', again, there are both earlier and later coins but none from the later 630s and 640s, 17-20. In fact the gold coins of the 630s and 640s occur only in the burial in the southwestern Crimea at Pampuk-kai, 13, from near Kerch' or Taman', 16, in another burial far to the northeast in the wooded steppe at Serpovoe, 9, and in the central northern Caucasus, 22,63 where there are coins of Heraclius and his two sons (though none are known of the first years of Constans II).

The northern Caucasian coins, however, are not isolated finds but form part of a series starting in the sixth century and continuing into the eighth; indeed, coins of Phocas and the first two decades of Heraclius' reign are much more common than those of the 630s. If the imitations of and impressions made from solidi are included, this pattern is reinforced. Then, after only a couple of specimens of Constans II with Constantine IV and from the latter's reign, coins or imitations become more common with the reign of Tiberius III Apsimar, the second reign of Justinian II and that of Leo III. The fluctuations described above are clearly connected with the liveliest periods of Byzantine relations with the region, such as the time of Heraclius' Persian campaign or the dealings of Justinian II and Leo III with the Khazars and Alans.

Excluding the northern Caucasian coins, there remain only the coins from Kerch' or Taman', 16, Pampuk-kai, 13, and Serpovoe on the Tsna, 10. It is significant that the latter are from a cemetery dating from the seventh or eighth century, and that a number of objects there have parallels in finds from the seventh century in the south.64 This raises the possibility that the Heraclian coins were also brought from further south and since they were, in any case, found with coins of Constantine IV, it is likely that they were buried considerably later than the coins from the Dniepr and around Poltava.

A second feature linking most of the finds from the Dniepr-Poltava area and further differentiating them from finds elsewhere is the presence of the light weight solidi, indeed their overwhelming preponderance compared with full weight specimens. Of the coins from the period 632-46 that can be identified, 66 are of 20 carats and only 6 are full weight. Of the coins from the other regions, that of Heraclius from Pampuk-kai is full weight, as is at least one of those from the northern Caucasus.65

The precise chronological span covered by the coins from the Dniepr and near Poltava provides a key to explain their occurrence, for they coincide exactly with the period of close relations between Byzantium and the "Great Bulgaria" of Kubrat. The coins of Heraclius and his sons, all crowned, are particularly suggestive. To follow the dating adopted by Hahn, the earliest and most common of this class to be found in southern Russia are those of the year 637/8 from Maloe Pereshchepino, and this is just about the time that Kubrat staged his revolt against the Avars—according to the patriarch Nicephorus—and entered into a firm alliance with Heraclius.66 This event is not dated precisely by the chronicler, but is sandwiched between events of 634 and late 639 or 640; there is no good authority for the commonly accepted date of 635, which derives from the suggestion of de Boor.67 Nicephorus adds that Kubrat received gifts and was honored with the dignity of patrikios. Could the coins of 637/8 have been part of these gifts? Even if the letters in the field on the reverse of these coins cannot be accepted as an indictional date, this does not substantively affect the correlation. The coins must then be placed in the period from 4 July 638, when Constantine Porphyrogenitus says that Heraclonas became co-augustus, to 641.68

To continue with the coins from Maloe Pereshchepino, by far the best documented find, the second largest group is that of Constans II, of 642-46. Could these be connected with a bribe from the new emperor—a "gift" designed to forestall any further Bulgar interference in Byzantine politics and to establish good relations between the two rulers once Martina and her sons had been removed from power? Kubrat, it was rumored in Constantinople (according to John of Nikiu), had supported Martina against Constans in their contest for supreme power in Byzantium and had planned to intervene actively on her side.69 The coins from the intervening years, 639-41, could perhaps have been sent as part of a series of annual gifts to cement the relationship between Heraclius and Kubrat, or even as the salary, or roga, to which the dignity of patrikios theoretically would have entitled the latter.

As for the earlier coins, leaving aside one of Maurice and two of Phocas which could have reached the steppe some time earlier,70 the six coins of Heraclius minted between 629 and the start of 632 stand out in the context of the Pereshchepino find, being both earlier and of full weight. To pursue the suggestion advanced, could these have been brought back by Kubrat or his entourage when he returned from his childhood in Constantinople?71 Kubrat's dates and career remain controversial,72 but this passage in John of Nikiu implies strongly that his stay in Constantinople took place in the reign of Heraclius, and a return home in 631 or 632, as the coins could suggest, would fit with events in the steppe. From 630 the Turkish empire in the east began to collapse, triggered by the revolt and death of Mohotu in 631, while in the west Avar power had been greatly weakened by the disastrous outcome of the Avars' siege of Constantinople in 626.73 It could well have been in Byzantium's interest to allow Kubrat to return home to the steppe as part of an effort to turn the new developments to its advantage. The years from 632 to 637/8 (637/8, the year of the alliance between Kubrat and Heraclius that the coins would suggest) would be the period when the Bulg�r ruler was expanding his authority, as is reported by John of Nikiu, and asserting his independence. Then in 637/8, once this was achieved, he put his relationship with Heraclius on a new footing. The coins of Constans, as described, close this short series, but it is impossible to be certain whether this can be connected with Kubrat's death. The date most commonly given for his death is ca. 642, but this is not based on any decisive evidence: neither Nicephorus nor John of Nikiu is explicit, Theophanes' chronicle offers no help, and the "List of Bulgarian Princes" has been interpreted as showing that Kubrat died both ca. 642 and ca. 665.74 There are good reasons for believing that his death occurred later than 642, and it is as possible that the cessation of coin exports to the steppe was linked with a shift in Byzantine policy as much as with any event north of the Black Sea.75

One problem with the proposed correlation between this series of coins and the Great Bulgaria of Kubrat is that of geography. The obvious implication of the relevant passages in Nicephorus and Theophanes is that Great Bulgaria lay beyond the Azov and reached south to the Kuban'.76 Some scholars, using the geographical excursus in Theophanes, extend its frontiers to include the Dniepr, but his text is most confusing on this point, and recent work has cast doubt on the reliability of his geographical knowledge.77 But even these scholars believe the center of Kubrat's domain to have been east of the Azov. The fact remains, nevertheless, that there is no other people or state with which the coins may be linked, and the correlation between this series of coins and the history of Bulgar-Byzantine relations appears too neat to be coincidental.

It is possible that all the coins reached southern Russia simultaneously under Constans II,78 in circumstances such as those outlined above. Indeed, five of the coins from Maloe Pereshchepino are still joined as part of a chain—three of Constans and two of Heraclius (as MIB 3, 66-67);79 but this only proves that they were made into that item of jewelry after the arrival of the Constans coins, not that they were all sent at the same time. Maybe the coins minted in 638/9 and in 639-41, of which we have only a few specimens (though the Maistrov find, 3, tantalizingly suggests there may have been more), did come at the later date, but this seems less probable with those of 637/8 which form by far the most common coins. A further fact argues against this hypothesis, too—the rarity of the coins. The two 2 1/2 solidus coins are unique, and one of the Heraclian 20-carat coins (MIB 67) is known only from Maloe Pereshchepino.80 It is unlikely that these coins, struck in very limited numbers as only one pair of dies seems to have been used,81 would have been minted in 637/8 and then held in Constantinople until sent out in the reign of Constans; this is even less probable in the case of the two coins minted between 629 and January 632. What is likely, however, is that the issues of all these light weight solidi were small, for other die links have been observed. Of those which have been studied, in addition to MIB 67, the other Heraclian coin of 637/8 (MIB 66) has two obverse dies and four for the reverse, in four combinations, as does the Constans coin.82 This suggests that the coins were struck in limited amounts, but at frequent intervals, to meet a regular need; and the very specific grouping of these coins, in both time and place, indicates that the selection of coins was not random and raises the question of whether they were produced specifically for shipment to Kubrat and the Bulgars.

In his study of the Byzantine light weight solidi, Adelson concluded that the 20-carat denomination was introduced to facilitate Byzantine trade with the west; certainly the find spots of coins minted in the period from Justinian to the middle years of Heraclius are predominantly in southern Germany, along the Rhine to Frisia, and in eastern England. Adelson's conclusion was amplified by A. Lewis.83 Grierson, while agreeing that the coins were concentrated in Germanic areas, remains quite unconvinced by this line of argument but points out that the denomination could fit with a Germanic weight standard, three shillings of 20 grains being equivalent to a 20-carat solidus.84 These explanations hardly suit the Turkic (or Alan-Bulgar, as some would have it) context of the southern Russian steppe. It may also be significant that 20-carat coins of Justinian I have been found in two sites in this area: one being the coin possibly included in the burial at Kelegeiskie khutora, 4, the other being a find of 50 coins from Beloyarovka, near Amvrosievka, southeast of Donetsk.85 Nevertheless, between Justinian and Heraclius there remains a break, with very little Byzantine gold of any kind being found in the Russian steppe and no light weight solidi, so there are insufficient grounds for at once supposing any particular association with the steppe.

Lewis, pursuing a link between these coins and Byzantine trade policies, has suggested that the switch from west to east in the distribution of coin finds in the middle of Heraclius' reign marks an attempt by the Byzantines to tap the trade of northern Europe by a new route, after routes further west had been closed, and to divert trade from the Sasanian or Islamic world toward the Byzantine ports on the Black Sea.86 Unfortunately, there is no evidence to show that the Dniepr route was open at that time and the indications are that the route from the east to the Baltic did not become important, if it was open at all, until the very end of the eighth century.87 Such trade as there was between north and south probably went up the Volga, perhaps via the Caucasus first, to the Kama-Vyatka region, where Byzantine vessels and coin are found. But these are of silver, not gold, whereas relatively little silver plate comes from the steppe and, as mentioned, no silver coin; conversely, there is no gold in the north.88

Adelson had explained the change in the distribution of finds by suggesting that the later coins of Heraclius and those of Constans II might have been used for trade with Russia, but a trade in the form of exports of jewelry to the peoples of the steppe, and he wondered if the coins were not minted specifically for conversion to such ornaments.89 This explanation is equally unsatisfactory, for the crudity of the jewelry sets it apart from pieces of definitely Byzantine manufacture found alongside the coins, and the skills of the nomads were certainly more than adequate for the task of adapting the coins and mounting the cabochon stones that most probably filled the settings now on them.90 Furthermore, many of the coins were not exactly made into items of jewelry, but simply pierced. There can be little doubt that the transformation was made locally, after the coins had reached the steppe. It seems unlikely, too, that the coins were traded as bullion, as were the later Kufic coins, quite apart from the fact that such exports were theoretically forbidden.91 However, given the parallel established between these coins and the relationship of Kubrat with Heraclius (which also excludes the possibility that the coins were booty), there is no need to assume that they had any commercial purpose or were anything other than gifts from the emperor in Constantinople to the ruler of Great Bulgaria.

This position is reminiscent of the theory that linked the issues of light weight solidi with the tribute paid by the Byzantines to the Avars. This theory is not tenable as it stands because the first coins antedate the Avars' arrival in Europe by some fifteen or twenty years, and any link with the Avars is made more doubtful by the fact that only one find is known from within their homeland, as has been detailed by Adelson.92 Despite this, the idea that light weight solidi should be linked with payments to "the barbarians'' remains current, especially in the case of the 20-carat denomination. Grierson, for example, believes this may have been their purpose, and Hendy would certainly consider it one of the uses for these coins.93 If payments were made to the barbarians in this light weight coin, it would be most improbable that the Avars did not receive it, in which case the puzzling distribution may be explained with the idea that it was passed on in trade by the Avars.

In favor of the tribute or the payments abroad theory is the presence of Justinianic coins in southern Russia, for these could have had nothing to do with trade or any other contacts between Byzantium and the Germanic world. On the contrary, it is tempting to link them with the payments made by Justinian in the 550s to the Kutrigurs and the Utigurs, the peoples that then dominated the steppe north of the Black Sea.94 The potential savings for the treasury deriving from the substitution of light weight coinage for that of full weight are evident, but only if such payments were calculated in numbers of coins and not by weight. In at least one later case this was indeed how they were calculated, for the amount that Heraclius promised to pay the Avars in 623 is expressed in terms of solidi—200,000 of them in fact.95 Yet it is hard to imagine that simple fraud was the motive for any such use of light weight coins, for even barbarians would have seen through such attempts at deception. As has been often noted, the light weight coins are clearly distinguished from standard solidi, and none of the barbarians were so isolated from Byzantium as to have been for long unaware of the difference. So this would not provide a very satisfactory explanation of the origins of the 20-carat coinage, let alone the other denominations of 22 and 23 carats.

The alternative suggestions for the origins of these coins are that they were minted for internal purposes (for instance as an economy measure to reduce the empire's wage bill)96 or as part of a broader plan to augment its revenues by fiscal measures, such as proposed by Hendy. He compares the sixth- and seventh-century coins with the tetartera of the tenth century and other light weight coinages, while allowing that they may also have been used to make payments abroad.97 But these hypotheses do not really explain why three denominations of light weight solidi were used and continued over such a long period nor, most particularly, do they account for the great irregularity with which the 23 and 20-carat types seem to have been minted in the seventh century. For the 23-carat coins we know of issues covering the period 603-35, none from the later years of Heraclius, three issues of Constans II of 648/9, then four very large issues from 652-54. After these there is a single issue of Constantine IV, of ca. 674-80, and the series ends under Justinian II with a somewhat larger issue from the years 685-87.98 For the 20-carat coins we have specimens covering most of the reign of Phocas and of Heraclius up to ca. 625. However, one of the issues for the period ca. 616-ca. 625 (perhaps the later of the two, since the mint mark BOXX is continued with the coins of the later 630s) is exceptionally abundant (MIB 65).99 After that there is a gap until the coins of ca. 637-ca. 646 of Heraclius and Constans II, which form the main subject of the present article. Last are a few issues of Constans II from ca. 662-ca. 667 and of Constantine IV from the periods 669-ca. 674 and ca. 674-80.100

In contrast to the irregularity of these seventh century issues, those of the sixth century, from the start of the coinages (Justinian I for the 20-carat coin, Maurice for the 23-carat) up to the first half of Heraclius' reign, seem to make a continuous series, although one 20-carat issue of Justinian I (MIB 15, of 542-62) and the 23-carat coins of Maurice (MIB 11, of 583-602) are noticeably more common than the others.101 The same is true for the 22-carat coins, issued from ca. 538-641 (there are two specimens from the few days in 578 when Justin II and Tiberius II reigned jointly); of these it is the coins of Justin II from 567-78 (MIB 8) that are the most common.102 It should also be noted that the 22-carat series is the only one minted outside Constantinople—there are regular issues from Ravenna and, later, Rome—with the possible exception of a 20-carat Alexandrian issue of Justin II.103 The proposal put forward by Hahn, that the light weight solidi were introduced to facilitate the equation of gold and copper coins at the times when the latter were being revalued,104 could meet this objection of irregularity, but the correlation does not always appear wholly convincing.

An internal, fiscal reason for these coins would provide a plausible explanation for their origin, but the problem with such a theory has been the extreme paucity of finds within the frontiers of the empire as compared with the numbers outside. Kent's view that such inferior coins would not have been retained, but would have left the empire, is surely insufficient; they may have been different from coins of full weight, and perhaps not too acceptable as currency, but they were still gold. Hendy comments that the distribution could be misleading because, if the weight of the light weight solidi did fit with the standards used by the "barbarians," those coins could have gravitated toward them, thus leaving only full weight coins within the empire.105 This could be an explanation, but begs the question of whether any such correlation was pure coincidence.

Some such sorting mechanism could have been at work among the coins from the Nikertai hoard of ca. 680 from Syria. Morrisson has shown that the average weight of the solidi is close to 23 carats, yet all except four were full weight coins, and the four exceptions are of 23-carat denomination.106 Their weight compares with that of the Islamic dinar, a coin of some 22.5 Byzantine carats, which began to be minted a few years later, and which was based on the standard of the Syrian or Arabian carat, a little heavier than that of Byzantium: one dinar equalled 20 of these Syrian carats.107 The coins from Nikertai, including the light weight solidi, could then have been selected to fit with the local standard. Morrisson would also consider the coins to be representative of those then in circulation, showing that the Arab conquest of Syria did not curtail the arrival of new coin. A contrary impression, however, is given by the Arab-Byzantine gold coinage, which copies much earlier types of Phocas and Heraclius, in other words of the period prior to the Arab conquest, and the standard of which approximates that of the normal Byzantine solidus, implying that those were the coins that remained current in the region.108 If that is so, the Nikertai coins must represent a stock specifically imported from Byzantium.

Certain peculiarities in the distribution of the light weight solidi are hard to explain by means of a sorting mechanism as suggested by Hendy, and, in addition, more coins are now known from within the frontiers of the empire. The following list does not pretend to be complete, but may serve to indicate the pattern of distribution. To continue with the 23-carat series, coins of Maurice (MIB 11) are known from Bulgaria (Sadovetz),109 the Caucasus, and Syria,110 and of Phocas (MIB 21) from Israel.111 Of Heraclius class 1 (MIB 54) there are three from near Ni�, in Yugoslavia, of class 2a (MIB 55) another from north of Ni� and probably one from the northern Caucasus, and of class 3 (MIB 56) one from the Nikertai hoard)112 For the reign of Constans II, five coins of class 3 (MIB 44-47) come from Syria, three being from Nikertai.113 In addition a coin of Maurice and several of Phocas, Constans II, and Justinian II are in the collection of the Archaeological Museum in Istanbul, so may be of "Byzantine" origin.114

No certain finds are recorded for the 22-carat coins of Justinian I, although there may be one from Cherson,115 but for those of Justin II there are many: as MIB 7, one from Syria;116 as MIB 8, nine from Syria, one from Egypt, one from Sadovets in Bulgaria, and another from somewhere in the Balkans.117 In addition, two of the three specimens as MIB 6 are in Romanian collections, as is one like MIB 22, from the Ravenna mint, and one as MIB 8 is in Vienna.118 The two 22-carat coins from the brief joint reign of Justin II and Tiberius in 578 also come from the east, one from Cyprus, the other from a burial in the northern Caucasus.119 In addition there is one coin of Tiberius II (MIB 5) from Egypt, one of Maurice (MIB 13b) from Syria, and one of Heraclius class 2a (MIB 59) from North Africa.120 A few more specimens of the coins of Tiberius and Maurice are in collections in central or eastern Europe, as are two of Phocas (MIB 18-19), while three more of the latter are in Istanbul.121

For the 20-carat series, from the reign of Justinian I there are finds from Scandinavia, northern Italy, and those from southern Russia mentioned above.122 Three coins as MIB 14 are in Belgrade, while the sole specimen MIB 11 is in the Hermitage, from the Tolstoi collection, as are another nine like MIB 15. Of the latter, two were bought in St. Petersburg, two in Odessa, and one in the Caucasus.123 For the reign of Justin II all the known finds are in Syria: one as MIB 9a, three as MIB 9c, and one as MIB 15 attributed to Alexandria; however, most of Justin's coins are in collections in central or eastern Europe.124 The only known provenance of a coin of Tiberius II (MIB 6) is also from Syria, 125 as is one of Maurice (MIB 14), though two examples of the latter are in Budapest.126 For Phocas only one provenance is known, Israel (see above, n. 111) for each of his first two issues (MIB 20 and 21) although there are several specimens in the Hermitage and Budapest collections, and two of the second in Istanbul.127 For the first part of Heraclius' reign, up to about 625, three of the four issues have known provenances. A specimen of the second (MIB 63) was found in Germany, as was one of the third (MIB 64), with two more from collections in Budapest.128 The fourth issue (MIB 65) is by far the most abundant and occurs in finds from Hungary to eastern England, the greatest concentration being in southern Germany, while more specimens are known from collections in central and northern Europe.129 The distribution of the later issues of Heraclius and of Constans II and Constantine IV is discussed below.

While it is certainly premature to draw definite conclusions from this summary list, it is worth pointing out that certain patterns seem to be indicated. There are finds from within the empire, especially of the 22-carat coins–in Syria, Egypt, North Africa, the northern Balkans, and perhaps Cherson, though none as yet from Greece or Turkey (the coins from the Istanbul collection do no more than indicate that possibility). Most of the provenances of the 23- and 22-carat denominations, however, are in the east, especially in Syria, and for 23-carat ones, perhaps in the Caucasus as well; there are fewer from the Balkans or elsewhere, though the presence of such coins in some European collections suggests there may have been more. The distribution of the 20-carat coins, on the other hand, is rather different: coins from Justin II to Phocas are in Syria, like those of the other denominations; but those of Justinian I and of Heraclius are all to the north and west of Byzantium, and this pattern is reinforced by the high proportion of other examples in collections from the same regions.

The finds from within the empire do lend some support to those who propose an internal explanation for the light weight coins, but it should be noted that the distribution is far from uniform. Many of the finds are from the frontiers or from zones of disputed territory, places where one might indeed expect to find coin hoards; but this also means that they cannot for certain be claimed as "Byzantine"—for the 23-carat coins of Heraclius from Yugoslavia.130 Further, the finds in Syria appear to form a series that continues after the Arab conquest, the 22-carat (and 20-carat) coins of the period of Justin II to Maurice being replaced by 23-carat ones from the reign of Maurice to that of Constans II. Given the proportion of known finds of these types from Syria, it is tempting to agree with the suggestion that they had some particular connection with that region.131 But if that is so, the fact that Italian mints issued 22-carat coin demands some further explanation. It seems most improbable, however, that any such mechanism can account for the very localized distribution of the series of coins that are the primary focus of this study.

To return to the 20-carat coins of the 630s and 640s, the evidence presented above indicates that they were used as gifts from Heraclius and Constans to Kubrat. This begs the question of whether this was not also their purpose. The coins are rare and their distribution specific—the only finds are from the lower Dniepr and around Poltava, and only four other specimens are known, three from the Hermitage in Leningrad, one from Berlin.132 The variety of issues within such a short space of time (even if the indictional datings are not accepted) and the presence of close die links within the three issues studied (but not, it seems, between them) suggest that there was a regular series of issues, but that each was only minted in limited quantities. If, therefore, the purpose of these coins was to fulfill some internal function, whether of a fiscal nature or anything else, and new issues were regularly needed for that purpose, why were they sent abroad so soon after being minted ? The obvious answer is that they were minted precisely for sending abroad. The appearance of the full weight coins in southern Russia would then be anomalous, but the present ratio of the one denomination to the other is so much in favor of the 20-carat coins that those of full weight may perhaps be considered adventitious, especially if the earlier coins of Maurice, Phocas, and of Heraclius with Heraclius Constantine are excluded.

Although the distribution of the coins of the 630s and 640s suggests that they were gifts, it does not necessarily follow that this was the purpose of all earlier issues of 20-carat solidi. The hoard from Beloyarovka, and the quantity of other Justinianic coins in collections in central and eastern Europe do, at least, suggest the possibility that the denomination may have been the traditional one for making payments to the barbarians of the north—to the Kutrigurs and Utigurs, to their successors the Avars,133 and then to the Bulgars. Could the first issues have been provoked by the great Bulgar (Kutrigur?) invasion of the Balkans in 540? Admittedly, this offers no reason for the choice of the denomination (unless some may be found in the Hunnic or Turkic world), and if the apparent link with a Germanic standard is real and the first issues precede the Bulgar invasion, perhaps they could be linked with payments to the Germanic peoples of the middle Danube, such as the Lombards. Although there is no reason why the coins could not have been minted for some internal purpose and then used at once to make payments abroad, it is certainly suggestive that large numbers of these coins should have left the empire so soon after being produced.

For the period from Justin II to Maurice, the validity of this theory appears more doubtful, for the few finds are all in the eastern part of the empire, in Syria, and the issues are more poorly represented in the collections of Russia and Central Europe. Also a 20-carat coin seems to have been minted at Alexandria under Justin II. Phocas' coins, too, are found in the east, but more occur in European collections. For the reign of Heraclius (up to ca. 625), however, the finds are again all in Europe. Furthermore, given that the fourth issue of Heraclius from this period (MIB 65) is so very much more common than those of the preceding emperors, it is remarkable that no finds are known in Syria, in fact from anywhere except the barbarian parts of Europe; the same applies also to the most abundant issue of Justinian I (MIB 15).

There is also some correspondence between the dates when the early seventh-century coins were issued—in this period, fortunately, they may be more precisely dated than most of the issues of the sixth century—and the changing fortunes of the conflict between Byzantium and the Avars.134 In 604 Phocas made peace with them, and thereafter paid them tribute in gold; the first known issue of his 20-carat solidi dates most probably to 604, the other two to 610. Heraclius continued this policy, and the first two issues of his coins cover the period 610-ca. 616. But in 623, after his narrow escape from the Avar ambush, he was forced to buy peace with a vastly increated tribute of 200,000 solidi in order to continue with his Persian campaign.135 With the outbreak of hostilities before the great Avar siege of Constantinople in 626, the payments would have been suspended, never to be resumed. This coincides with the coins of Heraclius' class 2b, of ca. 616-ca. 625, after which no coins are known until the issues of 637/8 and later. The first issue of this class (MIB 64) is quite rare; the second (MIB 65) is more than seven times as common, and was struck in eight officinae whereas both earlier and later issues of this denominations are from only a few—usually only the fifth, ε.136 Does this difference reflect the emergency administrative measures taken to meet the sudden demand for the 200,000 solidi? This type also introduces a change in the mint mark, from OBXX to BOXX, but it is not clear what significance this may have: the mark BOXX was retained on the coins of 637/8 and 638/9, but for those of 639-41 and for the first issue of Constans II OBXX was brought back, only to be replaced again by BOXX on the later coins of Constans and those of Constantine IV.

In conclusion, then, the current distribution of the 20-carat solidi of Justinian I and of the period from Phocas to Constans II and the correlation of the issues with historical events in the relations between the Byzantine empire and the peoples of the Russian steppe and of Pannonia point unmistakably to the fact that the coins were minted in order to be sent as tribute, bribes, or gifts to buy peace from and ensure good relations with the barbarians on the empire's northern frontier. Whether this was their original purpose it is impossible to say for certain, and this use does not necessarily exclude other ones—other uses must, indeed, be invoked to account for the distribution of the coins of the second part of the sixth century and their presence in the empire's eastern provinces.

Nothing definitive may be said concerning the 20-carat coins issued later in the reign of Constans II or during that of Constantine IV due to their extreme rarity: currently only two or three specimens are known of the Constans issue of ca. 662-ca. 667,137 and only one each for the issues of Constantine IV of mid-669 to ca. 674, and between then and 681.138 But given the fact that only one of the coins is not in a Russian collection (a coin of Constans II now in Dumbarton Oaks) and that another of the Constans coins seems to have been found at Chemburko, near Anapa on the coast of the Black Sea, it may be that these issues had a function similar to those of the 630s and 640s.

Unless some significant finds of these later coins are made, it is unlikely that much more may be inferred about their uses or purpose, but it may be expected that new discoveries and further study of the material in the collections of the Soviet Union, and elsewhere, will reveal new specimens of other, different issues of light weight solidi that will enable firmer conclusions to be drawn. In particular, one hopes that such work may confirm (or modify) the validity of the proposals made here for the 20-carat solidi of the 630s and 640s. If their association with Kubrat and the Bulgars is substantiated, it will provide a key to the interpretation of the other objects from Maloe Pereshchepino, and from related sites, which are of crucial importance for our understanding of the archaeology and the history of the European steppe.139

Addenda

5. Novye Sendzhary . Six of the seven coins are published, see A. T. Smilenko, "Nakhodka 1928 g. u. g. Novye Sendzhary (po materialam obsledovania A. K.Takhtaya). Slavyane i Rus' 1 (Moscow, 1968), pp. 158-66. The material from the find was destroyed in the Second World War, but photographs of a number of items survive, and it is these that Smilenko published. The coins are on pp. 162-63 and fig. 3. Fig. 3, 5, is the Phocas coin. The Heraclius ones were indeed all full weight: fig. 3, 3, is like MIB 41 (635/6); fig. 3, 2, is like MIB 48 (638/9); fig. 3, 4, is like MIB 50 (639-41); fig. 3, 1, is like MIB 53, attributed by Hahn to Heraclius II (641). The sixth coin, fig. 3, 6, was a 20-carat solidus of Constans II (MIB 48), officina θ; the seventh coin, of which no photograph survives, was presumably identical to this. The coins were in good condition.

6. Maloe Pereshchepino . An important study on this find has appeared since this article was completed: J. Werner, Der Grabfund von Malaja Pare�čepina und Kuvrat, Kagan der Bulgaren, Bayerische Akademie der Wissenschaften, Phil.-hist. Klasse, Abhandlungen, N.F., 91 (Munich, 1984). Werner deals with the coins on pp. 17-18 and 35-36, treating them as part of a chain made in imitation of Byzantine "ceremonial" or "marriage" chains. He notes (thanks to M. Mackensen) that the Constans II coins are of 20 carats and, from their date, would place the chain ca. 650, as the whole find itself, but does not otherwise discuss the coins or their significance, focusing rather on the plate and jewelry. He identifies (esp. pp. 38-45) the find as the grave of Kubrat himself, making explicit what is only implicit in the present article. For different arguments, building on those expressed here, but independently reaching the same conclusion as Werner, see my Ph. D. thesis, "Byzantium, the Crimea and the Steppe, c. 550-c. 750," the University of Birmingham, 1985, esp. chapters 3, 1, and 5, 2. I would, however, date Kubrat's death to ca. 660 rather than ca. 650 on historical grounds and suggest that the later coins of Constans II and those of Constantine IV may have been sent to his successors when they took power.

Since this article was completed, I have noted two more finds for the sake of completeness.

Near Volgograd: by the Volga; a nomadic burial containing a chain with one solidus of Phocas and one of Heraclius and Heraclius Constantine of 613-31 (perhaps to be associated with the Caucasian series), see K. S. Lagotskii and V. P. Shilov, in Arkheologicheskie Otkrytiya 1976 (1977), p. 159.

Kepoi: in the Taman' Peninsula; a chance find in the antique town of a hexagram of Heraclius (N. A. Frolova and D. B. Shelov, "Monety iz raskopok Kep, 1958-1963 gg.," Numizmatika i sfragistika 2 [1965], p. 192).

End Notes

1
J. P. C. Kent, review of Adelson, NC 1959, pp. 237-39; P. Grierson, DOC 2, pp. 14-15, and Coins, pp. 53 and 100; M. Hendy, "Light Weight Solidi, Tetartera and the Book of the Prefect," Byzantinische Zeitschrift 65 (1972), pp. 57-82, esp. 57-63 and 74-79; W. Hahn, MIB 1, pp. 26-27 and 49; 2, pp. 15-17; 3, pp. 16-17, 63-65, 88, and 126.
2
Notably V. V. Kropotkin, Klady vizantiiskikh monet na territorii SSSR, Arkheologiya SSSK, Svod E4-4 (Moscow, 1962; hereafter Kropotkin), pp. 9-10 and 14-15, and Ekonomicheskie svyazi Vostochnoi Evropy v I tysyachletii nashei ery (Moscow, 1967), esp. pp. 47-51, 52-53 and 116-17.
3
Kropotkin, pp. 21-51.
4
Kropotkin, p. 33, 196, comparing Sabatier 1, pl. 29, 18; MIB 21.
5
Kropotkin, p. 37, 274, as Adelson 161 and MIB 69. Hahn's Pr�getabelle 1 does not show the change in mint mark, from the BOXX of coins 65-68 to OBXX. The fact that it is the mark OBXX that continues into the reign of Constans II, not BOXX, suggests that this was indeed the last of the series, supporting the idea that the letters in the right field on the reverse are indictional dates; see below and n. 68.
6
Kropotkin, p. 32, 159; see MIB 50. Kropotkin gives a full description, yet mentions no letters corresponding to an indictional date (so not coins 40-49 of 635-39) and places the monogram to the left of the cross on the reverse (so not coins 39 or 51).
7
Kropotkin, p. 37, 268; N. Bauer, "Zur byzantinische Münzkunde des VII. Jahrhunderts," Frankfurter Münzzeitung (March, 1931), p. 228; Adelson, pp. 62 and 93 (not OBXX as on p. 18; nor should the Constans coin be listed with the mark BOXX as on p. 81). The BOXX coins of Heraclius would fit with MIB 66 or 68.
8
Kropotkin doubts that this coin came from the burial, and Bauer certainly does not mention it. The mint mark OBXX indicates it weighs 20 carats, yet the weight quoted by Kropotkin is 5.10 g.
9
Kropotkin, p. 36, 249; Bauer (above, n. 7), p. 228. Of the officina marks on the Heraclian coins quoted by Kropotkin – Г, Δ, ϵ, I only ϵ is known on contemporary light weight solidi, see MIB, p. 217. The Constans coins are as MIB 48. See Addenda.
10
Kropotkin, p. 36, 250, pis. 15 and 16; Bauer (above n. 7), pp. 227-28; Adelson, pp. 62 and 93-95. Kropotkin's totals, however, are confused: by reign he lists 68 coins (18 for Constans II, not 19); by mint mark he lists 70 (11 CONOB, not 10); he also lists eight coins with the non-existent mark BOOX, presumably in place of BOXX +.
11
The coins, pl. 15, 23, and pl. 16, 6, are among the most obscured, but the identification appears reason ably certain.
12
Kropotkin, pl. 15, 1-3. The caption identifies two coins of Maurice and one of Phocas, but the text has the reverse (as also Bauer [above, n. 7]). The illustrations are illegible.
13
Kropotkin, pl. 15, 4-9; MIB 29-37b. To judge from what is visible, Kropotkin 4 and 9 may correspond to MIB 29; 8 to 32; and 6 and 7 to 37a.
14
Bauer (above, n. 7), p. 228; Grierson, DOC, pp. 9-10, and Coins, p. 92; Hahn, MIB, p. 85. As Hahn says, the coins are presumably those in Kropotkin, pl. 15, 4 and 9, which are clearly distinguished from the others by the size of their flan. Their date of issue may be connected with the successful conclusion of Heraclius' Persian campaign.
15
Kropotkin, pl. 15: 10, 11, 13, 15-18, 22-26, 28; and pl. 16, 1-9, 11-13, 18; see Adelson 162 and 163; also, MIB 66.
16
Kropotkin, pl. 15: 12, 14, 21, 27, 29; pl. 16: 10, 16, 24 (the last being incorrectly assigned to Constans II in the caption); see Adelson 165; MIB 67.
17
Not illustrated, but Kropotkin's description permits the identification: the mint mark BOXX and the three emperors indicate the coins are the same as MIB 66 or 68 (68 = Adelson 164), as the indictional date is not mentioned (even perhaps 67, because Kropotkin seems not to acknowledge the extra cross, BOXX +, on the other coins of this type, see n. 10 above).
18
Again not illustrated, but the three emperors and the mint mark OBXX point to MIB 69 and Adelson 160-61.
19
Kropotkin, pl. 16: 14, 15, 17, 19-23, 25-32; MIB 48, and Adelson 166-68 (though all the coins are of officina 10, as Adelson 168, not officina 9, as 166-67); plus three coins not illustrated, but the description, with Constans II alone on the obverse and OBXX in the exergue, permits the identification. None of the coins have OBXX+ as suggested by Grierson, Coins, p. 100; nor are any known with BOXX, as Adelson, p. 81, or DOC, pp. 13-14 and 423.
20
Kropotkin, pl. 15, 19; MIB 44-53. Closer identification is not possible from the plate as the coin was made into a cuff link, part of which hides the right field. Sec Addenda.
21
V. V. Kropotkin, "Novye nakhodki vizantiiskikh monet na territorii SSSR," Vizantiiskii Vremennik 26 (1965), p. 178, 73.
22
Kropotkin, p. 37, 265, referring to BMC 1, pl. 30, 17; cf. MIB, pl. 20, 26.
23
Kropotkin, p. 29, 125. I have not been able to check the Delo Arkheologicheskoi Kommissii 94 (1892), pp. 75-125, which may have further details.
24
Kropotkin, p. 29, 126; pl. 17, 11-12; MIB 7 or 8 (the waving ribbons are diagnostic); BMC 2, pp. 314-15, 2 and 14, pl. 36, 2.
25
Kropotkin, p. 27, 84, pl. 18, 4 (the Leontius coin); Kropotkin (above, n. 21), p. 170, 17, quoting BMC 2, p. 315, 14, for the Constantine one (MIB 10), and BMC 2, p. 365, 2, for that of Leontius (MIB 1).
26
I. V. Sokolova, "Nakhodki vizantiiskikh monet VI-XII vv. v Krymu," Vizantiiskii Vremennik 29 (1968), p. 256.
27
Kropotkin (above, n. 21), p. 176, 61; A. L. Yakobson, Rannesrednevekovye poseleniya Yugo-zapadnoi Tavriki, Materialy i issledovaniya po arkheologii SSSR 168 (Leningrad, 1970), p. 147, fig. 97; MIB 50.
28
Kropotkin (above, n. 21), p. 177, 66, 4.32 g; comparing BMC 2, pp. 313-H, pl. 36, 2-3; MIB 7a. See also I. N. Osinovskii, "Nakhodka solida Konstantina IV bliz Sadaka," Sovetskaya Arkheologiya 1966, 3, p. 240.
29
Kropotkin, p. 34, 207, 4.39 g, comparing Tolstoi, p. 798, 8 variant, and BMC 2, p. 314, 4; see MIB 4C.
30
Kropotkin, p. 47, 487.
31
Kropotkin, p. 23, 33, comparing Tolstoi, pp. 777-81, 285-317. The presence of two figures on each side and the mint mark BOXX point clearly to MIB 49, and Adelson 176-77 (49 = 176).
32
Kropotkin, p. 22, 26; Kropotkin (above, n. 21), p. 168, 7; K. V. Golenko, "Klad vizantiiskikh monet VII v. naidennyi bliz Anapa," Vizantiiskii Vremennik 26 (1965), pp. 162-65 and plate. The three Constans coins (Golenko nos. 1-3) are like MIB 31; the solidi of Constantine IV (Golenko nos. 4-5) arc like MIB 4C; the hexagrams (Golenko nos. 6-7) are like MIB 63C (Kropotkin [above, n. 21), compares Tolstoi, pp. 806-7, 53, obviously in error for no. 54; see also the obverse legend given in Kropotkin, p. 22, 26).
33
Kropotkin, pp. 21-22, 12, 4.41 g.
34
Kropotkin, p. 22, 25, comparing BMC 2, p. 163, 10-11, and Tolstoi, p. 584, 8, and pl. 42, 9; MIB 9, 3.59 g.
35
Kropotkin, p. 23, 35, pl. 20, 15, 4.25 g; MIB 8A.
36
Kropotkin, pp. 30-31: 133, 135-39, 141a, 142; Kropotkin (above, n. 21), p. 171, 24 and 25; E. V. Rtveladze and A. P. Hunich, "Nakhodki indikatsii vizantiiskikh monet vblizi Kislovodska," Vizantiiskii Vremennik 32 (1971), pp. 219-22, and "Novye nakhodki vizantiiskikh monet i indikatsii v okrestnosli Kislovodska," Vizantiiskii Vremennik 37 (1976), pp. 151-55. There are also a number of unspecified contemporary coins, including some copper, and a silver coin each of Phocas and Constans II.
37
Kropotkin, p. 30, 136 (found with a dirham of 700/1), compared with BMC 1, p. 187, 26; see MIB 55, of 613-ca. 616.
38
Kropotkin, p. 30, 131; Kropotkin (above, n. 21), p. 171, 21.
39
Kropotkin, p. 180, 97, compare with Tolstoi, pp. 670-71, 226-30; MIB 160b.
40
Kropotkin, p. 31, 149; K. V. Golenko, "Imitatsii solida VII v. iz Podneprov' ya," Vizantiiskii Vremennik 11 (1956), pp. 292-94 and plate. All are of officina 10, like the coins from Maloe Pereshchcpino, see n. 19 above.
41
Adelson 162 and 163; MIB 66.
42
Adelson 166 and 167; MIB 48.
43
Adelson, pp. 185-86, nn. 112 and 116.
44
Kropotkin, p. 36, 250; pl. 16: 14, 15, 17, 19-23, 25-32.
45
Kropotkin, p. 36, 249. See Addenda: the museum coins clearly are not from Novye Sendzhary.
46
Kropotkin: no. 27, p. 37, 276; no. 28, p. 36, 252; no. 29, p. 36, 251; no. 30, p. 32, 161; no. 31, p. 37, 267.
47
Kropotkin, pp. 9-15, maps 3-12. Hoards or finds of the later gold coins occur from the Dniepr estuary north to Kiev and beyond, Kropotkin, pp. 13-14; p. 31, 154 and 158; p. 32, 176; p. 37, 270 and 281.
48
Kropotkin, p. 35, 233. It is said to be of 668-69, but MIB 3 recognizes no such classification, see Pr�getabelle 9.
49
K. V. Golenko, "Monety iz raskopok GIMa v Khersonesa 1958-1959 gg.," Sovietskaya Arkheologiya 1974, 4, pp. 214, 217 and 220; Sokolova (above, n. 26), p. 259. On the Cherson mint and the Bosporan countermarks almost certainly Chersonite too see W. Hahn, "The Numismatic History of Cherson in Early Byzantine Times – a Survey," N Circ. 86 (1978), pp. 251-23, and MIB, pp. 89, 121-22, 126, and 147-48; I. V. Sokolova, Monety i pechaty vizantiiskogo Khersona (Leningrad, 1983), pp. 26-29.
50
Kropotkin, pp. 33-34, 201, 204, 208, and "Novye nakhodki vizantiiskikh monet v Kerchi," Vizantiiskii Vremennik 32 (1971), pp. 217-18.
51
Kropotkin, p. 22, 19a.
52
Kropotkin, p. 35, 226; compare with Tolstoi, pl. 48, 223; MIB 129.
53
Kropotkin, pp. 9, 42, 44, and 50, nos. 366, 366a, and 405 for finds south of the Caucasus; Ekonomicheskie suyazi (above, n. 2), pp. 50-51 and 117.
54
B. Mitrea, "Date noi cu privire la secolul al vii-lea. Tezaurul de hexagrame bizantine de la Priseaca," Studi şi cercetari de numismatica 6 (1975), pp. 118-20.
55
Kropotkin, pp. 10, 15, 17, and 41-44, nos. 360, 364, 367, 370, 373, etc., and Ekonomicheskie suyazi (above, n. 2), pp. 50, 116-17.
56
Mitrea (above, n. 54), pp. 113-25; P. Radomĕrskϵ, "Byzantsk� mince z pokladu v Zemiansk�m Vrbovku," Pam�tky Archeologick� 44 (1953), pp. 109-22.
57
See below, nn. 86, 87.
58
D. Csall�ny, "Vizantiiskie monety v avarskikh nakhodkakh," Acta Archeologica Hungarica 2 (1952), pp. 235-44; L. Husz�r, "Das Münzmatcrial in den Funden der V�lkerwanderungzeit im mittleren Donaubecken," Acta Archaeologica Academiae Scientiarum Hungaricae 5 (1955), pp. 61-109, pis. 13-33.
59
There is no evidence from this period that substantiates H. N. Frye's assertion ("Sasanian Silver and History," Iran and Islam. In Memory of V. Minorsky, ed. C. E. Bosworth [Edinburgh, 1971], pp. 257 and 259) that silver was "par excellence" the coin of the steppe. See also Kropotkin, Ekonomicheskie svyazi (above, n. 2), pp.47 and 116.
60
Kropotkin, p. 21, n. 2, pp. 43-46, nos. 396, 402, 408, 435, 436, 448, 450, 475, 476, to mention just the seventh-century coins.
61
T. Ya. Abramishvili, "Nokalakevskii klad," Vizanliiskii Vremennik 23 (1963), pp. 158-65, pls. 1-2. An other 23-carat coin of Maurice was bought by Tolstoi in the Caucasus: Tolstoi, pp. 514-15, 30; MIB 11. Abramishvili also gives a list of coin finds in Georgia (pp. 163-65), but adds none to those in Kropotkin.
62
Huz�r (above, n. 58), nos. 110, 113, 119, 135, 140, 167, 186, 192, 204, and 236.
63
Kropotkin, p. 30, 138 (Kamunta), and (above, n. 21), pp. 171-72, 25. These two coins are both dated 638-41, so presumably all the figures are crowned and, according to MIB, of 636-41; Rtveladze and Runich (above, n. 36), p. 220, 1 (Dzhaga).
64
A. K. Alikhova, "Nekotorye khronologieheskie i plemennye otlichiya v kul'ture Mordvy kontsa I i nachala II tysyachletiya n.e.," Sovietskaya Arkheologiya 1958, 2, pp. 66-77.
65
Kropotkin (above, n. 21), pp. 171-72, 25, compare with Tolstoi, p. 702, 377 = BMC 55, see MIB 45.
66
Nicephorus the Patriarch, "Historia Syntomos," Opuscula historica, ed. C. de Boor (Leipzig, 1880), 24.9-15.
67
A point well made by I. S. Chichurov, Vizantiiskie istoricheskie sochineniya (Moscow, 1980), p. 174, n. 59. For 635 see, for instance, S. Runciman, A History of the First Bulgarian Empire (London, 1930), p. 14; D. Obolensky, The Byzantine Commonwealth (London, 1974), p. 89.
68
De Caeremoniis, book 2, chapt. 27, ed. J. Reiske (Bonn, 1829), vol. 1, pp. 627-28. For a good statement of the "indictional" view, see Hahn, MIB, pp. 85-86. Grierson, however remains unsure, "Solidi of Phocas and Heraclius: The Chronological Framework," NC 1959, pp. 139, 142-43, 148-49, and 154: DOC, pp. 114 and 223-25; and most recently Coins, pp. 86, 88, and 94. See above, n. 5.
69
John of Nikiu, Chronique de Jean, �v�que de Nikiou, tr. H. Zotenberg (Paris, 1883), p. 460; tr. H. H. Charles (London, 1916), p. 197.
70
Cf. nos. 20 and 27 above; perhaps too it is no coincidence that the early coins from Maloe Pereshchepino are among the most elaborately adapted, sec Kropotkin, pl. 15.
71
John of Nikiu (above, n. 69): Zotenberg, p. 460, Charles, p. 197.
72
G. Moravscik, Byzantinoturcica (Berlin, 1958), esp. vol. 1, pp. 108 and 112-31; vol. 2, pp. 98-101 and 161–62; Chichurov (above, n. 67), esp. pp. 112-13, 168-69, and 174-75. The argumentation is too complex to present here: I would follow the line that accepts John of Nikiu and Nicephorus, rather than rely on the "List of Bulgarian Princes," and would not place the start of Kubrat's reign in ca. 584 (as e.g. Hunciman [above, n. 67], pp. 11, 13-16, and 272-78), but ca. 630. 1 hope to deal with this and related issues in another article.
73
E. Chavannes, Documents sur les Tou-kiue occidentaux (St. Petersburg, 1903), pp.3, 25-26, 53-54, and 264-68; D. Obolensky, The Byzantine Commonwealth (London, 1974), pp. 77-79, 85-86, 88-90, 222-24, and 227; A. N. Stratos, Byzantium in the 7th Century, (Amsterdam, 1968-80), vol. 1, pp. 191-95 and 315-17; vol. 2, pp. 159-64. At this time, too, after 630/1, the Bulgars in Pannonia made an attempt on the Avar throne: "Fredegarii et aliorum Chronica," ed. B. Krusch, in Monumenta Germaniae Historica, Schptores Rerum Merov. 2 (Hanover, 1888), iv. 72, p. 157.
74
Nicephorus (above, n. 66), 24.9-15, 33.17-19; John of Nikiu (above, n. 69), Zotenberg, p. 460, Charles, p. 197; Theophanes the Confessor, Chronographia, ed. C. de Boor (Leipzig, 1883-85), vol. 1, pp. 257-58, esp. pp. 257.11-16. On the "Princes' List," see for instance Bunciman (above, n. 67), esp. pp. 272-78; O. Pritsak, Die bulgarische F�rstenliste und die Sprache der Prolobulgaren (Wiesbaden, 1955), esp. pp. 35-36 and 76-77; Moravcsik (above, n. 72), vol. 2, pp. 352-54.
75
It is most improbable that any break in Byzantine-Avar relations was the cause (as Grierson, Coins, p. 100), for at this time the Avars had little to do with the central Bussian steppe; but perhaps, in the end, it may have had some connection with Byzantine-Bulgar relations, see below.
76
Nicephorus (above, n. 66), 33-35, esp. 33.13-17; Theophanes (above, n. 74), pp. 356-58.
77
Theophanes, 356.20-357.8. For discussion, see Chichurov (above, n. 67), pp. 108-11 and 175-76, and "Ekskurs Feofana o protobolgarakh," Drevneishie gosudarstua na territorii SSSR (Moscow, 1976), pp. 65-80.
78
For such a suggestion, see Grierson's, Coins, p. 100.
79
Kropotkin, pl. 16, 14-18.
80
Hahn, MIB, p. 217; also Adelson, p. 62 and no. 165.
81
Bauer (above, n. 7), p. 228; Adelson, p. 62.
82
Bauer, p. 228; DOC, p. 14; Adelson, pp. 62-63 and 92.
83
Adelson, pp. 84-91, 98, and 103; A. R. Lewis, "Byzantine Light Weight Solidi and Trade to the North Sea and Baltic," Studies in Language, Literature and Culture of the Middle Ages and Later, ed. E. B. Atwood and A. A. Hill (Austin, 1969), pp. 131-47. Neither Adelson nor Lewis knew of the hoard from Beloyarovka and, in fact, the evidence of the finds is more complex than it then seemed, sec below.
84
DOC, pp. 14-17, and Coins, p. 53; also Hendy (above, n. 1), pp. 59-61. For a good statement of the alternatives to trade, see P. Grierson, "Commerce in the Dark Ages: A Critique of the Evidence," Transactions of the Royal Historical Society 1959, pp. 123-40, rpt. in P. Grierson, Dark Age Numismatics (London, 1971), sect. 2, pp. 123-40.
85
Kropotkin, pp. 9 and 253; MIB 1, 15 (Adelson 2-19) and perhaps 16. Several of the coins listed by Adelson also come from Russia or from the Hermitage collection (3, 4, 10-12, 14-17), of which two specimens were bought by Tolstoi in Odessa (Tolstoi, p. 295, 23 and 25 = Adelson 3 and 11), one in the Caucasus (Tolstoi, p. 295, 25 [bis] = Adelson 12) and one in St. Petersburg (Tolstoi, p. 295, 24 = Adelson 4). Further, might the hoards listed in Kropotkin, p. 35, 231 and 232, to the north of Beloyarovka, with coins of Justin I and Justinian I, have contained light weight solidi as well?
86
Lewis (above, n. 83), pp. 147-51.
87
That is the date of the earliest known hoards of Kufic silver, all of which contain Sasanian and early Umayyad coin, see V. L. Yanin, Denezhno-vesouye sistemy russkogo srednevekov'ya (Moscow, 1956), esp. pp. 79-85; T. S. Noonan, "When and How Dirhams First Reached Russia," Cahiers du monde russe et sovi�tique 21 (1980), pp. 401-69; and, in general, Kropotkin, Ekonomicheskie svyazi (above, n. 2), esp. pp. 116-19.
88
See also Kropotkin, pp. 10, 17, and 26, nos. 69 and 70, maps 3 and 12; E. C. Dodd, Byzantine Silver Stamps, DOS 7 (Washington D.C., 1961), nos. 37-78; R. N. Fryc, "Byzantine and Sasanian Trade Relations with Northeastern Russia," DOP 26 (1972), pp. 265-69. The finds at Serpovoe, halfway between the Dniepr and the Kama, may be explained otherwise, see above and n. 64.
89
Adelson, p. 92.
90
A. K. Ambroz, "Kochevicheskie drevnosti Vostochnoi Kvrazii i Srednei Azii V-VIII vv.," Stepi Evrazzii v epokhu srednevekov'ya, ed. S. A. Pletneva (Moscow, 1981), pp. 11-23, pl. facing p. 48, pp. 106-11, figs. 3-7; most of the objects from Pereshchepino were published by A. A. Bobrinskii, "Pereshchepinskii klad," Materialy po arkheologii Rossii 34 (1914), pp. 111-20 and plates.
91
Corpus Juris Civilis 4, 63.2, ed. P. Krueger et al. (Berlin, 1902-15), vol. 2, p. 188. Grierson considers it most improbable that the coins reached Russia by way of trade, preferring the idea of gifts or tribute, as does Hendy, above n. 84; see also Kropotkin, pp. 10 and 14, and Ekonomicheskie svyazi (above, n. 2), pp.47 and 116.
92
Adelson, pp. 19-29, on the Avar theory propagated by Stefan, and pp. 81-90, on the distribution. The one find from Avar territory is a coin of Heraclius and Heraclius Constantine from Szentes, Adelson 132; there are, however, many other coins in the Budapest museum which may be of local provenance (eg. 35, 88, 89, 100, 107, 109, 110, 122, 127, 128, 130); see DOC, p. 15. For a fuller discussion of find spots, see below.
93
Grierson, DOC, pp. 14-15, and Coins, pp.53 and 100; Hendy (above, n. 1), pp. 61-63.
94
Agathias, Historiae 5.11-12, 24-25, ed. R. Keydell (Berlin, 1967), pp. 176-79 and 195-97; Procopius, Bella 8.18-19, ed. J. Haury (Leipzig, 1962-63), vol. 2, pp. 581-89; Menander Protector, in Excerpta de legationibus, ed. C. de Boor (Berlin, 1903), vol. 1, pp. 170-71; Obolensky (above, n. 73), pp. 67-71 and 221-24. It is just possible that the Beloyarovka hoard could be booty, but unlikely; study of the coins should help resolve the question, particularly if many die links are found, for this would clearly point to a gift.
95
Nicephorus (above, n. 66), 17.18-19; see below, n. 135.
96
J. P. C. Kent (above, n. 1), pp. 237-38.
97
Hendy (above, n. 1), pp. 61-63 and 74-79.
98
MIB 2, p. 127, 13a-17; 3, p. 217, 54-57; pp. 241-42, 43-47; p. 256, 12; pp. 262-63, 10.
99
MIB 2, p. 127, 20-22; 3, p. 217, 62-65.
100
See below, and nn. 137-38.
101
For 20 carats, MIB, 1 p. 109, 11-16; 2, p. 92, 9 and 15; p. 104, 6; p. 109, 14. For 23 carats, MIB 2, p. 109, 11; cf. MIB 3, pp. 64-65.
102
MIB 1, pp. 109-10, 9-10, 38; MIB 2, pp. 91-93, 6-8, 22-23; p. 103, 2; p. 104, 5, 16; p. 109, 12-13; p. 111, 31; p. 127, 18-19; MIB 3, p. 30, N19; p. 217, 58-61. For the second coin of Justin and Tiberius, see Rtveladze and Runich (above, n. 36), p. 151 and plate nos. 1-2.
103
On Alexandria, see MIB 2, pp. 39-40; MIB 3, p. 65. For the ascription of the 22-carat coins with a reverse legend ending in ⊖ or ⊖S to Thessalonica or Antioch (Theupolis), see Adelson, pp. 99-102; MIB 1, pp. 49-50.
104
MIB 1, pp. 26-27 and 48-49; MIB 2, pp. 15-17; MIB 3, pp. 16-17, 63-65, 88, and 126.
105
Kent (above, n. 1), pp. 237-38; Hendy (above, n. 1), pp. 60-61.
106
C�cile Morrisson, "Le trésor byzantine de Nikertai," RBN 118 (1972), pp. 54-64.
107
P. Grierson, "The Monetary Reforms of 'Abd al-Malik: Their Metrological Basis and Their Financial Repercussions," JESHO 3 (1960), pp. 247-57, rpt. in Dark Age Numismatics (above, n. 84), sect. 15, pp. 241-64; see also Coins, p. 100.
108
Morrisson (above, n. 106), pp. 34-39; Grierson (above, n. 107), pp. 241-47, and Coins, p. 147.
109
J. Youroukova, "Contribution numismatique × le d�finition du charact�re des agglom�rations du vi siècle dans les Balkans," M�langes de numismatique, d'archologie et d'histoire offerts × Jean Lafaurie (Paris, 1980), pp. 273-80 and p. 280, Trésor mon�taire C.
110
Caucasus: see above and n. 61. Syria: J. Lallemand, "Sous d'or byzantins de poids faible eu Cabinet des Médailles de Bruxelles," Dont Numismatica Walter H�vernick zum 23 Januar 1965 dargebrachl, ed. P. Berghaus and G. Hatz (Hamburg, 1965), p. 40, 6-7; the coins published by Lallemand make a very odd collection indeed, but their Syrian attribution appears certain, see p. 39.
111
From Beth Galim: Coin Hoards 1 (1975), p. 66.
112
V. Popovic�, "Les t�moins archéologiques des invasions Avaro-Slaves dans l'Illyricum byzantine," MEFRA (1975), pp. 494-96, 15-18; Caucasus: see above and n. 37; Morrison (above, n. 106), p. 74, 216.
113
Morrisson (above, n. 106), p. 84, 415-17; Lallemand (above, n. 110), p. 40, 10-11.
114
Hahn, MIB 3, pp. 28, 30, 34, 35, 241-42, and 262-63.
115
T. I. Kostromicheva, "Neizdannye zolotye monety iz Khersonesa," Sotsial'noe razvitie Vizantii (Sverdlovsk, 1979), p. 128, 3 (4.08 g), compare with Tolstoi, p. 295, 22, as MIB 10a.
116
Lallemand (above, n. 110), p. 39, 4.
117
Lallemand, p.39, 3; Adelson 58-65 (Syria) and 51-52 (Balkans); 'Abd al-Mohsen al-Khachab, "Une nouvelle trouvaille numismatiqye," Annales du Service d'Antiquits d'�gypte 55 (1958), p. 142, 18 (Markaz al-Huseinieh).
118
MIB, pp. 91-93.
119
Adelson 79 and p. 179, n. 45 = MIB 2; Rtveladze and Runich (above, n. 36), p. 151.
120
'Abd al-Mohsen al-Khachab (above, n. 117), p. 142, 20; Lallemand (above, n. 110), p. 40, 9; Adelson 129, see DOC, pp. 12 and 249, 12.1.
121
Tiberius II: Adelson 82-83, 85-86; MIB 104. Maurice: Adelson 97; MIB, pp. 109 and 111 (an example of 31, from Rome, is in Bologna, see MIB 3, p. 34). Phocas: Adelson 105 and 111; MIB 2, 127; MIB 3, 30 and 35.
122
Fagerlie, pp. 74 and 75, 740-41; Adelson 2; MIB, p. 109, 11-16.
123
MIB 3, p. 32, 14; Adelson 1 = Tolstoi, pp. 295-96, 26; see above, n. 85.
124
Lallemand (above, n. 110), p. 39, 1-2; Adelson 36-38, 43, 46, 47, 49; MIB p. 92.
125
Lallemand, p. 40, 5; MIB, p. 104.
126
Lallemand, p. 40, 8; Adelson 88-89; MIB, p. 109.
127
Coin Hoards 1 (1975), p. 66. Adelson 100-102, 107-10; MIB p. 127; MIB p. 3, p. 35, 21.
128
Adelson 117, 128, 130, and 154; MIB 217.
129
Adelson 118, 122, 124, 125, 127, 128, 130, 132, 136-44, 148-50, 153, 155; MIB 3, p. 217; J. Hraha, "Byzantsky solidus ze stŕednich Čech," Archeologick� Rozhledy 16 (1964), pp. 512-16.
130
Popovic (above, n. 112), pp. 488 ff. There are hardly sufficient grounds for his assertion (p. 495) that the coins were not part of any tribute paid to the Avars.
131
Grierson, DOC, p. 15, and Coins, p. 53, 100.
132
Tolstoi, p. 707, 422-23 = Adelson 160 and 164; Adelson 168; M. Restle, Kunst und byzantinische Münzpr�gung von Justinian I. bis zum Bilderstreit (Athens, 1964), pl. 21; MIB, p. 217, 66-69; p. 242, 48.
133
See Bayan's demand that the Avars should receive the moneys previously paid to the Utigurs and Kutrigurs, Menander (above, n. 94), vol. 1, p. 196; for the general context, see Obolensky (above, n. 73), pp. 67-72.
134
Obolensky, pp. 76-79; Stratos (above, n. 73), vol. 1, pp. 66-67, 118-21, 145-50, and 178-95.
135
Nicephorus (above, n. 66), 17-18 esp. 17.18-19; A. N. Stratos, "Le guet-apens der Avars," Jahrbuch der Österreichischen Byzantinistik 30 (1981), pp. 113-35.
136
Hahn, MIB, p. 217; Morrisson (above, n. 106), pp. 47-53, provides a listing of the officina marks used on Byzantine solidi in the sixth and seventh centuries.
137
Adelson 176 (= DOC, p. 433, 39a = MIB, p. 242, 49) and 177; plus, most probably, the coin in 17, Chemburko khutor (n. 31).
138
Adelson 178-79 = Tostoi, p. 800, 15, and p. 802, 32 = MIB, p. 256, 13-14.
139
I thank Michael Metcalf and Wolfgang Hahn for their help on various points, especially for references I otherwise would have missed.

MICROCHEMICAL ANALYSIS OF THE METAL CONTENT OF SOME EIGHTH-CENTURY COINS OF Rome AND Ravenna

Wolfgang Hahn

The method used here is a form of wet chemical analysis, the dithiazone method, which has been applied elsewhere several times,1 but Has also been questioned2 because its results frequently conflict with results obtained using other physical methods. Although no fallacies in the chemical methods have been found yet, a project to investigate a larger series of Byzantine gold coins struck in western mints has been suspended for the present. Nevertheless, the results of the analyses already done need not be suppressed. They may either be confirmed or modified by future research.

The following list contains 20 eighth-century gold coins from Rome and four of Ravenna, with an addendum of three silver pieces. The analyses have been done by R. Mauterer at the Institute of Inorganic Chemistry of Vienna University. As the mint attributions for the gold coins are certain and the datings in most cases given by indictional years in the legends of the coins, the irregularity of the purity of the Roman manufacture is most strikingly demonstrated. Repeated efforts to halt the debasement of the gold coins seem to have failed, presumably because of a growing shortage of that precious metal.

The three silver coins from Rome have been added for comparative purposes. Their dating is less certain because the monograms on the reverse have been variously interpreted.3 All three appear to be papal strikings. One, DOC Constantine V 44, should be attributed to Pope Constantine (708-15), and another (BMC 66) is also of Pope Constantine. The third, DOC Leo III 92, should be attributed to Pope Gregory II (715-31) or Gregory III (731-41).

Despite questions about the reliability of the dithiazone method, the fluctuations in purity of the Roman issues in the troubled eighth century are apparent. Additional analyses are needed to define more precisely the peaks and troughs in the purity of manufacture. C. Morrisson has published results for Sicilian and Carthaginian gold4 obtained by proton activation together with a few microchemical results. Her results show discrepancies, but they are not as broad as the results here, although she does point out the differences in the silver to copper ratio. The results for one coin have had to be recalculated, the Syracusan semissis of Tiberius II (698-705, MIB 39). It was reported as Au 75.8%; Ag 7.4%; Cu 16.52%. After revision, the contents are Au 74.8%) Ag 22.1%, Cu 2.5% and Pb 1.6%. Another coin can be added to her listing, a Syracusan tremissis of Justinian H's second reign (705-11, MIB V22):5 Au 78.8%, Ag 12.1%, and Cu 9.1%. This last example is particularly of interest in that Oddy used the same coin in a test for specific weight. Assuming an alloy of gold and silver), then the gold content would be 87.3%; assuming a gold and copper alloy, then the gold content would be 91.8%.

Rome

Leo III

Denomination Date Control Marks Reference Weight Amount of stroke Au Ag Cu Pb Sn Color
1. Sol 719/20 ⌈|* BMC 58 4.26 g 38.72 g 74.7 8.1 16.4 0.8 golden
2. Sol 731/2 │�* BMC 36 4.19g 43.22 g 9.9 65.1 25.0 silvery
3. Sol 732/3 ▵|+ BMC 37 3.99 g 64.22 pg 34.6 40.3 24.8 0.3 silvery
4. Trem 732/3 *|A ANS(DOC 79) 1.40g 36.46\xg 86.2 13.8 golden
5. Sol 741/2 *|I BMC 35 4.08 g 66.46 g 25.7 19.6 28.2 26.5 silvery

Artavasdes

6. Trem 742/3 *|* BM 1.38g 35.87 g 59.4 22.6 17.7 0.3 silvery

Constantine V

7. Trem 743-54 *|R BMC S4 1.37 g 31.64 g 40.4 44.3 15.3 elektr
8. Sol 754-57 *|R BMC 59 4.17g 57.28 g 69.8 23.5 6.0 0.7 golden
9. Sol 754-57 *|R Cambridge 31.56 g 39.5 30.7 29.1 0.7 silvery
10. Trem 754-57 R|m BMC 65 1.25g 52.40 g 16.3 14.0 69.7 silvery + coppery
11. Sol 757/8 I|▵ Goodacre coll. (DOC 33) 47.93 g 12.25 2.25 85.5 coppery
12. Sol 757/8 I|▵ Goodacre coll. (DOC 33) 24.3 75.7 coppery
13. Trem 757/8 I|▵ BMC 61 1.36 g 54.58 g 28.7 27.7 43.6 silvery
14. Trem 757/8 I|▵ Hahn coll. (DOC 40) 1.375 g 12.73 g 7.7 38.1 54.2 silvery
15. Sol 764/5 |⌈ Hahn coll.(DOC 30) 2.94 g 17.12 g 23.3 76.1 0.6 coppery
16. Sol 765/6 |△ BMC 58 3.56 g 81.35 g 16.1 22.9 61.0 silvery/coppery
17. Trem 765/6 |△ BMC 62 1.43g 65.11 g 51.1 12.4 36.5 golden
18. Sol 771/2 R|I BMC 52 4.17g 40.91 g 32.1 52.7 14.6 0.6 faintly golden

Leo IV

Denomination Date Control Marks Reference Weight Amount of stroke Au Ag Cu Pb Sn Color
19. Sol 777/8 A|* BMC 71 (Leo IlI) 3.59 g 145.12 g 7.1 92.9 coppery
20. Sol 777/8 A|* Goodacre coll. (DOC Leo III 98) 92.92 g 8.7 91.3 coppery

Ravenna

Constantine V

Au Ag Cu Pb Sn Zn Color
21. Trem 746/7 |I� BMC 72 1.28g 64.36 g 52.0 33.5 9.9 4.6 silvery
22. Trem 747/8 |A ANS (DOC 48) 1.22g 24.68 g 56.5 13.5 27.0
23. Trem 747/8 |A BMC 74 1.39 g 84.33 g 46.7 20.5 25.4 2.5 2.2 2.7 faintly golden

Aistulf

24. Trem 754/5 |H BMC 76 (c. V) 1.35 g 25.62 g 73.0 2.1 24.9

Roman Petty Silver

1. Pope Constantine 708-15 ANS (DOC Constantine V 44) 31.17 g — 69.2 30.8
2. Pope Constantine 708-15 BMC (Constantine V) 66 33.83 g — 68.6 31.4
3. Pope Gregory II or III 715-41 ANS (DOC Leo III 92) 64.24 g — 36.9 63.1 traces

End Notes

1
The chemical method is described in H. Ballczo and R. Mauterer, "Zarst�rungsfreie Ultramikroanalyse archäologischer Fundst�cke," Fresenius Zeitschrift für Analytische Chemie 295 (1979), pp. 36-44; 298 (1979), pp. 269-72; 299 (1979), pp. 49-48. For its most recent numismatic utilization, see LNV 2 (1983), pp. 365-98.
2
C. Morrisson, report on the Symposium on the Application of Scientific Techniques to the Study of the Coinage of Europe and the Mediterranean World from 500 to 1500 A.D. in London, 6-7 April 1984, BSFN 39 (1984), p. 512.
3
M. D. O'Hara and I. Vecchi, "A Find of Byzantine Silver from the Mint of Rome for the Period of A.D. 641-752, SNR 64 (1985), pp. 105-56.
4
C. Morrisson, J. Barrandon, and P. Poirier, "Nouvelles recherches sur l'histoire mon�taire byzantine: evolution compar�e de la monnaie d'or � Constantinople et dans les provinces d'Afrique et de Sicile," J�B 33 (1983), pp. 367-86.
5
This tremissis issue was hitherto unknown, but could be presupposed as corresponding to the solidus MIB 13 and the semissis MIB 18.

THE DEBASEMENT OF THE PROVINCIAL BYZANTINE GOLD COINAGE FROM THE SEVENTH TO NINTH CENTURIES

W. A. Oddy

This is an interim version of a lecture given to the Royal Numismatic Society a few years ago. The analysis results are presented in graphic form for a large number of Byzantine gold coins from the mints of Italy, Sicily, Sardinia, and Carthage from the reign of Constans II until the cessation of minting activities. The results on the individual coins together with those on related series will be published in detail on another occasion.

METHOD OF ANALYSIS

All the coins were analyzed using the specific gravity method which has received much publicity in recent years.1 It has also previously been used by other scholars to study the composition of later Byzantine gold.2 The method is easy and is completely non-destructive. It also has the advantage of sampling the whole coin. However, it has the disadvantage of only being accurate for binary alloys (i.e. mixtures of gold with only one other metal) and towards the end of the series in both Italy and Sardinia it is clear that both silver and copper are alloyed with the gold. Nevertheless, for the purpose of studying debasement, the technique is indicative of general trends if not of exact values.

Until debasement becomes severe it is fairly safe to assume that only gold and silver are present in any quantity and that the analysis results for the gold content are within about 2% of the true value. In fact, Oddy and Blackshaw have shown that for binary gold/silver alloys the maximum error is from -1 to +3% of the true value for the gold content. In view, however, of the wide application in recent years of instrumental techniques of analysis, such as x-ray fluorescence or radiation induced activation methods, it may be asked why such an apparently archaic method as specific gravity is still being applied. The answer lies in the speed of the fact that it can be carried out in situ in coin cabinets, and the fact that it is quite adequate for the aim of the research which was to follow the debasement of the series.

THE GOLD OF CONSTANTINOPLE

Before discussing the debasement of the provincial Byzantine series it is necessary to establish what was the gold content of "pure" gold in this period. This rather paradoxical statement arises because gold invariably occurs in nature alloyed with other metals from which it has to be separated by a process of refinement. In antiquity the methods used to refine the gold were never completely successful, so "pure" gold always contained some alloying elements—usually mainly silver. Various studies have shown that up to the end of the ninth century the gold coinage of Constantinople was not debased and consisted of almost pure gold.3 In fact, the gold of this mint almost always lies within 4% of absolute purity and it is reasonable to assume that this represents "pure" gold as far as the technology of the period is concerned.

Specific gravity results on a random selection of coins from the mint of Constantinople are represented in the Table, from which it can be seen that all contain more than 95% gold, with an upper limit of about 98%. The median figure is 96.5% and the average of the results is 96.4%, hence consideration of the amount of debasement must be made by substracting the measured gold content from 96.4% instead of from 100%.4

Gold Content of solidi of Constantinople

Lowest Average Highest
Leontius, 695-98 96.3% 96.7% 97.1%
Tiberius III, 698-705 95.1 96.5 97.3
Justinian II, 705-11 95.7 96.6 97.8
Philippicus, 711-13 96.2 96.2
Anastasius II, 713-16 95.8 96.7
Theodosius III, 716-17 95.6 95.6
Leo III, 717-41 95.6 97.2
Constantine V, 741-75 96.2 96.3 96.4
Nicephorus I, 802-11 95.8 96.1 96.4
Michael I, 811-13 97.6
Leo V, 813-20 97.3 97.3
Theophilus, 829-42 96.6
Michael III, 842-67 96.1
Average 96.4%

End Notes

1
E. R. Caley, "Validity of the Specific Gravity Method for the Determination of the Fineness of Gold Objects," Ohio Journal of Science 49 (1949), pp. 73-82; E. R. Caley, The Analysis of Ancient Metals (Oxford, 1964); H. A. Das and J. Zonderhuis, "The Analysis of Electrum Coins," Archaeometry 7 (1964), pp. 90-97; M. J. Hughes and W. A. Oddy, "A Reappraisal of the Specific Gravity Method for the Analysis of Gold Alloys," Archaeometry 12 (1970), pp. 1-11; W. A. Oddy, "The Analysis of Gold Coins–a ComParison of Results Obtained by Non-Destructive Methods," Archaeometry 14 (1972), pp. 109-17; W. A. Oddy and S. M. Blackshaw, "The Accuracy of the Specific Gravity Method for the Analysis of Gold Alloys," Archaeometry 16 (1974), pp. 81-90.
2
P. Grierson, "Notes on the Fineness of the Byzantine Solidus," Byzantinische Zeitschrift 54 (1961), pp. 91–97; L. Brunetti, "Nuovi, Orientamenti statistici nella monetazione antica," RIN 52-53 (1950-51), pp. 3-74.

THE GOLD OF CARTHAGE AND SARDINIA

The gold coins of the Carthage mint at this period were very small and very thick, a characteristic shape which they had assumed gradually during the reign of Heraclius And his predecessors. They are sometimes dated by letters in the reverse inscription, but this is not invariably the case. The solidi of Constans II fall into six classes,5 corresponding to those at Constantinople. Of the ten coins analyzed for this reign (see Figure 1), eight of them are unalloyed, and only a solidus of class Ie and a rare semissis fall below 95% gold. The solidus appears to be perfectly normal and the low gold content must be attributed to carelessness in refining the gold, rather than to deliberate debasement.

image

Fig. 1 Coins from Carthage (Circles) and Sardinia (Diamonds)

Under Constantine IV, more than half the solidi analyzed fall below the "pure gold" level, two of them by an amount which is unlikely to be accidental. Grierson has divided the solidi of Carthage into four classes,6 but only III and IV are represented in the British Museum. The two coins which contain 80% (BMC 41) and 87% (BMC 42) gold are a problem. Stylistically there is nothing to suggest that the 80% coin is irregular, but the composition must raise the question as to whether it should be regarded as either a contemporary forgery or, perhaps, an Arab imitation. As far as the 87% coin is concerned there is slight evidence in this case that the coin is a copy. On the reverse the indictional letter Z has been transferred from its usual position to the right of the cross potent to the top of the staff held by the figure of Tiberius, where it replaces the usual cross. This mistake, together with a gold content of 87% perhaps suggests that it might be another Arab copy.7 How far these tentative suggestions can be substantiated will only emerge when more of these Carthaginian solidi have been analyzed. It is important to stress that these remarks are based only on a limited study of the British Museum material, together with that in the Dumbarton Oaks and Bibliothèque National catalogues, but Grierson did suggest in 19508 that certain members of class II are Arab copies. Although he has since changed his mind about the coins in question, it would now be wise to check on their gold content, as it may well emerge that Arab or possibly Berber imitations are more numerous at this period than has been realized.

Under Justinian II solidi of Carthage are rare and the British Museum only contains one which may be dated to 695 A.D., the year of the fall of Carthage to the Arabs. It is significant, in the light of the above remarks, that this coin has a gold content of 96% at a time when the hard-pressed Byzantines might be excused for allowing the quality of the gold to fall.

It is assumed that the mint of Sardinia was started up in the early 690s before the fall of Carthage, and it is interesting to find that the coins of the first reign of Justinian II are of good gold because in both Italy and Sicily at this time significant debasement had taken place. Under Leontius Carthage was retaken for a few months but no coins are known from this reoccupation, nor are any gold ones known of Sardinia. Gold was coined, however, at Sardinia under Tiberius III and, although only two coins are available in the British Museum, it seems likely that the gold coatent of 95% and 90% represents a definitive debasement rather than incompetent refining. If this is so, composition may possibly be used as dating evidence for the various types, of which Grierson has listed four according to the symbol in the reverse field left.9 Obviously further specimens of these solidi must be analyzed in order to test this hypothesis.

The output of the mint of Sardinia ceased in the early eighth century. DOC ascribes no coins to it for the reigns of Philippicus and Theodosius III and only one solidus for the reign of Anastasius II.10 The latest coin available for analysis was a fraction of a solidus of the second reign of Justinian II. This exceedingly rare coin has a gold content of 88%, providing further evidence for the decline in gold standard at Sardinia in the closing years of that mint.

Although the examples are few, especially for the later years, it may be concluded that the minting of gold in the regular Carthaginian mint continued using essentially fine gold right up until the fall of the city to the Arabs, and it is probable that coins which fall significantly below 95% are to be regarded as contemporary forgeries or imitations. Minting in Sardinia started using the same alloy, but after the first few years there is definite evidence for debasement, although how low the gold content had fallen by the time the mint had closed is not possible yet to say. However even though the gold content of these coins had been lowered by alloying, it was still greater than the gold content of the contemporary coinage of Italy and Sicily. The overall picture for the fineness of the coinage of Carthage and Sardinia has recently been confirmed by Poirier11 using proton activation.

End Notes

3
D. M. Metcalf, "Analyses of the Metal Constents of Medieval Coins," Methods of Chemical and Metallurgical Investigation of Ancient Coinage, ed. E. T. Hall and D. M. Metcalf, Royal Numismatic Society (London, 1972), pp. 383-434; J. Poirier, "Contribution a l'analyse de l'or antique: Application aux monnayages du monde meditteran�en dur iir au xivr Si�cle," Ph. D. diss., University of Orléans, 1983, p. 125; C. Morrisson, "Numismatique et histoire, l'or monnay� de Rome × Byzance: Purification et alterations, Académie des Inscriptions et Belles-lettres, Compte Rendu (1982), pp. 203-23.
4
Proton activation analyses conducted in France indicate that gold was 98.2% pure during the time from Anastasius I to Constantine IV.
5
DOC, pp. 468-74.
6
DOC, pp. 545-48.
7
C. Morrisson, J. N. Barrandon, and J. Poirier, ''Nouvelles Recherches sur l'histoire mon�taire Byzantine: �volution compar�e de la monnaie d'or × Constantinople et dans les provinces d'Afrique et de Sicile," J�B 33 (1983), p.283, report two other coins with lower gold content, Hahn coll. (MIB 25) 87.7% and BNC 14/Ct/A/11 (MIB 28) 86.5%.
8
P. Grierson, "A Barbarous North African Solidus of the Late Seventh Century," NC (1950), pp. 301-5.

THE GOLD OF SYRACUSE

The coinage of Sicily is also reasonably straightforward as far as debasement is concerned, and Figure 2 illustrates the analytical results from the reign of Constans II until the end of that of Basil I. During the reign of Constans II, when the regular mint at Syracuse began striking on a large scale, all except one of the available solidi contained more than 95% gold. The fractional coinage had a slightly greater tendency to fall below 95%, but the amount is not significant.

During the reign of Constantine IV less gold seems to have been coined in the mint of Syracuse than in previous years. All the coins available for analysis range between 93.4 and 94.8% fine, indicating a definitive drop in quality since the previous reign. Whether this is an intentional debasement by half a carat is impossible to say with certainty on the basis of so few results but it seems possible that it is.

image

Fig. 2 Fineness of Coins from Syracuse

Figure 2 clearly shows that during the reign of Justinian II there was a debasement at the mint of Syracuse. Grierson had classified the gold coinage according to the symbol in the reverse field, pointing out that Z, H, and θ could be a date sequence, but if so K and *K are anomalous.12 It might be possible to arrange the coins in a chronological order based on the analyses, but with so few coins available it is only possible to suggest the order which may need some rearrangemant after further analytical studies.13 The danger of reading too much into a single analysis is illustrated by considering the four Sicilian solidi of Leontius. Their gold content ranges from 84.5 to 87.5%, suggesting that their intended fineness was in the range 21� to 22 carats. If these are representative of the norm for the short reign of Leontius, then a semissis of Justinian II with a gold content of only 79% must be regarded as exceptionally low. Two solidi of Leontius of type a, without symbols in the reverse field, have a gold content of 87%, while two solidi of type b, with four pellets and a letter I in the reverse field, have a gold content of 84 and 86%; whether type b is consistently less fine than type a is impossible to tell from these few results.

For the period from 700 to 720 A.D. the results on the available Sicilian gold coinage show a general decline in gold standard which follows that of the Italian mainland at the same time but which is not in line with the gold standard of the mint of Syracuse in the period 720 to 820 A.D. The coins of Tiberius III, Philippicus, and class I of Leo III are credited to Sicily on the basis of provenance, especially that of the so-called Sicilian Hoard.14 It is a great pity that no Sicilian coins of the second reign of Justinian II, Anastasius II, or Theodosius were available for analysis, but the general picture of these twenty years seems fairly clear with a gradual decline from 84% gold down to 60% in the first years of Leo III. Grierson's visual examination of the coins led him to put the low water mark of this debasement in the second quarter of the eighth century15 with a revival under Constantine V, but the evidence here indicates that this is a little late and that the coinage was reformed soon after the start of the reign of Leo III.

In fact, from the inception of the class II solidi of Leo III the gold content was raised to about 80%, as can be seen from Figure 2. The change in style from class I is rather abrupt and DOC attributes them to Syracuse on the grounds of provenance and stylistic affinities to the copper series. The analytical results support the attribution to Sicily because, although the change in style is accompanied by a surprising increase in the gold content of the coinage by 20%, no mint on the Italian mainland could have been producing coins of such a fineness at this time. None of the coins which are attributed to Rome and to the other Italian mints, apart from those of class I, have finenesses in excess of 60% and the contemprorary gold coinage of the Duchy of Beneventum fell from about 80% in the first few years of the eighth century to about 50% in the 730s.

The same spread of results, with a mean gold content of about 80%, continues under Constantine V. No coinage of Syracuse in known for Leo IV or Constantine VI and only very rare solidi for Irene. The next Syracusan coins are of Nicephorus I, which were clearly struck to the same standard as those of Constantine V. Four coins of Micheal I have similar analytical results, the two members of class I having a gold content of 81% and the two members of class II having somewhat lower content, about 77%. Under Leo V the gold coinage is again divided into two classes, all available specimens of which had a gold content from 77 to 79%.

For a century the gold coinage of Sicily remained at a constant standard, but during the reign of Michael II a debasement was started which was complete by the reign of Michael III. The earlier coinage of Mitcheal II ranged from 75 to 77% fine, but by the end of his reign it had fallen to between 69 and 71%. However, the picture is not very clear as the only available specimen of class II has a gold content of more than 77%, while the four members of class III range from 69 to 76% fine, and the five members of class IV range from 70 to 74%. Although a spread of analytical results over a range of 4 or 5% is not unusual for members of a single class, these results give marginal support to the suggestion in DOC that classes III and IV may have been concurrent issues at separate mints as the political situation on the island was far from stable in the later part of the reign.

Under Theophilus the available coins of classes I and II had a gold content between 68 and 73%, but the more numerous members of class III have a gold content ranging from 77% down to 30%, many of the more debased pieces obviously containing large proportions of copper. The so-called gold coinage of Syracuse entered its final phase under Michael III. Only semisses are known, and one from each class was available for analysis. Class I follows the issues of the end of the reign of Theopilus, having a gold content in the region of 20 to 30%. The class II specimen has a small amount of gold, and on the basis of these two results it is possible that classes I and II are consecutive. Class III which is firmly dated to the last two years of the reign has no gold, nor have the various specimens of Basil I which continued until the final fall of Sicily to the Arabs in 878.16 The overall picture for the debasement in Sicily has recently been confirmed by Poirier17 using proton activation analysis on 30 coins.

End Notes

9
DOC, pp. 633-34.
10
Sardinian coins of Theodosius III (MIB 5 and N5) and Leo III (MIB 5) are known and Morrisson et al. have published a proton activation result for Leo III giving 93% gold (above, n. 7), p. 284.
11
Poirier (above, n. 3), p. 126; C. Morrisson et al. (above, n. 7), pp. 267-86.
12
DOC, p. 572.
13
It is now possible to distinguish between the Sicilian coins of Justinian II�s first (685-95) and second (705-11) reigns (MIB, p. 193); the debasement started in the intervening reign of Leontius II (695-98).
14
P. Grierson, "Two Byzantine Coin Hoards of the Seventh and Eighth Centuries at Dumbarton Oaks," DOP 19 (1965), pp. 207-28.
15
DOC, p. 83.
16
Morrisson, however, has published a coin of Basil I that is 27.3% gold.
17
Poirier (above, n. 3), p. 127.

THE GOLD OF ITALY

On the Italian mainland the regular mints had ceased production by 780 A.D., but for their last 40 years the so-called gold coins contained little or no gold. The analysis results are illustrated in Figure 3 where the different mints (according to the attributions in DOC) are indicated by different symbols. At the first glance Figure 3 looks rather complicated, but several trends are clear. The results for Rome show a steady fall across the graph, as do the results for Ravenna, although most of these are confined to the years 685 to 720 A.D. The results for Naples cover a similar period. Finally there are two groups of results for unidentified mints, one starting about 670 and ending about 695 and the second starting about 705 and ending between 730 and 740. The gap between these two series, if they are series, coincides with the exile of Justinian II.

image

Fig. 3 Fineness of Coins from Italy

Byzantine control over Naples waned toward the end of the eighth century, although the emperor was still recognized nominally. In the early ninth century there are two groups of large flat solidi, struck in the names of Nicephorus I and Theophilus, but which are usually credited to the mint of the duke of Naples. The first group struck in the names of Nicephorus and Stauracius appears at the right hand side of Figure 3. That they are anomalous can be seen from the fact that they contain 30 to 37% gold at a time when imperial gold coin was otherwise absent in Italy.

Much work remains to be done on gold coinage in Byzantine Italy from 660 to 780 A.D. and scholars are not in agreement about the attributions of some of the coins. It is possible that careful appraisal of the analyses presented in Figure 3, together with results from other collections, may help to ellucidate some of the problems. Meanwhile the overall picture can be some what simplified if the results for the coins of Rome and Ravenna are plotted alone (Figure 4). It is interesting to note that the inception of the debasement in Italy coincides with that in Sicily, but in Italy there was no reform corresponding to that at Syracuse during the reign of Leo III.18

image

Fig. 4 Fineness of Coins from Rome and Ravenna

End Notes

18
I am grateful to Michael Metcalf who first suggested the Byzantine provincial gold coinage as a fruitful field of research and allowed me to analyze the relevant coins in Oxford and John Kent who made the British Museum coins available for study and discussed the results at some length. In addition Michael Hendy willingly gave me access to the Barber Institute collection in Birmingham and Simon Bendall, formerly of A. H. Baldwin and Sons, lent coins for analysis, as did a number of private collectors.

BACK

PLATES JOINT REIGN OF JUSTIN I AND JUSTINIAN I

Plate 1

image

Plate 2 JOINT REIGN OF JUSTIN I AND JUSTINIAN I

image

Plate 3 JOINT REIGN OF JUSTIN I AND JUSTINIAN I

image

Plate 4 MONTE JUDICA HOARD

image

Plate 5 MONTE JUDICA HOARD

image

Plate 6 MONTE JUDICA HOARD

image

Plate 7 MONTE JUDICA HOARD

image

Plate 8 MONTE JUDICA HOARD

image

Plate 9 CARTHAGE: JUSTINIAN I AND JUSTIN II

image

Plate 10 CARTHAGE: JUSTINIAN I AND JUSTIN II

image

Plate 11 CARTHAGE: JUSTINIAN I AND JUSTIN II

image

Plate 12 CARTHAGE: JUSTINIAN I AND JUSTIN II

image

Plate 13 CARTHAGE: JUSTINIAN I AND JUSTIN II

image

Plate 14 CARTHAGE: JUSTINIAN I AND JUSTIN II

image

Plate 15 THESSALONICA

image

Plate 16 THESSALONICA

image

Plate 17 THESSALONICA

image

Plate 18 THESSALONICA

image

Plate 19 THESSALONICA

image

Plate 20 THESSALONICA

image

Plate 21 THESSALONICA

image

Plate 22 THESSALONICA

image

Plate 23 THESSALONICA

image

Plate 24 THESSALONICA

image