Studies in Constantinian Chronology

Bruun, Patrick.
Numismatic Notes and Monographs
American Numismatic Society
New York
Worldcat Works




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


Table of Contents




Contemporary history is always written by the victors, by the conquerors. When posterity treads the laborious path to truth in an attempt to find out wie es eigentlich gewesen — to quote the famous words of Ranke — great difficulties lie ahead. The written account, biased already when conceived, leads us astray. The chronicles with their unimaginative lists of sundry events and officials, compiled by industrious clerks and civil servants, had a long way to go before they reached the present day editions. The record which confronts us is the product of many rounds of mechanical copying in the monasteries, where the eye of a monk all too easily missed a line and thus created a chronological confusion.

In discussing the reign of Constantine the Great, one of the great watersheds of history, out of which was to emerge Western Christian Civilization, we are faced with these very difficulties. Yet there is one source, continuous and inexhaustible when treated properly, which can corroborate or refute the eulogies of the panegyrists, namely, the coinage.

Constantine's road to sole-rulership covered a period of eighteen years, from the summer day in 306 when his father's faithful soldiers acclaimed him Augustus in York, to the September day in 324 when he finally, at Chrysopolis, defeated his last opponent, Licinius, his fellow-ruler and rival during thirteen of these years. The triumph of Constantine indubitably was the triumph of the Christian faith, but the question whether the Church would have been persecuted and paganism championed by a victorious Licinius can never be truthfully answered. The school of the Belgian Henri Grégoire, ingeniously claiming Licinius as the true defender of Christianity, 1 and disputing the authenticity of Eusebius' Vita Constantini, one of the main sources for the history of the Constantinian epoch, has by now suffered defeat. 2 Nevertheless, it would be wrong to accept at their face value all the accusations directed against Licinius by the supporters of Constantine. During the last three years of Licinian rule the rupture between East and West was apparent. This period in the East has been named the "last persecution," 3 though on feeble grounds. Admittedly, the freedom of the Church was curtailed, Christian courtiers, civil servants and officers dismissed, and, possibly owing to over-zealous provincial governors, some adherents of the Church killed. But this does not warrant calling Licinius a persecutor. 4 The religious policy of the years immediately preceding the final clash was an intrinsic part of the general policy of the Emperors; Licinius, indifferent in religious matters, might well have appeared as a sponsor and supporter of Christendom had the wheel of fortune turned the other way.

End Notes

1 The main results of these studies were presented in two lectures at the 1960 Summer Seminar of the American Numismatic Society of New York. I am greatly indebted to the Society, and particularly to its former president, Mr. Louis C. West and to its Executive Director, Mr. Sawyer McA. Mosser, not only for encouraging the preparation of these studies for publication, but also for the stimulating experience I gained at the Seminar. Prof. A. R. Bellinger, Chairman of the ANS Publications' Committee, has read through the typescript more than once with keeneyed criticism. I am much obliged to him for numerous suggestions arising out of his familiarity with the subject matter and his demands for lucidity in presentation. Mr. George Kustas of Buffalo University, kindly read the manuscript and corrected its English. For this I want to express my sincere gratitude. However, it goes without saying that the author is solely responsible for the views expressed and the errors remaining in this volume. My thanks are also due to the staff of the ANS for friendly co-operation during the process of finishing the typescript and producing the plates. Mr. R. A. G. Carson of the British Museum has not only encouraged my research work but also assisted in providing casts for the illustrations.In the text below references to the actual coins have been made through reference to the numbers of the items listed in the Index of Reverse Legends (comprising the gold coins only). Illustrations have been indicated by inserting the number of the plate together with the number of the rev. leg., e.g. Plate IV, 100. To avoid confusion the bronze coins illustrated have been accorded letters instead of numbers. The corresponding reference will accordingly be e.g. Plate III, a. All the bronze coins illustrated are from the ANS collection except the Domitius Alexander coin (Ratto, December 1932, 722).
1 Grégoire, Rev. de l'Univ. de Bruxelles XXXVI, pp. 231–272, lately in Bull. Acad. Belgique 1953, PP. 466–483.
2 A. H. M. Jones, JEH V, pp. 196ff., further Aland Stud. Patr. I, p. 564 with ref. to Vittinghoff, Rhein. Mus. 1953, p. 373 and Telfer, Stud. Patr. I pp. 162f.
3 Cf. Roberto Andreotti's exhaustive study in Ruggiero's Dizionario Epigrafico IV, fasc. 31–33, pp. 1018ff.
4 Ibid.


Three battles mark Constantine's progress to supreme power; the Milvian Bridge, Cibalae and Chrysopolis. The first gave him Italy, the second the Balkans, the third the entire East. Now, let us follow him from Trier to Antioch, from a barely accepted Caesarship within the framework of the tetrachic system to acknowledged sovereignty, surrounded by oriental splendor, in the whole Roman empire.

In view of later events we may assume that Constantine from the beginning aspired after something more than a subordinate rank in the Western provinces. The precarious position of Constantine and Maxentius drew the rulers of Gaul and Italy together, and the alliance was confirmed through the agency of the old Herculius when Constantine in 307 was married to Fausta and encouraged to assume the title of Augustus. A year later Maximian, ambitious of power but failing to discharge Maxentius, was expelled from Rome and forced to seek refuge with Constantine. 1 The relations between Trier and Rome were broken, but about the same time Maxentius' attention was distracted from the Gallic scene by the usurpation of Domitius Alexander in Africa.

The two following years may be described as a time of gathering storms. The quiet was broken when, as the story goes, the old conspirator Maximian was found responsible for an attempt on Constantine's life. In 310 he was killed, or killed himself, in besieged Massilia. 1

This event was probably the signal for a major change in the political constellations. The filial devotion of Maxentius flared up once more and, according to literary sources, he was resolved to avenge his father's death. 2 The African usurper, however, prevented him from taking instant action. In these circumstances Constantine and Domitius Alexander naturally formed an alliance, the evidence of which can be found on the African milestone mentioning these two Augusti only. 3 At the same time, possibly a little later, Constantine appears to have broken the isolation of Gaul in Europe; both Licinius and Daza are depicted on the coin obverses of this time. 4 The consent of Licinius was of paramount importance to Constantine when engaged in a war for supremacy in Italy. A ruler de facto since 307, de iure since 308, and specifically charged with the reconquest of Italy,5 Licinius might in ordinary circumstances have been reluctant to accept Constantine's interference south of the Alps, but in 310 Galerius was taken seriously ill 6 and the stakes in the prosperous East were higher than those in the West. Both Licinius and Daza concentrated on the future division of the Eastern provinces, and the West was left to settle its disputes on its own. The fact that after the death of Galerius Daza captured Asia, reached the Bosporus and was halted only by the arrival of his rival 7 demonstrates how far-sighted Licinius had been.

Thus the stage was set. Maxentius made the first move and in early 311 dispatched Rufius Volusianus to cope with the rebellion in Africa. 1 At the same time that tension grew in the East, Galerius' condition was rapidly deteriorating. In the spring of 311 Constantine celebrated his second quinquennalia in Trier and struck, among other reverses, the exceptional type SPQR OPTIMO PRINCIPI in gold at Trier (Plate I, 201–2).

The date of this gold type can be established with a high degree of certainty. The early Treveran gold portraits of Constantine are very distinct in character; after the very first coins naming Constantine Caesar, 2 we get a series of solidi of small module, thick flan, and high relief. The minting of this kind of solidus comes to an end when, shortly after the decisive victory at Pons Milvius, Trier discontinues its gold series and Constantinian gold minting is concentrated in Italy, first at Rome and Ostia, later at Ticinum. Gold coining was resumed at Trier only towards the end of 313, possibly not before 314, and then with coins of distinctly thinner flans and larger module. 3

Now the second group of Constantinian obverses can be classified chronologically, and the systematic arrangement shows the coins with the obverse legend CONSTAN-TINVS PF AVG (Plate I, 201) to be of earlier date, the coins with the obverse legend CONSTANTI- NVS PF AVG (Plate I, 202) to be later. By a stroke of luck the chronological dividing line between these two kinds of obverse breaks can be dated to Constantine's quinquennial year, because a part of the obverses of the type VOTIS V MVLTIS X/VICTORIA AVG has the break N-T (Plate I, 312), another part the break I-N (Plate I, 313). The coins of the type SPQR OPTIMO PRINCIPI can be grouped in the same way. 4

Thus it remains to establish when the Constantinian quinquennalia were celebrated. The point of departure for the reckoning of regnal years was the natalis imperii of the ruler, regardless of the title conferred upon him on his ascent to the throne. With regard to Constantine the situation is unusually complicated. His original dies imperii, regularly celebrated in later years, was July 25th. The realization of the weakness of his position — the very fact which led him to accept the early alliance with Maximian and Maxentius — made him to a certain extent dependent on his allies. Thus he appears to have accepted the title of Augustus from Maximian on the occasion of his marriage to Fausta, although he seems to have delayed using it for some time. 1 March 31, 307 was thus considered his natalis imperii 2 until the conquest of Italy; when the Roman senate conferred upon him the title of Maximus Augustus, he can be supposed to have resumed his original natalis and dies imperii. His quinquennalia, however, would have been celebrated according to the earlier reckoning, on March 31, 311 and 312.

We return now to the legend SPQR OPTIMO PRINCIPI. The reading was no novelty, but it had scarcely been in use since the days of Trajan. 3 A slogan revived after 200 years is not commonplace, but in this context two additional facts enable us to appreciate the true significance of this issue. The first is that the same type was struck by the usurper Domitius Alexander (Plate I, a) in Africa. 4 This can be no mere coincidence. Clearly we have here another piece of evidence of the alliance between Gaul and Africa. But let us take the other fact, the fact that the type SPQR OPTIMO PRINCIPI was struck in two mints Trier and Carthage, out of reach of the Senate. The implication of the type is that the Senate pays homage to an Emperor (Constantine or Alexander) on the verge of war or actually engaged in a war with the master of the Senate (Maxentius) ! Therefore, there is but one way of interpreting the evidence of the coins, namely that Constantine and, accordingly, his African ally, had reached an understanding with the Senate or with an opposition faction of the Senate. In itself this is not surprising, for we have ample proof of senatorial discontent with the tyrant of Rome. 1

Thus the Treveran gold coins of the type SPQR OPTIMO PRINCIPI have given us the ramifications of the Italian campaign of Constantine and, indeed, the date: the spring 311 instead of the generally accepted year 312. We conclude that the battle of the Milvian Bridge was fought one year earlier than traditionally has been believed. 2 Additional numismatic evidence can be brought forward in support of this.

One of the primary literary sources of the events of these years is Lactantius' De mortibus persecutorum. Lactantius remarks "Immine- bat dies quo Maxentius imperium ceperat, qui est a. d. sextum Kalendas Novembres, et quinquennalia terminabantur." 3 Now the quinquennial celebrations were concluded in 311; Mommsen, however, escaped the difficulty by explaining that the starting point for the reckoning of Maxentius' regnal years must have been October 307 when, according to the great German scholar, he assumed the title of Augustus. 4 This day, not the day of his usurpation, should have been regarded as his natalis imperii, Mommsen maintained. In the light of recent research concerning the Maxentian coinages, his views have to be revised. 5

Of the Maxentian mints Carthage was closed before Constantine had been given the title of Augustus, but on the two last issues of the mint Maxentius appears as Augustus (on the image first as Princeps Invictus). 1 Similarly the co-ordinated issues of Ticinum, Aquileia and Rome show the rulers as Maxentius Augustus and Constantinus Caesar, respectively. 2 The subsequent issues from these three mints are the last before the break between Herculius and Constantine on the one hand and Maxentius on the other, and record Constantine as Augustus. 3 The mint of Carthage was closed at the time and had not as yet been reopened in Ostia. We can see that when the mint of Carthage closed in Spring 307, Maxentius had used the title of Augustus for some time (the duration of almost two mintmarks). 4 Again, this proves his natalis imperii to have been 306, not 307 as Mommsen believed.

The Maxentian coins imply that Maxentius was defeated in 311. Some Constantinian gold coins suggest the presence of Constantine in Rome not later than March 312, namely, the VICTORIAE LAETAE PRINC PERP, coins of thick flan and small module with, exclusively, VOT X (Plate I, 274) in Rome, VOT X (275) and VOT PR (Plate I, 276) at Ticinum. 5 The VOT X coins must have been struck in the course of Constantine's quinquennial year which ended in the Spring of 312. This means that the vota were suscepta. To regard the vota as soluta and, accordingly, the coins as having been struck in 315/316 is impossible, because coins vot(is)x(solutis) mul(tis)xx (susceptis) of entirely different module and fabric were struck at Ticinum in the course of the decennial year (Plate I, 212; III, 214), and these coins were clearly preceded by coins of similarly large module struck to Constantine's fourth consulship in 315 (Plate III, 72). 6 I should therefore interpret the evidence of the VICTORIAE LAETAE PRINC PERP solidi as follows: The type was created at Trier with VOT PR on the vota shield, i.e., prior to the initial quin- quennial celebrations, as a part of the billon triad celebrating the alliance or understanding between Daza, Licinius, and Constantine, 1 in advance of the Italian campaign. Shortly after the war in Italy gold coinage ceased in Trier. By that time the first Constantinian gold coins had been issued at Rome. The interesting point is that exclusively Treveran prototypes were used (the same applies to the mint of Ostia). 2 Only in early 312 were the first solidi struck at Ticinum, among them the VICTORIAE LAETAE PRINC PERP, before the expiration of the quinquennial year with VOT X, after Constantine's dies imperii, with VOT PR. Whether the gold coins of Ticinum are indicative of Constantine's presence in Northern Italy at the time is difficult to say, but it is quite likely. The other contemporaneous types are no mere repetitions of earlier Roman reverses, but to a large extent new creations. After the conference at Milan the Ticinese gold issues were discontinued; Constantine returned to Gaul and Trier once more became the mint responsible for the coining of gold. 3

End Notes

1 The chronology of these events is not undisputed. I have here accepted the theory propounded by Seston (La conférence de Camuntum et le "dies imperii" de Licinius, Röm. Forsch. in Niederösterreich III, 1956, pp. 178ff.) which breaks away from the traditional adherence to Lactantius' account, more closely follows Zosimus, and enables us to restrict the number of Maximian's visits to Gaul to two. According to Seston the conference at Camuntum, a much less elaborate meeting with Diocletian than has usually been believed, took place in 307, when Licinius was charged with the reconquest of Italy, whereas the rank of Augustus was conferred upon him only a year later. Cf. also Dr. Kent's arguments in NC 1957, PP. 21 ff. Andreotti, op. cit., p. 982, does not entirely accept Seston's theory, but concedes that preliminary discussions might have taken place before November 308, though not earlier than the Spring 308.
1 Stein, Histoire du Bas-Empire I 1, pp. 86f., particularly I 2, p. 454 with relevant notes and references.
2 The statues of Constantine were overthrown in Italy (Seeck, Untergang I4, p. 118). Cf. also Stein, Histoire I 1, p. 87, who does not exclude the possibility that Maximian in accord with Maxentius had acted against Constantine.
3 CIL VIII 22183 = Dessau 8976.
4 Fabre-Mainjonet in their account of the Montbuoy hoard (Gallia, Suppl. XII, p. 145) show how Gaul was isolated from the rest of the Empire at least in the year 310. The hoard does not show the subsequent trend. One would have expected the isolation to have been reflected in the hoard some years earlier. Note that Constantine did not accept the Eastern consuls of the years 309–310.
5 Andreotti, op. cit., p. 985.
6 Cf. Lactantius 33, 1ff., also Andreotti, op. cit., p. 985.
7 Lactantius 36, 1. 2.
1 Salama's contention ("Α propos de L'usurpateur africain L. Domitius Alexander," Bulletin van de Vereniging tot bevordering der kennis van de antieke beschaving te's-Gravenhage XXIX, 1954, pp. 72f.) that the rebellion of Domitius Alexander was suppressed very soon after the rupture with the old Herculius (in 309) depends on a gross misinterpretation of the coin evidence. He finds support in Laffranchi's study of the Aquileian bronzes, Numismatica XIII, 1947, pp. 17–20. For a correct evaluation of the coins, cf. Cathy E. King, "The Maxentian Mints," NC 1959, Conspectus on p. 78.
2 Maria Alföldi, "Die constantinische Goldprägung in Trier," Jhrb. f. Numismatik und Geldgeschichte IX, 1958, pp. 102f.
3 Cf. pp. 8, 48f. below.
4 Cf. my "The Battle of the Milvian Bridge," Hermes, 1960, pp. 366f. When staying at the BM in early 1960 and discussing the proofs of that paper with Mr. Carson and Dr. Kent, the latter informed me that he had arrived at the same conclusion with regard to the chronological significance of the observe breaks. The coins of the BM are now arranged accordingly.
1 Cf. King, NC 1959, pp. 56f., also note 5, and ANS Museum Notes IX, 1960, pp. 119, 126f., 130, 136 maintaining that Constantine could not have assumed the title of Augustus before July 25, 307.
2 Seston, REA 1937, pp. 197ff., Kent, "Pattern," Appendix II, pp. 74–77.
3 Mattingly, Roman Coins, p. 153, Grant, Roman Anniversary Issues, p. 149. Cf. Cohen VIII2, Index p. 436, records also SPQR coins of Septimius Severus, all of them of the Adventus type (RIC IV, pp. 82, 113, 147) and of Gallienus (not in RIC V 1). Cohen V2 nos. 998ff. records them all as being of the reverse type with the leg. in laur. wreath. The type with 3 standards, however, was not struck after the reign of Trajan (Maria Alföldi, "Trier," p. 107).
4 Kent, "Pattern" no. 589, cf. also Grant, loc. cit.
1 Groag, RE XIV, col. 2454.
2 I have dealt with the literary evidence in my "The Battle of the Milvian Bridge," Hermes 1960, pp. 361–365.
3 Lactantius 44, 4.
4 In comments to Philocalus' Calendar, CIL I, p. 405.
5 Sutherland, "The Folies of Ticinum," NC 1954, pp. 68ff., Carson-Kent, NC 1956, particularly pp. 106–117, Kent, "Pattern," pp. 19ff., and King, op. cit., pp. 53f., 56f., 61, 67, 69.
1 King, op. cit., p. 61, Carson-Kent, p. 116.
2 King, op. cit., p. 69 hesitates to accept these issues as co-ordinated. The pattern is nevertheless quite clear.
3 King, op. cit., pp. 53, 56f., 69f.
4 King, op. cit., p. 61.
5 Cf. "The Battle of the Milvian Bridge," p. 366.
6 A. Alföldi, J RS 1932, p. 17 also realized this, although he did not evaluate the significance of his observation, cf. "The Battle of the Milvian Bridge," p. 366.
1 The series of analyses of the metallic contents of these so called billon coins, not yet completed at the moment of writing, suggests the possibility that the whole series was a normal bronze series. The Iovi conservatori aug and the Victoriae laetae princ perp would have been struck after Civil War I, whereas the Daza coins of the rev. Soli invicto comiti should be regarded as ½-folles struck before Mid-313. It should be noted that the two former types were struck in two officinae, the latter in one off. only, and that the image series was the first to be struck in two off. And then, until well after Civil War I, the off. were marked A and B, whereas the Iovi conservatori aug and Victoriae laetae princ perp both have the off. mark P and S. If this suggestion is correct, the gold rev. Victoriae laetae princ perp would have been a creation of the mint of Rome and this would, in fact, strengthen my contention that the coins had been struck in the course of the decennial year. I propose to return to this question in the forthcoming volume of Roman Imperial Coinage (vol. VII). Cf. also note 1, p. 49 below.
2 "The Milvian Bridge," p. 366 and pp. 48f. below.
3 Maria Alföldi, "Trier," pp. 105–III mistakenly assumes a steady flow of gold from the Treveran mint at this time, although she (p. 102) remarks for the first period of Constantinian gold coining that Die Goldprägung erfolgte … stossweise.


I. The Western Mints

Let us now proceed to the Bellum Cibalense. In the seventeenth century the date of this war was established by Gothofredus as 314. 1 When arranging the bronze coins of Arelate some years ago, I found it impossible to accept this dating because of the evidence of the coins; 316 appeared to be a much more likely time for the first war between Constantine and Licinius. A study of the literary sources supported my theory, though not unanimously; 2 other scholars have in the meantime shown that the year 316 in actual fact is the only date compatible with all the sources. 3

After studying the Arelate coinage I had an opportunity of working through the coinages of all other mints. Therefore, I propose to demonstrate how the coins in no single mint speak against the redating of the Bellum Cibalense, and in several instances permit no other conclusion.

Let us start with Arelate. We know that the mint was created as a substitute for Ostia, and that the personnel of the Ostian mint was moved to Southern Gaul, 4 with the possible exception of some skilled die-cutters sent to Ticinum.5 The time can be established roughly as shortly after the break of relations between Constantine and Licinius on the one hand and Daza on the other, i.e., any time after the conference at Milan in early 313. The exact time when the mint of Ostia was closed cannot be ascertained, but Daza is adequately represented on all Ostian issues. 1 Now my British colleagues have argued that Arelate must have started working in mid-314, immediately before the Civil War I, traditionally fixed to the same year. 2 Thus the war would not have made any impact on the coinage of this mint. Their main argument appears to be that about a year was required for the transfer of the mint of Carthage to Ostia and that an opening date of 313 at Arelate "presupposes an operational celerity not evinced in the earlier move." 3 It is, however, quite clear that Messrs. Carson and Kent have exaggerated the duration of the transfer of the African mint to Italy. Miss Cathy E. King has pointed out in her closely argued study of the Maxentian mints that the closure of Carthage must have been later than the British team supposed, 4 and that the reopening of the mint in Ostia took place almost six months earlier than they had thought. 5 Applied to Arelate this means that minting must have started well before the New Year's Day in 314; not even a start in the summer 313 should be considered impossible. 6

This, of course, is conjecture based on an estimate of the duration of the transfer of the Ostian mint. The coins themselves are much more outspoken, regardless of the fact that of the early series comprising mainly the reverse types MARTI CONSERVATORI and SOLI INVICTO COMITI, the date of two coins only can be definitely as sessed as 315, namely, two consular coins marked PARL and image 1 On the other hand the sequence of mintmarks is established beyond doubt as PARL, image followed by the three issues marked with image in the field. 2 Again, the three (?) issues marked with image comprising the first coins of the Caesars are firmly con nected with the image series, 3 the very series illustrating the conflict with Licinius as Licinius disappears from the obverses in the course of the first series and is absent from the two following. This conflict immediately preceded the nomination of the Caesars (officially confirmed at the settlement at Serdica on March 1, 317). The absence of Licinius from certain series cannot be due to pure chance, as the coins preserved to our days are fairly numerous. Of the image mark a single Licinius obverse is known against 27 Constantinian; for the subsequent image my material contains 6, for the image mark 31 obverses of Constantine. The ratio of Constantinian to Licinian obverses for the other pre-317 marks is PARL 78/31 (15/7 63/24) image 50/22 image 19/3 image 55/31

Except for the image issue we have a substantial representation of Licinian obverses. Assuming for the sake of argument that Arelate started coining in mid-314, and that Civil War I was fought in 314, as the British scholars contend, we are faced with the inexplicable fact that Constantine throughout the war and the preliminaries to the war faithfully remembered his adversary on the coins, but a couple of years later, when peace ruled the Empire, ignored his fellow-ruler. The short series of PARL coins (with large flans and with the long obverse legend IMP C CONSTANTINVS PF AVG) undoubtedly struck before the consular coins of 315 marked PARL, 1 actually have a higher ratio of Licinian obverses than the second part of this issue with short obverse legends (15/7 and 63/24, respectively).

It is time to leave Arelate and continue to the other Western mints. Londinium, strangely enough, remained completely unaffected by the events, and regularly coined Licinian obverses up to the settlement at Serdica; but almost as an afterthought the Licinii do not appear on the coins after the war at all, obviously an indication of the isolated position of the British Isles. 2

Trier reacts differently to the political development. Licinius disappears well in advance of the coinage with the Caesars, to reappear only in the course of the VICTORIAE LAETAE PRINC PERP coinage, about 319–320. The highly complicated system of coining makes it difficult to discover the exact point at which Licinius was eliminated from the coinage. From the death of Daza Licinius is well represented on the issues marked image and image (the first 2-officinae issue). An analysis of the following series with image in the field and the officinae marked A and B, reveals a highly elaborate system, 3 and proves the Civil War to have immediately preceded the nomination of the Caesars. The sequence of issues was

  • image
    • first part comprising
      • Soli invicto comiti, Sol stg. l. (PLATE I, c–d) Constantine
      • Marti conservatori, Mars stg. r. (Plate II, e–f) Licinius
      • Genio pop Rom (Plate II, g–h)
    • second part comprising
      • off. A Soli invicto corniti, Sol stg. 1., globe low (Plate II, i) Constantine
      • off. B Marti conservatori, Mars stg. r., with a later addition of
      • Claritas reipublicae, Sol adv. l. (Plate II, j) Sons of Constantine
  • image
    • a transitory stage, first part comprising
      • off. · A Claritas reipublicae, Sol adv. l.
      • off. B Claritas reipublicae, Sol stg. r., look. l.
    • second part comprising
      • Soli invicto comiti, off. · A Sol stg. l. (Plate II, k) off. B Sol stg. r., look. l. (Plate II, l)
      • Principi iuventutis, off. · A unbroken rev. leg. (Plate II, m) off. B broken rev. leg. (· I–V)
      • Claritas reipublicae, off. · A, B Sol stg. l. (Plate II, n–o)

The interesting point is that Licinius disappears from the coinage a very short time before the Caesars, i.e., the sons of Constantine, appear, a corroboration of the theory that Constantine elevated his sons to princely rank in advance of the settlement at Serdica. 1

The gold coinage of Trier also contributes to our understanding of the course of events. In his analysis of the political development of 314–316 the German scholar Chr. Habicht 2 concludes that the plot of Bassianus, which ultimately led to the Civil War I, was contrived shortly after the decennial celebrations on July 25, 315. Now the vota coins struck for Constantine's decennalia follow two different formulas, mentioning either a plurality of Augusti (283–5) or simply one Augustus (215–6). Although, at the present state of knowledge, we have no means of telling definitely the internal order of these vota coins, it seems natural to assume that the decennalia were celebrated twice, just as ten years later first in Nicomedia, then in Rome, and that the coins with the plural Augg were struck in 315, those with the singular Aug 316 when the war was imminent.

Moving south to the coins of Lugdunum, we find a mint extremely reluctant to remember Licinius; altogether I have recorded four coins only, all of the type GENIO POP ROM marked image with the aid of similarly marked consular coins dateable to 315. 1 Thus they give no indication of the date of the war.

The Italian mints constitute no major problem. Ticinum strikes a long sequence of marks with the types Soli invicto comiti and Marti conservatori of gradually decreasing module, commencing with the short obverse legend and ending with the long obverse legend IMP CONSTANTINVS PF AVG. 2 The last issue before the nomination of the Caesars was marked image and comprised no obverse of Licinius. This mark was of very short duration and, obviously, before the settlement at Serdica superseded by the mark PT, which introduces the types PRINCIPIA IVVENTVTIS and CLARITAS REIPVBLICAE for the Caesars Crispus and Constantine II, respectively (early 317). 3

The structure of the coinage of Rome is largely similar. After a great many marks of the Sol coinage depicting both Constantine and Licinius, we find the last marks before the appearance of the Caesars without any Licinian obverses: image with the unusual reverse legend SOLI INVICTO COMITI DN (and, curtailed, SOLI INVICT COM DN) 4 and RP: SOLI INVICTO COMITI, Sol with captive. The fact that Licinius is represented on a few coins only of the first and very substantial issue with obverses of Crispus and Constantine II once more illustrates that Constantine nominated his sons Caesars prior to the settlement at Serdica. 5

Aquileia did not coin continuously during the years 313–317; one short issue only was struck without any bearing on our chronological problem. It seems to have been a continuation of the Sol type with captives, struck originally with obverses of all the three Augusti, Constantine, Daza and Licinius.1 The only unusual feature at this early stage is the absence of obverses of Licinius, a very surprising phenomenon, particularly in a mint situated on the border of the Licinian part of the Empire.

End Notes

1 Habicht, Hermes 1958, p. 360.
2 The Constantinian Coinage of Arelate, pp. 17–22, the main difficulty being the text of the Consularia Constantinopolitana (Chron. I, p. 231).
3 Particularly Habicht, Hermes 1958, pp. 360–370. Andreotti, op. cit., p. 1002 while accepting the validity of the assumption of a conflict in 316, maintains that hostilities broke out for the first time in 314. The joint consulship of Constantine and Licinius would thus have been a sign of a temporary settlement before New Year 315.
4 Arelate, p. 3.
5 A. Alföldi, "The Initials of Christ on the Helmet of Constantine," Studies in Honor of Allan Chester Johnson 1951, pp. 303–311 obviously is wrong when assuming that the best of the die-cutters were transferred to Rome and Ticinum.
1 Arelate, p. 4, coin lists pp. 58ff. Messrs. Carson and Kent accordingly point out that Ostia must have been working in Mid-April (p. 117). They refer to a letter by St. Augustine (Ep. 88,2, not 82,2 as given by Carson- Kent). St. Augustine, however, records a letter by the proconsul Africae Anicius Anulinus to Constantine, dated April 15, 313, in Carthage (as to the genuineness of the letter, cf. H. Kraft, Kaiser Konstantins religiose Entwicklung, pp. 32ff., particularly p. 33, note 2). But the letter is not valid evidence in this context; it only shows that the news of the impending war had not reached Carthage by Mid-April. If this were the only evidence and the argument otherwise sound, Ostia might have ceased coining in March, if not earlier. It seems that the controversy regarding the time of the transfer of the Ostian mint to Arelate is a point of academic interest only. The fact that Daza appears to be present in all issues proves that coining easily could have been discontinued before Daza's disappearance from the obv., but not later than the last date when Daza can be attested as Augustus, acknowledged in the West (April 15, 313 in the inscription CIL VI, 507 preserved by Cyriacus Anconitanus Parm., and recording the third consulships of Constantine and Daza).
2 Carson-Kent, pp. 116f.
3 Ibid.
4 King, op. cit., p. 62.
5 King, op. cit., cf. Conspectus on p. 78.
6 Cf. note 1 above.
1 Arelate, pp. 17, 64, 66.
2 Arelate, pp. 23f., Carson-Kent, pp. 117f.
3 Arélate, pp. 24f.
1 Arelate, pp. 15ff.
2 Kent, "Pattern," pp. 37f.
3 The system has been outlined in general terms in my "The Disappearance of Sol," Arctos 1958, N. S. II, pp. 17ff.
1 "The Disappearance of Sol," p. 36, Arelate, pp. 25ff., cf. also Habicht Hermes 1958, p. 366.
2 Habicht, op. cit., pp. 374f.
1 Kent, "Pattern," p.44; the consular coins, cf. Maurice II, pp. 193f., rev. IV 17.
2 "The Disappearance of Sol," pp. 21f., cf. Kent, "Pattern," p. 47.
3 "The Disappearance of Sol," p. 21.
4 Kent, "Pattern," pp. 50f.
5 "The Disappearance of Sol," pp. 20f.

2. Siscia and Thessalonica

Civil War I gave Constantine almost all the Licinian territory in Europe, and two new mint cities, Siscia and Thessalonica. The coinage of Siscia is straightforward; the last Licinian gold issues have no obverses of Constantine, the very last actually stressing the singular Augustus 2 and demonstrating that the Western ruler was not accepted by Licinius.

The Licinian bronze coinage similarly ends with a short issue with obverses of Licinius alone. 3 This, however, does not conclusively prove the war to have been fought in 316. If we date the war to 314, we could postulate a discontinuance of coining for the two years between the war and the settlement in Serdica, a solution suggested by Dr. Kent. 4 That this cannot have been the case is shown by the mint of Thessalonica, whose fortunes were different from Siscia in that it remained in the hands of Licinius throughout the war. After the outbreak of the war against Daza we have four series of bronze coins struck at Thessalonica: 5


The sequence of issues is quite clear, and equally clear is the fact that the Iovi conservatori aug issue was struck during the hostilities between Constantine and Licinius. The puzzling feature of the sequence is the Iovi conservatori with the Constantinian obverses dominant. Dr. Kent dates this type to 317 and presumes a discontinuance of coining between 314 and 317. 1

The two first of these series were struck in the names of both Constantine and Licinius with Licinius dominating, as is natural. The third had exclusively Licinian obverses, whereas the fourth had both Constantinian and Licinian obverses, the former being predominant. The type remains the same throughout, Jupiter standing holding Victory on globe with an eagle carrying a wreath at his feet.

The puzzling character of the Iovi conservatori issue calls for an explanation. Viewed in the light of the events in the theatre of war, the fate of Thessalonica can be interpreted in four different ways:

Dr. Kent accepts the first hypothesis and remarks that no bronze coins are attributable to Thessalonica in the years 314–317. 2 This, however, appears to be incorrect as is shown by the gold multiple VICTORIAE LAETAE AVG N/VOT X/MVL XX (Plate IV, 272) struck to Constantine's decennalia. The immediate prototype was the Ticinese medallion VICTORIAE LAETAE AVGG NN/VOT X/MVL XX of exactly the same imagery (Plate III, 273). Now, supposing that both Ticinum and Thessalonica were in the hands of Constantine during the decennial year (315/316), why should the Emperor have chosen to hail the concordia of the Augusti at Ticinum but not in Thessa- lonica ? It should also be noted that scarcely any other gold pieces are attributable to Thessalonica at so early a juncture. 1 No satisfactory explanation for these extraordinary circumstances can be found.

Should hypothesis (ii) be accepted, we can find no reason for Licinius to discontinue coining after the war in one of his chief mints. The fact that it started working as soon as it fell into the hands of Constantine, proves that the staff had not been evacuated. Two years of inactivity at one of the major mints appears impractical and inexplicable. The probability that Licinius should have struck the Iovi conservatori dominated by Constantine at Thessalonica during these years is negligible.

Turning from numismatic surmise to political reality, hypothesis (ii) can be dismissed as impossible. If military operations had ceased in 314 or early 315 and the peace treaty was signed two years afterwards, Constantine must have withdrawn his troops from Licinian territory during the negotiations for peace. It should be remembered that during the campaign he advanced very far into Licinian territory, actually to Philippopolis, thus cutting off his adversary, who had retreated to the northwest in the direction of Beroea, 2 from Macedon and Thessalonica. Now, can we really believe that Constantine gave up this advantageous position, his sole guarantee of gaining anything in the would-be-settlement ? The answer is an emphatic "No" !

The disposition of the opposing armies in the theatre of war should be kept in mind when discussing the two remaining possibilities (iii) and (iv), both of which imply that the war was fought in 316. In either case Thessalonica would have been isolated from the Licinian main army during the war, while the Constantinian fleet threatened the communications with its sovereign. 3 In actual fact the problem whether (iii) or (iv) is the correct solution boils down to the question whether Constantine, having gained possession of Thessalonica, continued striking the Licinian Iovi conservatori as the Augustan counterpart to the types of the Caesars (PRINCIPIA IVVENTVTIS, CLARITAS REIPVBLICAE, VIRTVS MILITVM DD NN).

A close examination of the bronze coins shows that the patron of Christianity, who many years earlier broke away from the tetrarchic system, after the war hailed Jupiter his tutelary god. For Constantine two different obverse legends were used in the course of the Iovi conservatori issue, namely Imp Constantinus pf inv aug (Plate II, s) and Imp Constantinus pf aug (Plate II, u). Of these obverse legends the former is the earlier; in prewar times Constantine had been depicted as a bearded middle-aged man (the portrait employed was probably a Licinian one), in the earlier part of the Iovi conservatori issue we find a broad tetrarchic portrait of a youthful man without beard; in one instance a Licinian portrait with beard is connected with the obverse legend Imp Constantinus pf inv aug. This portrait is subsequently superseded by a new, narrow Constantinian portrait, the same portrait we encounter on the first Constantinian gold coins of Thessalonica (Plate IV, 235), and except for a few instances of the inv (ictus) obverse legend (Plate II, s), all these portraits are coupled with the short obverse legend pf aug (Plate II, u).

The pattern is now clear. At Thessalonica a new portrait of Constantine was created in the traditional style, and issued together with obverses of Licinius. The obverse legend for this part of the issue was Imp Constantinus pf inv aug. Before the obverse legend was curtailed once more a new Constantinian portrait was created, this time an entirely different one, breaking away from the tetrarchic tradition. It is a fair inference that this happened when Constantine took possession of the city after the settlement at Serdica. At the same time the Licinian obverses disappeared. Expressed in terms of political history this shows that Constantine did not capture Thessalonica during the war, but that the mint city itself during the war tried a policy of appeasing both Constantine (named invictus) and its legitimate sovereign Licinius, a very natural reaction for a city in such an awkwardly isolated position as Thessalonica's. After the war the city fell into the hands of Constantine, the tetrarchie portrait was superseded by a new one, gold coining started and Licinius disappeared from the bronze coinage. Thus hypothesis (iv) proves to be the correct solution of our problem. 1

In the Thessalonican coinage there is one more detail to explain, the decennial multiple VICTORIAE LAETAE AVG N. I propose, however, to return to this piece later. 1

Before continuing with the bronze issues of the Eastern mints, it may be noted that, as regards gold, Siscia is the only mint mirroring Civil War I. The last Licinian gold issue Iovi conservatori aug was struck without obverses of Constantine, whereas the preceding issue had depicted both rulers. The exact date of the gold issues is, however, impossible to establish.

End Notes

1 Kent, "Pattern," pp. 58f.
2 The sequence of the last Licinian prewar issues in gold is:
  • IOVI CONSERVATORI, Emp. holding Victory on globe, m. m. SISC, obv. IMP LICINIVS PF AVG (113) with obv. of Constantine also.
  • IOVI CONSERVATORI, as previous, m. m. image Obv. LICINIVS PF AVG (112), no obverse of Constantine.
  • IOVI CONSERVATORI AVG, Emp. holding thunderbolt, m. m. image
3 "The Disappearance of Sol," pp. 24f.
4 Kent, "Pattern," pp. 6of.
5 Kent, "Pattern," p. 63. Daza actually was eliminated from the coinage in the course of the IOVI CONSERVATORI AVGG issue.
1 Ibid.
2 Ibid.
1 Cf. below p. 56.
2 Anon. Val. 17, Andreotti, op. cit., pp. 1005–1008.
3 Seeck, Unter gang I4, p. 163.
1 With regard to the Iovi conservatori as struck by Constantine after the war, there is always the possibility that the Jupiter type was continued as an emergency, very much in the same way as anachronistic gold types were issued. But would any mint official have taken the responsibility for this if Constantine really was considered Christian or an ardent champion of the Christian faith? We should also note that Constantine was in Thessalonica very shortly after the settlement at Serdica, and that the Iovi conservatori must have continued to be struck for some time. The material is abundant and the obv. of curtailed obv. legg. (pf aug) appear to be twice as common as the pf inv. aug legends. If we assume that the pf inv aug were contemporaneous with the Licinian obv., all of them struck at Thessalonica during the war, i.e. during at least 2–3 months, the numerical strength of my material (35 coins of each against about 70 of the obv. leg. Imp Constantinus pf aug), suggest that the Iovi Conservatori, almost literally struck under the very eyes of Constantine, cannot have lasted a shorter time than the previous series. On the other hand we should note that the presence of the Emperor and the court in the mint city naturally increased the local need of cash, and therefore the intensity of striking. Again, the theory that the Iovi conservatori was allowed to continue as an emergency finds little support in the fact that at the same time new types were introduced both for gold and for bronze coining. It is therefore very unlikely that Jupiter appeared on the bronze coins against the intentions of the Emperor.

3. The Licinian Mints

Of the five Easternmost mints Cyzicus and Alexandria are of particular interest because of the bronze coins of Valerius Valens struck during Civil War I. In a recent study Mr. Carson has shown that despite the numerous forgeries known, one coin of each mint of all the specimens known appears to be genuine. 2 Dr. Kent, dating the Civil War as 314, and realizing the difficulties involved if these coins be accepted as genuine, prefers to disregard them entirely. 3 In my opinion they conclusively prove the date of the war to be 316. This can be seen particularly clearly in Alexandria, where two series of coins were struck with the mintmark image 1

  • Iovi conservatori aug for Licinius
    • Iovi conservatori augg for Licinius and Valens
    • Iovi conservatori augg for Constantine, Licinius, Crispus, Licinius II.

In other words, Licinius, once having broken his relations with Constantine, first appears alone on the coins (aug!), then appoints Valens as a colleague (augg), and; finally, after the peace treaty, coins with Constantine and the two oldest Caesars (coins of Constantine II may yet be found; otherwise the absence of the younger son of Constantine might reflect a late stage in the negotiations between Constantine and Licinius when the latter had deposed Valens and had agreed to accept one of the sons of Constantine as Caesar, but not the other. In theory, three members of the House of Constantine would have given the West a majority in the imperial collegium. 2 Such an Alexandrian issue could possibly be connected with the papyrus 3 which shows that Licinius, ready to compromise, actually accepted the consuls of 317 before Constantine). 4 The chronological coherence of the events is proven by the use of a single mintmark for all these coins.

The coinage of Cyzicus is less explicit although the issue with Valens immediately precedes the postwar series with obverses of all five rulers. 5

To these proofs I may add that whereas the Nicomedian and Heraclean coinages give no clues as to the chronology of Civil War I, the Antiochene sequence of marks and obverses conforms to the observations regarding Cyzicus and Alexandria. After the fall of Daza and before the rise of the Caesars we have three marks and four series of coins: 1

(a) image (b) image both Iovi conservatori augg, obv. head laur. r., (c) image Iovi conservatori augg nn, bust laur., drap. r. with curtailed obv. leg. IMP LICINIVS AVG, and, finally, (d) Iovi conservatori marked image with the same obv. leg. and a bust rad., drap., hold. sceptre, globe and mappa. This bust is later to be used with the remaining Licinian bronze issues of this mint, all of which included the sons of the Constantinian and Licinian Houses, and entitled the mint S(acra) M(oneta).

Of the remaining Eastern gold coinages only Nicomedia sheds additional light on the course of events. Nicomedia appears to have been the chief mint city of Licinius. In 315/316 Licinius commemorated the decennalia of Constantine by issuing the type VICTORIAE AVGG NN/VOT X/MVL XX; shortly afterwards, obviously before 317, the Iovi conservatori without any Constantinian obverses was minted. The issue appears to coincide with Civil War I. 2

End Notes

1 Cf. below p. 56.
2 The "Geneva Forgeries," NC 1958, pp. 47–58.
3 Kent, "Pattern," pp. 31, 68, 73. Seston, Relazioni di X Congresso internationale die scienze storiche (1955) VI, p. 782, note 3 is in error when assuming the existence of a gold coin of Valerius Valens, cf. also Habicht, op. cit., p. 365, note 2.
1 Kent, "Pattern," p. 73 publishing both parts of (ii) as one issue (sic).
2 Habicht, op. cit., p. 366.
3 PThead. 57. Habicht's interpretation (pp. 365ff.) takes no account of this papyrus, nor does Lafaurie in Bull. Soc. Franҫ. Num. 1954, p. 293.
4 Entered in the Fasti for February 17, 317 (Degrassi, I Fasti Consolari, p. 79) as noted also by Lafaurie, loc. cit.
5 Kent, "Pattern," p. 68.
1 Kent, "Pattern," pp. 70f.
2 An obv. of Licinius II shows the m. m. image to have been in use after March 1, 317. The preceding issue of the same m. m. was a IOVI CONSERVATORI without any obv. of Constantine.


The brilliant Austrian numismatist Georg Elmer once made the provocative suggestion that the late Roman gold mint was a travelling mint. 1 Unfortunately the paper in which he developed this idea was never published in full; it has therefore been impossible to test his arguments in detail. Nevertheless Dr. Kent of the British Museum later followed the same line of thought in his study of the late fourth century gold coinage 2 and in a survey of the tetrarchic gold coinage Professor Pink hinted at a similar phenomenon by noticing how the center of gravity with regard to the coining changed according to the political activity of the Emperor. 3

I now propose to demonstrate that the travelling mint was a reality, and venture to illustrate how the organization worked in Constantinian times. I do not pretend to be able to give more than an outline, and I am fully aware of the fact that future research might occasion revisions in details.

The notion of a travelling mint is closely connected with the fact that the Emperor, particularly before the foundation of Constantinople, had a moving headquarters and no fixed capital as an ad- ministrative center. Thus it becomes of primary importance to establish his itinerary; again this has previously largely been done on the basis of his legislation, his edicts, rescripts and so on, grouped around certain events, unanimously fixed geographically and physically by our literary sources. Yet the milestones available are few, and the itineraries have remained largely conjectural. It is my conviction that a close analysis of the gold coinage enables us to construct and reconstruct the Imperial chronology with an accuracy almost equal to that derived from a study of the legislation. The final result will, of course, emerge out of an effort to reconsider the "Regesten" in the light of numismatic research. In this context, however, I have thought it wisest to touch upon these legal matters only when they directly have affected the arrangement of the gold issues; a systematic revision of the dates of the Theodosian Code, the most important source for a study of Constantinian legislation, would certainly be outside the compass of this paper.

Before attempting a discussion of details, a few words should be said with regard to the gold coinage as a whole. A striking characteristic is the wide range of types employed, particularly in the earlier part of the reign of Constantine — contrary to the practice adopted for the bronze coinage. Whereas the bronze coinage, at least in geographically restricted areas, generally speaking, was uniform, the gold coinage clearly was highly differentiated. Certain mints never coined any gold at all (Londinium, Lugdunum), others short series only (Rome, Arelate, Aquileia, Cyzicus, Heraclea). Only rarely can we find exactly parallel sets of gold types in any two mints. We may find Sirmium exclusively striking the vota suscepta to the quinquennalia (263, Plate V, 264, 266, 267) of the Caesars, Aquileia as the only mint commemorating the fifth consulship of Constantine, Nicomedia striking votis x (i.e. soluta) caess nn (315) but Thessalonica coining votis x caes n (314) of exactly the same reverse type. This, I believe, is an indication that many of the gold issues are not contemporaneous but successive.

On the other hand, it should be stressed that in most cases the individual mints, also when striking gold, retain their particular style of portraiture, an indication that the mints remained stationary whereas the organizers or controllers were mobile.

In order to establish the chronological sequence of the gold issues we should begin by studying the travels of the Emperor and as far as possible ascertain his whereabouts throughout his reign. On the whole, I propose to do this without the aid of the numismatic material. Once Constantine's itinerary is worked out, the next step will be to discover how far the gold minting coincides with the temporary residences of the Emperor. Once the general principle is placed beyond dispute, we may be entitled to infer that certain locally limited gold series presuppose a visit of the court to the mint city in question. My aim is to reconstruct the itinerary and to demonstrate how the literary sources, mainly the Imperial constitutions, and the numismatic material can be correlated.

End Notes

1 "Wanderungen römischer Münzämter im IV Jahrh. n. C.," Mitt.d. num. Ges. in Wien 1930, p. 136. The expression "the travelling mint," though not strictly accurate, will be used in the sequel as a matter of convenience. In actual fact it seems impossible to ascertain at present to what extent the minting was mobile and to what extent the mobility of the mints coincided with the travels of the Emperor. Though the general rule probably was that the mints (the bronze minting) were stationary and that the Emperor (or court) when travelling employed the services of these mints, there are certain signs suggesting a great deal of mobility among the mint employees, signs not entirely compatible with Dr. Kent's theory of the fourth century bronze as struck not continuously "but in spasmodic outbursts of intense activity" (Pattern, p. 31).
2 "Gold Coinage in the Later Roman Empire" in Essays in Roman Coinage presented to Harold Mattingly 1956, p. 198.
3 "Die Goldprägung des Diocletianus und seiner Mitregenten," NZ 64, 1931, p. 13.


The basic source of our knowledge of the travels of the Emperor is the Theodosian Code. 1 A great number of Imperial constitutions record both date (datum) and place of issue. The year is indicated by the consular year. In many instances either the date or the place is fragmentary or missing.

Occasionally the subscriptions of the constitutions state also, or exclusively, the date and/or the place where they were received (acceptum) and properly exhibited or proposed (propositum). If nothing but the acceptum or propositum has been preserved, we get a terminus ante for the constitutions as such but, of course, no indication of the date and place of issue. If, however, we know from other sources (possibly other constitutions) the Emperor's temporary residence, it is permissible to calculate the approximate time between the datum and the acceptum or propositum. If the place of the datum coincides with the place of the acceptum or propositum, a very short time is likely to have elapsed between the one and the other. 2

Owing to the varying and uncertain sources of the Theodosian Code, the preserved dates are to a high degree conflicting. This is eloquently illustrated by Mommsen's endeavour to work out the itinerary of the Emperors in his introduction of his edition of the Code. 3 The great German scholar was compelled to confine himself to general conclusions such as (for the year 319 a. d.) "De hoc quoque anno historia nobis utilia non suppeditat. Constitutions, quae quidem locum habeant, ducunt ad Illyricum potissimum. Neque Roman neque Galliam hoc anno Constantinus adiit." 4

The only way of escaping contradictory statements about dates and places is an examination of the principles underlying the com- position of the Code. Mommsen, of course, laid a firm foundation for future research, and Otto Seeck, using his intimate knowledge of the Late Roman Empire and the legislative procedure finally succeeded in working out a logical and consistent arrangement. 1 Any such arrangement necessarily involves a great deal of emendation and interpolation. The ruling principle is to make certain dates, established independently of the constitutions, points of departure, and then consider the constitutions which by their dates and/or places of issue clearly conflict with other evidence. Thus we are compelled to evaluate the possible sources of errors occurring in the Code. Seeck devotes some 150 pages to an excellent analysis of all the problems arising in this context.

It would certainly take us too far to recapitulate Seeck's arguments. Some points of importance for the subsequent discussion should, however, be mentioned.

The fundamental reason for the inaccuracies with regard to the consular years is the Emperor Theodosius' instructions that in any case of conflicting constitutions the chronologically later should be valid. 2 All the material of the Code should therefore be arranged in chronological order within each book and title. In addition each constitution was to be dated in an easily recognisable way, according to the Fasti. 3 Seeck shows conclusively that the subscriptions frequently are interpolations based on the Fasti, and more regularly so for the earlier parts of the period covered by the Theodosian Code (311–438). The result is, to quote his words: "Wir haben hierdurch den methodischen Grundsatz gewonnen, daβ Datierungen, die den Fasten genau entsprechen, minder vertrauenswürdig sind als solche, die von ihnen in der Form irgendwie abweichen." 4

For the Constantinian time the dates of the subscriptions therefore on the whole correspond to the Fasti except for "kleine Text- fehler, me die Weglassung oder Entstellung der Iterationsziffern, Verwechslungen ähnlich lautenden Namen, z. B. Constante und Constantio u. dgl. m." 5 Once occurs a nonexistent consulate, and the only deviation from the wording of the Fasti is the use of the formula ipso A VII et C conss, which occurs three times. This wording is very likely original. 1 The close correspondence to the Fasti is also shown by the complete absence of postconsulates (to be discussed later) for the years 312–336. 2

(1) The first point of importance is accordingly that the consular years of the Constantinian period frequently are interpolations, that the iterations easily may be faulty 3 and that a confusion between rulers with similar names (Constantine I and II, Constantius and Constans) can be expected to occur. The same, of course, applies to Licinius and his son.

Generally speaking the consulships of private persons can be expected to be correct. 4

Invaluable aid for chronological considerations is rendered by extant or reconstructed lists of officials, particularly by the list of urban prefects. 5 In fact, a study of Late Roman prosopography is intimately connected with any examination of the Theodosian Code. Seeck regarded his Regesten as a "Vorarbeit zu einer Prosopo- graphie der christlichen Kaiserzeit" and he consistently tried to reconstruct the lists of the holders of Imperial offices. When the consulships are in doubt, the addressees of the constitutions frequently give valuable clues as to the date.

(2) The second point of importance is therefore that we have to examine dubious dates (as indicated by the consular years in the subscription) in the light of the prosopographical data available; these as well as the geographical dates given in the subscriptions are even more important as chronological criteria than the consulates of the Emperors. "Im allgemeinen gilt für die Kaiserkonsulate der methodische Grundsatz, daβ sie sich bis auf die Zeit des Valens herab fast beliebig untereinander vertauschen lassen. Welches in jedem ein- zelnen Falle zu wählen ist, darüber entscheiden weniger die Iterations- zifjern oder die Kaisernamen, als Ort und Adressat, mitunter noch auch der Inhalt des Gesetzes" 1

The postconsulate is another conception, which should be mentioned here. Dating by postconsulates was employed when the consuls were unknown or unacceptable. 2 Seeck states two important facts with regard to the postconsulates, (a) that they do not occur exclusively in the beginning of a year (when conceivably the authorities concerned had not been notified of the consuls of the year), and (b) that the compilers did their utmost to eliminate the postconsulates from the consular reckoning in the Code. 3 Another source of error was, of course, the unconscious omission of a postconsulate. 4

(3) The third point is thus that a seemingly incorrect date may depend on an omitted or deleted postconsulate. In these cases the right date would be either a year earlier 5 or later 6 than suggested by the subscription. Again, if the subscription records Imperial consulates, we may have several starting points when counting backwards or forward. 7

We have noted above that the accepta and proposita have little chronological significance with regard to our particular purpose. The computers of the Theodosian Code appear to have held the same view. When, however, constitutions recording solely the note of acceptance or proposal were included in the Code, this appears to be due to lack of complete records of the Imperial legislation. 1 Consequently texts with solely accepta or proposita were added as a kind of emergency. In an effort to reconstruct the original text an acc. or p(ro)p(ositum) frequently appears to have been substituted by a dat. 2 Otherwise the guiding principle seems to have been that preferably the datum should be recorded rather than the acceptum or the propositum. Datum and propositum in the same text are clearly due to negligence. 3

(4) The fourth point emerging out of Seeck's examination is accordingly that a datum may be a substitution either for an acceptum or a propositum. This, however, applies only to constitutions where the place of the propositum is located in the administrative area of the addressee except in the case of high court officials. 4 In certain cases this obviously enables us to date a constitution a year later than suggested by the text. 5

Seeck regards the passages giving the time of the year (month and day) as more reliable than the consulates. Dubious points can usually be interpolated by ordinary philological methods. As the exact date of each constitution was required, we may expect to encounter some interpolations in this context, too. 6 The place of issue or of acceptance was, however, of secondary importance to the compilers. It is therefore little surprising that this part of the subscription in many cases has been omitted. On the other hand, we may assume the places recorded mainly to have been rendered correctly. Seeck regards them as by far the most trustworthy part of the subscription, 7 though mistakes do occur. Dubious points can usually be explained with recourse to philological analysis. Most cases suggest a confusion of common, frequently occurring names with little known localities of similarly sounding names.

It remains to say a few words about the preambles of the constitutions, consisting of two parts, (a) the name of the legislator, and (b) the addressee. The former must frequently have been supplied with the aid of the consular year noted in the subscription. The original wording may have been e.g. idem A or idem AA. This had to be altered in the text of the Code, and consequently this called for a conclusion of the computer. 1 Occasionally the names of the Emperors may aid us in identifying the Imperial consulships recorded in the subscriptions. 2

On the other hand, the second part of the preamble may be classified as one of the trustworthiest passages of the constitution, 3 though slight corruptions occur. This concerns the names, whereas particularly the abbreviations of the offices held by the addressees could have been easily distorted or confused. 4

(5) As a fifth point we may note that, (a) the dates of the Calendar year, (b) the geographical names, and (c) the names of the addressees are by far the most reliable parts of the passages of the constitutions important for dating purposes.

(6) The sixth point is that in dubious cases little heed should be paid to the titles of the addressees.

In the sequel the dates of certain Constantinian constitutions will be discussed with the explicit aim of establishing the Emperor's itinerary. This implies that only dated constitutions with an indication of the place of issue will be considered. In so far as revisions of the dates of certain constitutions appear necessary, the picture of Constantine's legislation should be drawn anew. This task without direct bearing on the chronological problems must, however, be reserved for future research.

The basis of the subsequent survey is the redating of Constantine's first Licinian war, proposed earlier by me, 5 further confirmed in the first part of these studies, and recently wholeheartedly accepted by Dr. Habicht.6 The revision of the date of this war affected the dates of certain constitutions, and Habicht went even further on the path suggested. Now the redating of certain con- stitutions 1 is likely to affect the dates of others. Seeck's Regesten are more or less a series of interpolations based on a brilliant analysis of the principles governing the conception of the Theodosian Code. Once a part of the Emperor's itinerary is adjusted, a chain reaction is started. The purpose of the following pages is to see how this chain reaction started by the redating of Civil War I affects our conception of the Constantinian Chronology as a whole.

The first brief part of our examination indicates, with reference to earlier studies that the travels by Seeck dated to 313 really belong to the year 314.

As a point of departure for our study of the dates of the Theodosian constitutions, we may take the fact that Constantine left Milan after the conference with Licinius in the spring of 313 to fight the German tribes on the Rhine frontier. 2

Curiously enough we have no proofs that Constantine took up residence at Trier until December 27, 313 (CTh XI 29, 1; 30, 1). 3 All the other edicts dated at Trier and by Seeck referred to the year 313 4 because, in his opinion, Constantine in 314 was campaigning against Licinius, really belong to 314, namely CTh VI 35, 1; IX 40, 1; XI 30, 2; 36, 1; I 12, 1. 3; VIII 10, 1; X 15, 1; XI 1, 2; 7, 1; 5 I 2, 1. 6 Nevertheless, despite the war operations on the frontier we may assume that Constantine made Trier his headquarters and center of the civil administration much earlier, and after the German wars we find him there late in December 313. His presence is subsequently attested in January 7 and on March 26, 314 (CTh III 30, 1). Towards the end of the year the Theodosian Code shows him to have been in Trier on October 29, November 3 and 8 and on December 30, 8 obviously for the consular procession on the New Year.

The following section purports to show that Constantine stayed in Gaul, using Trier as his headquarters until and past his initial decennial celebrations. The first part of the inquiry concerns the constitutions accepted by Seeck as showing Constantine staying elsewhere during this period. This will take us through the controversial years 314–5 to the Civil War in 316, and actually to 318 as there seems to be little disagreement with regard to the Emperor's movements in 317.

The rich and varied coinage of the year of Constantine's fourth consulship and of his decennial celebrations (315/316) suggests that Constantine remained in Gaul until the initial decennial celebrations, and a study of the Constantinian constitutions confirms our hypothesis. The presumed decennial celebrations at Rome prove to be based on an erroneous conjecture. Modern research has without hesitation accepted Mommsen's views as expressed in CIL I, p. 397 "Testibus legum subscriptionibus circa illud tempus in urbe mora- batur." The Theodosian Code, however, gives little support to his theory. The incorrect date of the CTh VIII 7, 1 has been demonstrated in other context. 1 The CTh II 30, 1 issued at Sirmium with the date Constantino A IIII et Licinio conss appears to be from the year 319; the correct reading certainly was Constantino A IIIII et Licinio Caes conss. 2 That Constantine should have stayed in Licinian territory in 315 is quite impossible.

The next decree in Seeck's Regesten is more controversial, CTh VIII 18, 1 "dat XV kal Aug Aquil, recitata aput Vettium Rufinum pu in senatu non Sept Constantino A V et Licinio C conss" Seeck 3 builds his chronological considerations on the fact that Vettius Rufinus was urban prefect in 315/316 and concludes that the dating must be wrong. The correct consular dating according to him was Constantino A et Licinio conss, i.e., 312, 313 or 315. By choosing the latter year he is forced to alter the place of issue, Aquileia to Aquaviva, some twenty miles outside Rome, because according to Mommsen's interpretation of the Calendar of Philo- calus, Constantine must have entered Rome either on July 18 or 21. The crucial point is the mention made of the urban prefect Vettius Rufinus. There were two contemporary Vettii Rufinii, the urban prefect of 315/316, C. Vettius Cossinius Rufinus, 1 and the Vettius Rufinus who was pretorian prefect in 319/320 2 and possibly consul in 323. The text is fairly corrupt. At least it varies in different manuscripts. The apparatus criticus records "apud uettium (uittium N, vectium OC, ricium G), rufinum (om. O), image (sic PMLS, image H, pfm E, image N, pfp C, image G, image B, om. O)." It is therefore very likely that the pu is corrupt, that the official in question was either the former urban prefect C. Vettius Cossinius Rufinus, in 319 holding some other office, or the pretorian prefect Vettius Rufinus, although the latter interpretation makes the words in senatu more difficult to understand. That the decree thus was issued on July 18, 318 or 319 in Aquileia appears extremely likely, preferably in 319, since Constantine's presence in Aquileia can be established for July 1 and 30.

Next we turn to the Frg. Vat. 274 of August 13. The subscription runs "data idibus augustis Romae Constantino et Licinio augg conss." As the consulships are unspecified, the theoretically possible years are 312, 313 and 315. 3 In 313 Constantine had left Italy, whereas we know that he stayed in Rome in 312. Thus our interpretation of the date of this fragment entirely depends on our interpretation of the general course of events.

Another fragment, Frg. Vat. 33, has the subscription "proposita idib Aug Romae Constantino et Licinio augg IIII conss." That this decree was issued not more than a few days before August 13, is clear. 4 The inconsistency of the Vatican fragments as regards the recording of consulates makes it very difficult to establish the correct dates. Seeck 5 therefore proposes a choice of any one of the years 312, 313, 315 and 319 (Constantino A V et Licinio Caes conss). Thus 312 appears to be a possible date also for this fragment.

Three more items in the Theodosian Code establish Constantine in Rome at 315, namely XI 30, 3 (August 25), I 2, 2 (propositum on August 29) and X 1, 1 (September 18). The dating appears to be in order. Accordingly we have to accept them as proofs that Constantine stayed at Rome at least during the time August 25 –September 18, 315.

Before concluding this negative chronological argument, we should consider evidence in favor of his staying elsewhere. That is, what evidence can be brought forward to show his movements in the early part of 315 ? Four dates can apparently be established, March 21 in Cabillunum, April 28, June 1 and August 3 in Trier. The date of the Emperor's visit to Cabillunum rests on CTh IX 40, 2. Seeck 1 refuses to accept the consulships Constantino A IIII et Licinio IIII as genuine because, according to him, Constantine was in Thessalonica at that time. Recently Dr. Habicht has shown the date recorded to be correct. 2

The date April 28, altered by Seeck to February 27, 316 3 has previously been restored by me 4 and the reading suggested was later accepted by Dr. Habicht. 5

The date June 1 is recorded in CTh XIII 5, 2, which gives the fourth consulships of Constantine and Licinius. Seeck, assuming Constantine to have been in Italy at the time, was forced to alter the year to 314. 6 Our revision of the Emperor's itinerary enables us to preserve the original dating.

August 3 is the date of CTh I 16, 1 recording the fourth consulates of Constantine and Licinius. Seeck 7 is compelled to alter the dating to Constantino et Licinio conss because of the supposed decennial celebrations in Rome at that time. Again there is no necessity to take such an extraordinary step, since the Emperor cannot be proved to have been elsewhere at that time. We may thus conclude that Constantine celebrated his initial decennalia at Trier on July 25 and travelled south in early August to be in Rome not later than August 25.

The dates of the Theodosian Code are highly controversial for the years 318–319. Of fifty-six decrees fifty-two have the dating of 319, Constantino A V et Licinio C conss, whereas only one is dated with Licinio V et Crispo C conss. 1 Seeck decided that two series of decrees had been confused and endeavoured to reconstruct the itinerary of the Emperor with the aid of addressees, acting under the assumption that the name of an addressee was less likely to be corrupt than the subscription.

The urban prefect Septimius Bassus, who held office from May 15, 317 to August 31, 319, is the addressee of several of the constitutions in question. This implies that no decree directed to Bassus in his capacity of praefectus urbi can be later than August 31, 319; thus we are compelled to alter the year of CTh III 17, 1, given at Aquileia, to 318. In actual fact, similar alterations of the dates of several other constitutions appear necessary; it would be unnatural if no decrees had been issued in 318. One particular case illustrates how a dating with a Licinian consulate might have been altered. CTh XI 29, 2 gives Constantino A V et Licinio C conss, whereas the corresponding CIust III 11, 3, from a different source than the present day version of CTh, records the probably correct Licinio V et Crispo. On the whole, however, the reconstruction of the Emperor's itinerary should be made warily, with due regard to the prosopographical implications.

Accepting the Emperor as staying in Aquileia on October 12, 318, we may also accept Seeck's theory concerning the dates and the places of issue of the following decrees: 2 CTh XI 30, 9 (June 22, Aquileia), XII 1, 6 (July 1, Aquileia), VIII 17, 2 (September 7, Milan). The three constitutions by Seeck recorded from the end of the year 318, CTh II 6, 2 (October 24, Sirmium), V 2, 1 (December 1, Sirmium) and XIV 25, 1 (December 12, Sirmium) create some difficulties if we maintain that Constantine celebrated his processus consularis in Aquileia on the New Year 319. Seeck's arrangement means that the Emperor had left Italy by the autumn 318 and was moving east. The crucial one of these three decrees is CTh XIV 25, 1, addressed to the proconsul Africae Aconius Catullinus. Here Seeck boldly has corrected the recorded consular dating Constantino A IIII et Licinio IIII conss to ipso aug cons. 1 Now the date in itself is no stumbling block, as the list of African proconsuls needs to be revised. The following dates for the early Constantinian proconsuls in Africa can be established:

  • Anulinus, beginning of April–October 31, 313. 2
  • Aelianus, February 15–November 8, 314. 3 We have at least five constitutions, obviously belonging to the same group and originally probably parts of the same law. 4 Of these five, namely CTh I 12, 1 (October 30), XI 1, 2; XI 7, 1 (both November 1); VIII 10, 1 and X 15, 1 (both November 8), the last but one is the only complete one, with the subscription dat VI id Nov Treviris, acc XV kal Mart Carthagine Constantino A IIII et Licinio IIII consulibus. The consular year, of course, refers to the acceptum, 5 and thus the day of issue was November 8, 314. As all these decrees originally belonged to the same law, the acceptum actually refers to them all, although the pertinent passage has been omitted in some of them. That the same consulates were recorded in all decrees is only natural. 6 The year of issue was thus, in all instances, 314. The latest decree recording Aelianus as proconsul Africae was of November 8, 314. Again, St. Augustine records him as holding the same office on February 15, 314. 7
  • Petronius Probianus, established as proconsul Africae on August 25, 315 (CTh XI 30, 3). On the second date recorded by Seeck 8 , August 13, 316, his title or office is not mentioned, and the two constitutions in question must therefore be disregarded in this context.
  • Aconius Catullinus. He is mentioned in three constitutions of November 3, 314 (CTh IX 40, 1 ; XI 30, 2; XI 36, 1) but by name only; they may therefore be disregarded here. He next appears in CTh XIV 25, 1 of December 12, 315, the decree mentioned above. The date appears to be in order, but the place of issue, Sirmium, is impossible at so early a juncture. This difficulty can be eliminated by reverting to Seeck's expedient in another context 1 of correcting Sirmio to Sirmione, Sirmio being located at the Garda lake.
  • Proculus, established in office as early as April 319, since the propositum in Carthage on May 7, 319 presupposes a date of issue in April (CTh XV 3, 1). He is further established on July 27, 319 (CTh VI 35, 2) although the consular dating appears to be corrupt. (Constantino A IIII et Licinio conss instead of Constantino A IIII I et Licinio C conss). Other decrees do not record his title and office.

For the sake of clarity it should be recorded that of the two constitutions mentioning Catullinus at a time when Proculus already had succeeded him, the first, CTh XI 16, 1 of August 27, 319, is a pro- positum; the date of issue, impossible to ascertain, must fall within Catullinus' tenure of the African proconsulate. The other decree, CIust VI 56, 3 of July 28, 320 is badly corrupt. Seeck incorrectly supplies proconsuli, although the mss. record praefecto and pp. He also alters the consular dating to ipso A et C conss although the ms. records Constantino A VII et Constante conss. 2 The most likely dating appears to be 326, the year of Constantine's seventh and Constantius' first consulship, Constante standing for a corrupt Constantio.

We may now conclude that Constantine spent the end of the year 319 in Sirmium ; there is no necessity to consider the two constitutions dated October 24 and directed to Felix, praeses Corsicae, as belonging to 318, for the only evidence of an official with that title rests with the constitutions addressed to Felix (CTh I 16, 3; II 6, 2) 3 . The consular year of the former is clearly corrupt, Constantino A IIII et Licinio IIII Caes cons instead of Constantino A IIIII et Licinio Caes cons, 4 whereas the latter has the correct consulships of 319. Finally, for the CTh V 2, 1 the consulships are correctly recorded, and the addressee, the pretorian prefect Rufinus is the first one recorded for the Gallic prefecture. 5 Nothing prevents us from accepting the subscription of the mss. rather than Seeck's conjecture.

With the Emperor's whereabouts for the end of 319 established, two points of the itinerary proposed by Seeck involve some difficulties, namely the two constitutions of July 25, 319, issued at Naissus (CTh II 15, 1 and II 16, 2). 1 The latter, directed ad Bassum is simple to deal with. The inference of Seeck and Mommsen-Meyer that this Bassus was the urban prefect of 317/319 is supported solely by the comments of the O(xonianus Bodleianus Seldenianus B 16) ms. which states "Ad Bassum intellege praefectum urbi." 2 As the ms. identifies Bassus with the urban prefect, the compiler necessarily must have altered the dating, for this is the only ms. giving the consulships as Constantino A IIII et Licinio conss, probably instead of Constantino A IIIII et Licinio C conss. In other mss. the consulships are given as Constantino A VIII et Constantio C, obviously the dating of 329 (Constantino A VIII et Constantino C IIII). Thus the real slip is confined to Constantio (pro Constantino C). Mommsen- Meyer, however, both in their comments and in their introduction 3 regard the consulships of the vast majority of the mss. (NBGEPMLS) as a later interpolation necessitated by the fact that CTh II 16, 1 has been securely dated to 326. The above discussion has shown that this interpretation cannot be valid, that the correct date of CTh II 16, 2 is 329, and that the Bassus in question was the pretorian prefect of Italy Iunius Bassus, 4 who succeeded Ablabius in 329. 5

The second constitution is directed "Ad Symmachum vicarium" (CTh II 15, 1). On formal grounds no objections can be raised against it: a Symmachus is known from February 4, 318 (CTh II 4, 6 1), and this decree is cited in another of October 12 of the same year (CTh III 17, 1). The dating appears to be correct, too. Nevertheless, as we are compelled to move CTh II 16, 2 to 329, the same should be done with CTh II 15, 1, because it would be highly unlikely that the Emperor should have stayed twice on exactly the same day in such a small township as Naissus (despite the fact that it was his birthplace; on the other hand, Constantine does not seem to have attached much importance to this. Other cities were renamed after him or after members of his family, Cirta-Constantina in Africa, Arelate-Con- stantina in Gaul, Byzantium - Constantinopolis, Helenopolis, the new foundation at the Propontis, but not Naissus), and because Constantine, even regardless of these two decrees is known to have been in Naissus in 329 (CTh XI 27, 1 of May 13). 1 In addition it should be mentioned that Symmachus by 329 must have been a man of considerable standing; he was one of the consuls of 330. The character of his office, by CTh II 15, 1 recorded with vicarius only, is uncertain. The necessity to alter the Imperial consulships recorded is no serious obstacle for our arrangement, particularly if one considers the unreliability of the records concerning the consulates of the Imperial Houses.

It remains to discuss the CTh IX 37, 1 issued at Serdica on November 26, addressed to a certain Ianuarinus. First of all the proper dates should be restored to the CTh V 10, 1 of August 18, and the CTh XII 1, 16 of September 29, both issued at Serdica. The consulates recorded are correct for 329 except for the slight error Constantino A VIIII et Constantio IIII conss instead of Constantino A VIII et Constantino C IIII. 2 Formally the date of CTh IX 37, 1 is in order, but physically it is impossible for the Emperor to have been in Sirmium on October 24, 319 (CTh I 16, 3), in Serdica on November 26 and anew in Sirmium on December 1 (CTh V 2, 1). This must be explained as a case of an omitted postconsulate. 3 This would give us the year 320, the end of which the Emperor spent in this important city. The fact that we have a decree (CTh IX 34, 3) addressed to the same Ianuarinus and posted on December 4, 320 in Rome (thus issued in November ?) makes the likelihood of this arrangement even greater. Furthermore, both deal with actions at law, CTh IX 34, 3 with defamatory writings and CTh IX 37, 1 with annulments of suits.

The constitutions discussed show that Constantine spent the entire year 320 in Serdica except on two possible occasions, May 22 (CTh IX 1, 5) and July 22 (CTh XV 1, 4). Now the consular dating of the former is Constantino A VII et Constantio C conss, the correct formula of 326. Although the years 320 and 326 are easily confused 1 , there seems to be no reason for an alteration in this case. It is true that the editions of CTh record the addressee as Maximus p(raefectus) u(rbi), and that a Maximus (Valerius Maximus Basilius) held that office from September 1, 319 to September 12, 323. On the other hand, two generally reliable mss., B and N, 2 record the office as P(raefectus) p(raetori)o. This reading has not been accepted though it offers a far easier solution of the problem of the decree than either the acceptance of the date or the insertion of a p(ost) c(onsulatum) in the subscription. As, however, the decree was accepted, and accordingly posted in Rome, this implies that the pretorian prefect must have been the administrative head of Italy. According to Seeck 3 such was not the case, since starting from the appointment of Rufinus (from December 1, 318 according to Seeck, from the same day in 319 according to me) 4 we get pretorian prefects in Africa, Italy and Gaul. 5 This, however, appears to be a misinterpretation of the evidence although the persons referred to by Seeck must have been officials of considerable standing. Neither Menandros (June 22, 320–July 6, 322, in Africa), nor Acilius Severas (December 18, 322–January 23, 324, in Gaul), nor Volusianus (August 1, 321, in Italy) is addressed as pretorian prefect in any constitution in CTh. The only official referred to in that capacity is actually Vettius Rufinus, and Seeck justly remarks that his tenure of office appears to coincide with Crispus' first command in Gaul. 6 Seeck argues further that the custom of addressing a decree to a pretorian prefect became necessary when the number of prefectures was increased and the holders were acting far away from the court. As long as there had been one prefect only there had been no need to direct decrees expressly to him. Obviously Vettius Rufinus created a precedent, and when after the defeat of Licinius the East fell to Constantine, and the pretorian prefects necessarily must have been two, 7 the custom adopted during his tenure of office was developed further. Ulpius Maximus appears to have been the Western prefect in charge both of Gaul and Italy, and this accounts for the decree having been posted in Rome, although Maximus later appears solely in his capacity of praefectus praetorio Galliarum.

The second constitution mentioned (CTh XV 1, 4) is more easily dealt with. Despite the physical impossibility that Constantine stayed in Sirmium on July 22, 320, it is more natural to regard CTh XV 1, 4 as a case of an omitted postconsulate and date it as 321. The subscription is also otherwise corrupt with Constantino A IIII et Constantio C conss instead of Constantio A VI et Constantino C conss.

For the year 326 Seeck 1 records two constitutions (CTh IX 24, 1 of April 1, and CTh IX 8, 1) which, if correctly dated would be incompatible with the decree CTh IX 1, 5 addressed to Ulpius Maximus, considered above. The addressee of the latter is Bassus, vicarius Italiae, this being the only constitution where he is referred to in this capacity. This Bassus must be Iunius Bassus, praefectus praetorio Italiae in 329 and consul in 331. 2 The only evidence that he was vicarius Italiae is CTh IX 8, 1. Thus the other decrees mentioning a Bassus (possibly Septimius Bassus, the former urban prefect) do not prevent us from redating this decree as 318, when Constantine's presence in Aquileia is otherwise firmly attested. The list of vicarii Italiae as recorded by Seeck 3 is headed by Iulius Severus, first in evidence on June 27, 318. That Iunius Bassus was the first vicarius Italiae is thus very likely. This arrangement implies that the subscription of CTh IX 8, 1 is restored as Licinio A V et Crispo C conss from the Constantino A VII et Constantio Caes conss in the present day editions.

The unreliability of the Imperial consulships as recorded by CTh has been demonstrated fully before and stressed by all editors. An extract of the apparatus criticus of the Mommsen-Meyer edition goes to show how much at variance the textual tradition is: "Constantino (constantio PM, const. C ) a (om. H) uii (i G ) et constantio (sic VBEC, Constantino NGPMS, constante HO) caes (sic VHN, om. reliqui) libri."

The second Aquileian constitution CTh IX 24, 1 is closely connected with the one discussed above, both with regard to content, place and date of issue. The proper year is accordingly 318.

With regard to the constitutions of later years we may in passing note that the proper year for CTh XII 5, 1 1 is 327, the constitution being probably a case of an omitted postconsulate, and restore CTh XI 7, 4 and IX 12, 2 to their proper dates, those indicated in the subscriptions, i.e. the years 327 and 326, respectively. 2 Similarly the date of CTh VI 1, 4, by the manuscript given as June 27, 317, can be maintained despite Seeck's attempts to change it to 330.

Reconstructing the Emperor's itinerary we arrive at the following results:

314 March 26, Trier CTh III 30, 1
October 29, Trier CTh VI 35, 1
October 30, Trier CTh I 12, 1
November 1, Trier CTh XI 1, 2; 7, 1
November 8, Trier CTh VIII 10, 1; X 15, 1
315 March 21, Cabillunum CTh IX 40, 2
April 28, Trier Optatus Milev. ed. Ziwsa p. 212
June 1, Trier CTh XIII 5, 2; XI 30, 4
August 3, Trier CTh I 16, 1
August 25, Rome CTh XI 30, 3
September 13, Rome CTh X 1, 1
October 19, Milan Frg. Vat. 273 3
December 12, Sirmio CTh XIV 25, 1
316 January 11, Trier CTh I 22, 1
May 6, Vienna CTh II 6, 1
August 13, Arelate CTh XI 30, 5. 6
September 20, Verona Frg. Vat. 290
October 8, Cibalae 4
Campus Mardiensis
December 4, Serdica CTh IX 1, 1
317 March 1, Serdica Chron. I p. 232, Anon. Val. 5, 19
March 8, Thessalonica CTh VIII 7, 1
April 17, Serdica CTh VIII 12, 2; IX 10, 1
June 6, Sirmium CTh XI 30, 7
June 27, Thessalonica CTh VI 1, 4
318 April 1, Aquileia CTh IX 24, 1
April 4, Aquileia CTh IX 8, 1
June 22, Aquileia CTh XI 30, 9
July 1, Aquileia CTh XII 1, 6
July 18, Aquileia CTh VIII 18, 1
July 30, Aquileia CTh VII 22, 2 1
September 7, Milan CTh VIII 18, 2
October 12, Aquileia CTh III 17, 1
319 March 11, Sirmium CTh X 8, 2
April 13, Sirmium CTh II 19, 1
April 27, Sirmium CTh VI 35, 3
June 2, Sirmium CTh II 30, 1
October 24, Sirmium CTh I 16, 3; II 6, 2
December 1, Sirmium CTh V 2, 1
320 January 31, Serdica CTh III 2, 1; IV 12, 3; VIII 16, 1; XI 7, 3; CIust VI 9, 9; VI 23, 15; VI 37, 21
May 19, Serdica CTh X 1, 4
June 30, Serdica CTh IX 3, 1
November 26, Serdica CTh IX 37, 1
December 17, Serdica CTh XVI 10, 1
321 February 6, Serdica CTh II 19, 2
February 27, Serdica CTh IX 42, 1
April 11, Sirmium CTh XV 1, 2
April 17, Sirmium CTh XI 19, 1; CIust XI 62, 2
May 27, Viminacium CIust VIII 10, 6
June 12, Sirmium CTh II 18, 1 ; XI 30, 11 (uncertain)
September 14, Sirmium CTh I 4, 1; IX 43, 1
322 May 23, Sirmium CTh II 4, 2; II 18, 2
June 12, Sirmium CTh IV 8, 4

Campona, defeats the Sarmats, Porph. Opt. VI 18, Zos. II 21

Margus, C. returns to Roman territory, Porph. Opt. VI 23

Bononia, C. divides the spoils, Porph. Opt. VI 26

July 6, Bononia CTh XI 27, 2
July 20, Sirmium CTh IV 8, 5
July 26, Savaria CTh I 1, 1
December 18, Serdica CTh III 32, 1
323 February 15, Thessalonica CTh IV 8, 6
April 13, Byzantium CTh XI 30, 12; XII 1, 8
May 25, Sirmium CTh XVI 2, 5
324 January 19, Sirmium CTh XII 17
January 23, Sirmium CTh VI 22, 1
March 8, Thessalonica CTh XIII 5, 4
April 9, Thessalonica CTh II 17, 1
July 3, Hadrianople, Licinius defeated, p. 232 CTh VII 20, 1, Chron. I
Byzantium, siege of September 18, Chrysopolis
325 February 25, Nicomedia CTh I 15, 1
May 20, Nicaea, opening of Council, Socrat. I 13, 13
May 23, Nicaea CTh I 2, 5; II 18, 3
July 25, Nicomedia, vicennalia, Hier. Chron. 2342, Eus. V.C. III 15
August 11, Nicomedia CIust VI 21, 15
September 17, Nassetis CTh XI 39, 1
October 19, Aquae CTh VII 4, 1
326 February 3, Heraclea CTh IX 3, 2; 7, 1
March 5, Heraclea CTh X 4, 1
March 8, Constantinople CTh II 10, 4
April 18, Sirmium CTh IX 12, 2
May 22, Sirmium CTh IX 1, 5
July 18, Rome, C. enters the capital, CIL2 p. 268 1
July, Rome CTh XV 14, 3 1
July 25, Rome, vicennial celebrations, Hier. Chron. 2342, Chron. I p. 232
September 25, Spoleto CTh XVI 5, 2
September 27, Rome, Profectio augusti
October 23, Milan CTh IV 22, 1
December 31, Sirmium CTh III 32, 2; X 1, 5 = CIust V 71, 18; VII 62, 17; X 1, 7
327 February 27, Thessalonica CTh XI 3, 2
May 18, Serdica CTh XI 7, 4
June n, Constantinople CTh II 24, 2
July 30, Nicomedia CTh XII 5, 1
328 March 1, Nicomedia CTh XIV 24
July 5, Oescus CTh VI 35, 5
September 27, Trier CTh I 4, 2
December 29, Trier CTh I 16, 4; VII 20, 5
329 March 9, Sirmium CTh VI 4, 1 2
March 15, Sirmium CTh II 16, 1 ; III 30, 3; CIust IV 32, 25; V 72, 43
May 13, Naissus CTh XI 27, 1
May 29, Serdica CTh IX 9, 1
July 25, Naissus CTh II 15, 1; II 16, 2
August 3, Heraclea CTh XI 30, 13
August 18, Serdica CTh V 10, 1
September 29, Serdica CTh XII 1, 16
October 13, Constantinople CTh VII 20, 3
October 25, Heraclea CTh XII 1, 17

For the remaining years of Constantine's life Constantinople becomes the center of his activities and there is no need to pursue the inquiry further. Any problems arising from the dates of the Theodosian Code will be discussed in connection with the coins.

End Notes

1 Occasionally Imperial constitutions have been preserved elsewhere, in inscriptions, in literary sources and the like. A part of the constitutions earlier published in the Theodosian Code are now preserved in the Codex Iustinianus.
2 But even when the places coincide, two weeks could pass from datum to acceptum, and a further fortnight to propositum (Seeck, Regesten, p. 10). Even a period of eight months between datum and propositum seems acceptable to Seeck (ibid., p. 58) in extraordinary cases.
3 Cf. pp. CCIXff. Tempora et loci.
4 Op. cit., p. CCXIII.
1 Regesten der Kaiser und Päpste, Stuttgart, 1919.
2 Seeck, p. 18. CTh I 1, 1 makes any undated edict or constitution null and void, cf. Wieacker, Textstufen klassischer Juristen, p. 39, also note 68.
3 Seeck, loc. cit. (Consularia Constantinopolitana as recorded by Hydatius).
4 Seeck, p. 21.
5 Ibid.
1 Ibid.
2 Seeck, p. 22.
3 For the Valentinian period Seeck notes that of 33 constitutions the iteration is proved to be correct in 11 cases, possibly but not probably correct in 8 cases and shown to be wrong in 14 cases (p. 36). For the sons of Constantine the situation is even worse.
4 Seeck, e.g. p. 69.
5 Chronica Minora I, p. 67.
1 Seeck, pp. 65f., cf. also p. 1 "Die große Mehrzahl der Datierungen scheint allerdings richtig zu sein; dock mischen sich darunter so viele erweislich falsche (im ersten Buck kommen auf 142 Fragmente 36 falsch edierte…), dass Mommsen zuletzt keine einzige für sicker begtaubigt und jede beliebige Anderung für ertaubt erhalten hat (Gesammelte Schriften II, p. 401)."
2 For various reasons. A good example is given by Kent, "Pattern," pp. 24f. who records the different consular reckonings in the dominions of Constantine, Maxentius and Galerius-Daza-Licinius, resp., in the years 306–313.
3 Seeck, pp. 66, 74ff., cf. also p. 22. This was done to avoid confusion in the legislation. The postconsulate (in the subscription noted as p. c.) was all too easily overlooked. The compilers therefore tried to restore the dating without recourse to postconsulates "mit derselben Inkonsequenz und Lieder- lichkeit, die wir auch sonst bei ihnen beobachten" (Seeck, p. 66).
4 Seeck, pp. 68ff.
5 Cf. e.g. Seeck, pp. 68f., 155.
6 Cf. e.g. Seeck, pp. 76, 155.
7 E. g. Seeck, p. 71, lines 33ff.
1 The archives of Constantinople obviously were the foremost source, but particularly the selection of the material of early date shows that a considerable number of the consitutions were not to be found in the capital. Material from other places must therefore be brought together, and the compilers in some instances most certainly had to be satisfied with incomplete constitutions (cf. Seeck, op. 2ff.).
2 Seeck, p. 80.
3 Seeck, p. 81.
4 Seeck, p. 80.
5 Seeck, pp. 82ff.
6 Seeck, pp. 96 ff.
7 P. 106.
1 Seeck, pp. mf.
2 Seeck, p. 112.
3 Ibid.
4 Seeck, pp. 113ff., e.g. p(raefectus) p(raetori)o, p(rae)p(ositus) (alternatively pp = praefectus praetorio), p(raefectus) u(rbis) and proc(onsul). These frequently occurring titles were occasionally inserted instead of other more unusual titles (Seeck, pp. 115ff.).
5 Cf. my Arelate pp. 17ff., particularly p. 18, note 5.
6 Hermes 1958, pp. 367–370.
1 With regard to those discussed in Arelate, it was mostly a case of maintaining (or, more accurately, restoring) the dates recorded in the subscriptions.
2 Seeck, Regesten, p. 161 referring to Eumenes Paneg. XII (IX) 21,5.
3 Seeck, Regesten, pp. 78, 162. Obviously we have to accept Seeck's insertion of the postconsulate before the consular years unless we prefer Mommsen's suggestion that the date is related to the datum of the edict.
4 Loc. cit., pp. 161f.
5 Habicht, Hermes 1958, p. 369.
6 Habicht, ibid., p. 368, note 1.
7 The propositum of January 22 (CIust VII 32, 10) cannot have followed very long after the issuing of the decree.
8 All the constitutions mentioned above, by Seeck referred to 313.
1 Cf. my Arelate, p. 18, particularly note 5, Habicht, op. cit., pp. 368, 370, note 1.
2 Habicht, op. cit., pp. 368, 370, note 1.
3 Regesten, pp. 59, 108, 163.
1 Schuurmans, De Samenstelling van den Romeinschen Senaat in de IVe Eeuw n. Chr. Diss., Gent 1943 (in typescript only), no. 776.
2 Schuurmans, no. 777, cf. Seeck, Regesten, p. 143. As to the corrected date of CTh V 2, 1, according to Seeck (p. 167) establishing him as pretorian prefect already in 318, cf. below p. 39.
3 Regesten, p. 50.
4 Regesten, pp. 8f.
5 Regesten, p. 23.
1 Regesten, pp. 68, 164.
2 Op. cit., pp. 368f.
3 Regesten, pp. 142f.
4 Arelate, p. 18, note 5.
5 Op. cit., pp. 367f.
6 Regesten, p. 98.
7 Regesten, pp. 55, 161.
1 Regesten, p. 56.
2 Regesten, p. 56.
1 Regesten, p. 167, cf. also p. 56.
2 Regesten, Index p. 477.
3 Not November 8, 313—April 26, 315 as recorded by Seeck, ibid.
4 Regesten, pp. 78f.
5 Regesten, ibid., Habicht, p. 369.
6 As to the slip for CTh I 12, 1, cf. the comments in the Mommsen-Meyer edition of the Theodosian Code.
7 Cf. Mommsen-Meyer, comments to CTh I 12, 1.
8 Regesten, Index p. 477.
1 Regesten, pp. 55, 160 when discussing the CTh VII 22, 1.
2 Krueger suggests Constantino A IIII et Licinio IIII.
3 Cf. Mommsen-Meyer, Introduction, p. CXCIX.
4 So corrected by Mommsen-Meyer.
5 Regesten, p. 143.
1 Regesten, p. 168.
2 As to the general reliability of the O ms., cf. Mommsen-Meyer, pp. LXVIf.
3 Cf. pp. CXXXIIf.
4 Schuurmans, no. 120.
5 Regesten, Index p. 474.
6 Regesten, pp. 57, 166.
1 Regesten, pp. 54, 179.
2 Regesten, pp. 65, 168f.
3 As to postconsulates, cf. Regesten, pp. 66ff.
1 Mommsen-Meyer, Introduction, pp. CCXIV, CCXVIII.
2 Particularly the latter, Mommsen-Meyer, pp. CXXXIIIf.
3 Regesten, pp. 143f.
4 Cf. p. 39 above.
5 Regesten, Index pp. 473f.
6 Regesten, p. 143.
7 As to the praefectus praetorio Orientis, cf. Regesten, pp. 144f.
1 Regesten, p. 176
2 Schuurmans, nos. 119–120.
3 Regesten, Index p. 479.
1 By Seeck published twice, Regesten, pp. 175, 178.
2 Seeck gives the years 328 and 329.
3 Possibly 312.
4 My Arelate, pp. 17ff.
1 Possibly 326.
1 Cf. Mommsen-Meyer, p. CCXVIII.
1 Cf. Mommsen-Meyer, comments to this decree.
2 Possibly 320.


1. From Trier to Cibalae

If we approach the chronological problem from the numismatic angle, the mint of Trier, of course, comes first. As Trier was the only mint to strike gold for Constantine in the five first years of his reign, we have little reason to go into details of internal chronology. Suffice it to say that the earliest group comprised aurei marked TR depicting Constantine as Caesar, the second solidi marked PTR showing Constantine as Augustus. The latter group was struck in the years immediately preceding the Italian campaign. 1 This group, homogeneous in point of style and fabric, can chronologically be divided into two parts, the earlier with the obverse break CONSTAN- TINVS PF AVG, the latter with the break CONSTANTI-NVS PF AVG. 2 The change of break can be dated as Constantine's quinquennial year, March 31, 311/312. 3 The solidi struck with the obverse break CONSTANTI-NVS PF AVG were:

  • VOTIS V MVLTIS X (Plate I, 313)

The next Treveran group of gold coins is different in many respects, the flan larger and thinner, the relief flat, the portrait broad and squarish. To this group belong the dateable gold pieces commemorating both the fourth consulate and the decennalia of the Emperor, although certain types appear to have been struck before these years (315–July 316). However, before dealing with this group of coins it seems appropriate to study the early Italian gold coins struck by Constantine.

The influence of Treveran prototypes upon Italian issues has been noted by previous research. 1 The resemblance is one of types as well as of fabric. The first Italian solidi all had thick and small flans. This is attested by the coins of Ostia, struck not later than the spring 313. 2 The reverses of the small module solidi were:

  • VBIQVE VICTORES (Plate I, 205)
  • VICTORIA CONSTANTINI AVG (two types, Plate I, 228)
  • Ticinum: PERPETVA VIRTVS AVG (124)
  • VICTORIAE LAETAE PRINC PERP/VOT X or PR (275, Plate I, 276).

We note that all the types of Rome and Ostia had Treveran prototypes, 1 whereas the types of Ticinum were new creations, except for the Victoriae laetae. Their early date is attested by the fact that both the Victoria Constantini aug of Rome and the Principi iuventutis of Ostia have obverses of Daza also. The parallelism of types indicates that the coins were struck in accordance with models brought to Italy from Trier. As most of the types are very scarce — of some of them a single specimen only is known—it is impossible to draw any conclusions on the basis of the absence of some of the Treveran types. The coin with the vot x (suscepta) proves the date at least of the vota coin itself, i.e., prior to the expiration of the quinquennalia (March 31, 312). 2 Again, the Ticinese coinage shows that the vota cannot be regarded as soluta, i.e., the type cannot have been struck for the decennalia in 315/316. 3

This survey suggests that Rome and Ostia were the first Italian mints to coin Constantinian gold, that the first solidi in these mints were struck before the end of March 312 4 , and that all types were reproductions of types previously struck at Trier. The first solidi of Ticinum were struck later, possibly in connection with the conference at Milan. The magnificent gold multiple announcing the Adventus (64) of the two Augusti points to this date. The new types demonstrate that the die-cutters could work at ease, that the authorities responsible for the medallic propaganda had had an opportunity to adapt the imagery according to the requirements of the day. Of the Ticinese Victoriae laetae with either vot x or vot PR the former variety could be regarded as an instance of "anachronistic" reverses; the type could have been brought to Ticinum from Rome to be mechanically and thoughtlessly copied at first until it was refashioned with the insertion of the vot PR. 1

Thus at this early stage of Constantine's career we can see how the gold minting followed in the Emperor's tracks, from Trier to Rome and Ostia, and finally to Ticinum. 2

The session at Milan was not a long one. Germanic tribes threatened the Rhine frontier and Constantine left Italy in the Spring 313. Uncertainty prevails as to the way he chose when travelling to Gaul and the time of his arrival in the Gallic capital. It is tempting to assume that he journeyed by way of Southern Gaul, i.e. Arelate, and to regard the first solidi of the mint of Arelate as struck during a visit of the Emperor to the mint city, but, as Constantine's presence in Milan is attested for March 10, 313, his visit to Arelate would have taken place in the end of March or the beginning of April. 3 Conceivably Daza could have been eliminated from the coinage at this juncture, but, nevertheless, it is difficult to believe that the transfer of the Ostian mint was carried out so rapidly. 1 A possible alternate interpretation regarding the early solidi of small module from Arelate is that Constantine slightly later, the same year or early 314, paid a visit to this city and that the unusual reverse types PRINCIPIS PROVIDENTISSIMI-SAPIENTIA (164) and the varieties of VIRTVS AVGVSTI (289) were struck then. 2 The unique character of the reverses cannot, I think, be explained except by assuming an Imperial visit to the mint city to have been the reason for creating them.

The return of the Emperor to Trier in 313 brought about a new period of gold coining in the Gallic capital. We cannot tell exactly when he took up residence in Trier, nor when the coining started. It is quite possible that this happened only after the Francian war. This is suggested by the rich array of consular coins and decennial types. Except for Ticinum, no other mint struck anything corresponding. Thus the coins corroborate the results of our study of the Theodosian Code that Constantine celebrated his decennalia for the first time in Trier, and travelled to Rome only after August 3, 315.

Constantine's stay at Rome in 315 appears to have been of very short duration, 3 at any rate not long enough to make it worth while to bring the court and the administration down to the old capital. The absence of gold coins struck at Rome would otherwise be inexplicable. His main reason for visiting Rome might have been the inauguration of his triumphal arch.

Instead of Rome Ticinum became his residence, a fact amply attested by the rich issue of gold coinage. Comparing the Ticinese gold types with the earlier Treveran ones, we find two different groups:

  • of identical reverse type and legend
  • FIDES EXERCITVS (Plate III, 80–81)
  • PAX AETERNA AVG N (Plate III, 121–122)
  • VICTOR OMNIVM GENTIVM (Emp. with standard, 206–207)
  • VICTORIA CONSTANTINI AVG (Vict. adv. l., captives, 230–231)
  • VIRTVS AVGVSTI N (Emp. mounted, 290–291)
  • of identical reverse type or legend Ticinum:
  • GAVDIVM ROMANORVM (Soldier dragging captive, 89)
  • VICTORE AVG N VOTIS X/MVL XX (Vict., trophy, Plate I, 212 ; 213; III, 214)
  • VIRTVS AVGVSTI N (Mars, captives, Plate III, 292)
  • Trier:
  • VICTORE AVG N/X/XX (216)
  • VOTA PVBLICA (308)

Arranging the Ticinese gold coins according to mintmarks and issues 2 we get a first series of gold comprising the multiple VICTORIAE LAETAE AVGG NN/VOT X MVL XX (Plate III, 273) and the following solidi, all marked SMT:

Felix processus cos IIII aug n, Fides exercitus, Gaudium Romanorum/Fran et Alam, Pax aeterna aug n, Pm trib p cos IIII pp procos, Restitutori libertatis (Emp. receiving globe from Roma), Victor omnium gentium (Emp. hold. standard, capt.), Victoria Constantini aug, Virtus augusti n (Emp. mounted), all struck in the later part of 315. The following issue, marked PT, is shown by a consular coin to have been struck within the same period. The PT coins were: Gaudium Romanorum[Franc et Alam, Pm trib p cos IIII pp procos, Restitutor libertatis (Emp. seat. w. parazonium, 167), Victore aug n votis/x mul xx. The two former types were struck with the well-known facing busts.

The next mark, ·Τ·, known from one type only, FELICIA TEMPORA (56), was probably struck for the New Year 316 1 both with a nimbate, facing portrait (without consular attributes) and with an ordinary laureate head. Next we get the S · Μ · T, dateable by a vota coin to the decennial year. The types were Adlocutio aug (1), Securitas perpetua (Plate III, 175), Victore aug n votis/x/mul xx.

The last group, closely connected with the previous ones, was marked SMT, like the first issue, and comprised the following types: Fortunae reduci (82), Gaudium Romanorum (Sold, dragg. capt. to Emp.), Liberalitas XI imp IIII cos ppp (120), 2 Rector totius orbis (165), Restitutor libertatis (Emp. seat. w. parazonium, 166), Soli comiti Constantini aug (193), Victor omnium gentium (Vict. crown. Emp., Plate III, 211), Victore aug n votis x/mul xx, Victorioso semper, Virtus augusti n (Mars adv. r.).

Analyzing this list of reverse types we find Treveran influence gradually weakening. All solidi of the first series had their counterparts in the earlier Treveran coinage, the second phase of coining introduces the new facing portrait and, for the old legend Restitutori libertatis, a new imagery. The symbolism of the few coins marked ·T· is new in the Constantinian coinage, as are two of the three types of the following issue. In the last series, the second SMT issue, only the legends remind us of the Treveran gold coinage, whereas a number of types were continued later in other mints.

The fact that no obverses of the Caesars appear in the Ticinese coinage during any of the series mentioned shows that they all were issued before the settlement at Serdica, probably even before the New Year 317.

End Notes

1 There is no real evidence for a solidus reform as early as 306/307 as Mrs. Alföldi contends (Trier, pp. 104f.). Her dating of the irregular 1/2 solidus PONT MAX PPP PROCS is not convincing nor the dating of the CONSE- CRATIO type (consecration coins of Chlorus of a much later date are known). Coining was not necessarily continuous (Alföldi, JRS 1932, pp. 10ff., Kent, "Pattern," p. 31) and the argument that the Emperor at a certain time must have needed hard cash is deceptive.
2 Cf. my "The Battle of the Milvian Bridge," Hermes 1960, pp.366f. and above p. 5.
3 Ibid. The question whether the natalis imperii was March 31, 307 or later the same year (Miss King, ANSMN IX, p. 130 remarks that Constantine could not have assumed the title of Augustus very late in 307) does not affect the argument.
1 Mrs. Alföldi, "Trier," p. 107.
2 Cf. p. 11, note 1 above.
1 Of the Iovi conservatori augg (117) a 1/2 solidus only was struck in Trier, the Ubique victores (204) can be regarded as the specifically Licinian type and was therefore not listed above. The prototype of the Victoriae laetae princ perp was struck in billon at Trier (but cf. note 1, p. 9 above).
2 The theory concerning the "anachronistic" reverse types discussed below (pp. 61f.) is not applicable to the Victoriae laetae of Rome. A new creation can never be anachronistic, it must necessarily reflect reality or fiction accepted as reality. Regardless of whether we accept the Victoriae laetae of Rome as a new creation or as a parellel to the billon coins of Trier, the vot x must refer to the quinquennalia; otherwise the vot PR of the billon type should have been maintained in Rome.
3 All decennial issues of Ticinum have large flans, and chronologically they were preceded by large flan consular coins recording the fourth consulate of Constantine. Alföldi (JRS 1932, p. 17) pays attention to this, but does not grasp the implications.
4 For details, cf. my "The Battle of the Milvian Bridge," pp. 366ff.
1 Alternatively, the vot x coins were struck in Ticinum before the close of the decennial year, but the scarcity of early Ticinese solidi makes this unlikely.
2 It is, of course fairly difficult to establish the exact time when gold minting was discontinued at Trier. I have earlier contended that this probably happened after the Italian campaign, i.e., that, e.g., the SPQR optimo principi and the vota coins with the later obv. break CONSTANTI-NVS PF AVG were struck after the battle of the Milvian Bridge, but before the expiry of the quinquennalia on March 31, 312. This is not necessarily correct, though there is no doubt as to the chronological significance of the two obv. breaks N-T and T-I. Gold coining possibly continued in Trier until the mint of Rome could start supplying Constantine with cash in that precious metal; there might even have been some overlapping as regards coining in these two mints. I cannot believe that the personnel of any one mint, nor even certain groups of workers in precious metals, were transferred to other mints as long as their original mint continued to work, except in special cases (cf. Mrs. Alföldi, Trier, p. 109, note 55). Otherwise it would be very difficult to account for the individual style of the mints.
3 Seeck, Regesten, pp. 160f. To this cf. above pp. 10f. It is impossible to fix any terminus post for the transfer. The necessary terminus post for the establishment of the mint of Arelate is the rupture between Constantine and Licinius, on the one hand and Daza on the other.
1 Cf. above p. 11, note 1.
2 The VIRTVS SAECVLI (Maurice II, pp. 152f., my Arelate, p. 61, no. 5a) is a misreading of VIRTVS AVGVSTI (307). The rev. leg. is partly obliterated, Dr. Bruck of the Kunsthistorisches Museum informs me, so as to allow the imaginative interpretation… SAECVLI.
3 It is, of course, possible that Constantine did not visit Rome at all this year and that the decrees mentioned above (pp. 35, 43) have incorrect dates.
1 Cohen, 577 records a solidus exactly similar to the Ticinese one in coll. Caylus. Not verified.
2 The first thorough analysis of the Constantinian gold of Ticinum was made by Andreas Alföldi (J RS 1932, pp. 16–23) who, however, missed the significance of the mintmarks.
1 Cf. Toynbee, Roman Medallions, p. 90.
2 This coin is the actual proof that Ticinum struck two different issues marked SMT. The terminus post is December 10, 316 when Constantine received his eleventh Imperial salutations (cf. Seston, REA 1937, p.218 and Laffranchi in Atti della Pont. Accad. R. d' archaeologia 1921).

2. From Cibalae to Chrysopolis

Gold coining at Ticinum was discontinued for the obvious reason that the court, the administrative center, owing to Constantine's successful campaign against Licinius, followed the Emperor north to Siscia. Here a scarce issue of gold, a 1½ solidus and an aureus of the reverse legend Soli invicto comiti (195), both marked ·SIS· reflect the Ticinese Soli comiti constantini aug (193) as, indeed, does the nimbate and facing bust of the aureus. 1 Here the solidi Victoria Constantini aug (Plate IV, 232) and Virtus exercitus Gall (Plate IV, 305) obviously were worked out according to models from Ticinum.2 The very earliest Siscian solidus, however, appears to be the hitherto unpublished GAVDIVM ROMANORVM/FRANCIA/ PS (Plate IV, 95), now in Berlin, of the well-known Treveran type showing Francia seated to l. on the ground with a trophy behind her back. The portrait enables us to interpret the exergual letters as P(ercussa) S(isciae). New creations are the 1½ solidi marked SIS, CRISPVS ET CONSTANTINVS IVN NOBB CAESS (47) and FELIX ADVENTVS AVG N (62), the latter at least suggesting a visit of the Emperor to Siscia in the early months of 317. The established dates for this period of time 3 make a stay at Siscia in January or February quite possible, and, in view of the numismatic evidence, even probable.

Before following the travels of the Emperor and the court any further, a few words should be devoted to the gold struck in Gaul at this time. Both Arelate and Treveri were coining, the former mint for a short period only. Two types common both for the West and for the East, originating in Ticinese solidi of the second SMT issue, illustrate how the central mint distributed prototypes as modified versions of the original reverses. The Virtus exercitus Gall (Virtus augusti n in Ticinum) was struck at Arelate (304), Trier (303), Siscia (Plate IV, 305) and Thessalonica (306), the Felicitas perpetua saeculi (Soli comiti Constantini aug in Ticinum ) at Arelate (59) and Thessalonica (Plate IV, 60). These two were the only later gold types of Arelate, whereas the Treveran gold coinage was more varied, and included also special types for the Caesars. Some of the types of 316–317 were clearly influenced by earlier Ticinese types. The chief Gallic mint, however, retained a very high degree of independence until the last years of Constantine's life, possibly owing to the fact that Crispus in 318 was given the praefectura Galliarum (after Civil War II Constantine II was in charge of Gaul). The mint of Trier can therefore be disregarded in the sequel.

To return to the Danube provinces, the court and Constantine can scarcely have stayed long in Siscia, and on March 1 the literary sources attest the Emperor's presence at Serdica for the reconciliation with Licinius. The peace treaty was confirmed by the nomination of the three Caesars, Crispus, Licinius II and Constantine II. Licinius gave up his European provinces except for the Thracian diocese. On the other hand, Constantine did not succeed in keeping all the territory he had conquered during the war. 1 An important result of Civil War I was that Licinius was forced to resign his legislative powers ; from now on the Eastern Augustus was restricted to distributing and enforcing the laws and edicts of Constantine. 2

Eight days later Constantine signed a letter to the consularis aquarum Versennius Fortunatus in Thessalonica.3 The subsequent rich gold issues of the Thessalonican mint suggest that the Macedonian capital was the imperial residence for some time.

Once more it is highly instructive to compare the reverse types of the mint in question with the types of the immediately preceding time, in this case struck at Ticinum.

The reverses of the first Constantinian gold issue of Thessalonica were (as recorded together with their Ticinese counterparts):

CLARITAS REIPVBLICAE (stg. l., kneeling captive, 7) 1
VICTORIAE LAETAE AVG N/VOT X MVL XX (Plate IV, 272) the same but AVGG NN (Plate III, 273)

Only a few specimens are known of each type. It is possible that the correspondence between the Ticinese and Thessalonican types was even more complete than the list suggests. Clearly here is another case of models brought from one mint to another. The two vota types are of particular interest, partly because both were struck after the actual decennial year, partly because of the significant singular aug. Ticinum struck the Victoriae laetae type with the plural augg nn, whereas Thessalonica used aug n, thus showing that the legend must have been altered before the peace treaty of 317. The "anachronism" of a decennial type being struck after the decennial year is explained by the celerity employed when starting the Constantinian issues at Thessalonica.

From Thessalonica the central mint moved to Aquileia. The exact time is in doubt, but although gold issues of the years 319–322 are known from Ticinum also and from Sirmium, the fact that the Emperor's fifth consulate (in 319) was celebrated in Aquileia only (73), makes it a fair inference that the consular coin was struck for the New Year 319. An Adventus coin of the same mint (2) implies that the processus consularis that year took place in Aquileia.1 The heavy multiple FELICIA/TEMPORA with the exceptional mintmark MAQ (57) expressing a New Year's wish 2 must have been struck for the New Year 319 when Licinius II entered his first consulship.

The itinerary of the Emperor has been established above. 3 Generally speaking, my survey of the Imperial constitutions has shown that the Emperor spent almost all the year 318 in Aquileia, 319 in Sirmium and 320 in Serdica. The last Serdican decree (of this period) was issued in February 321. On April 4 his presence is established in Sirmium, and in Sirmium he appears to have remained until July 20, 322 with a necessary break for the Sarmatian War. After the war he returned to his residence, but in the autumn he starts moving; in December he is in Serdica, in February 323 in Thessalonica preparing the Gothic War, in April he violates Licinian territory when marching against his enemies. The victory must have been easily won, because on May 25 he is back in Sirmium. By now it is clear that an open war against Licinius is unavoidable, and the seriousness of the situation is stressed in a letter to Helpidius (CTh XVI 2, 5), in which Constantine threatens severe punishment to any official found guilty of such anti-Christian measures as evinced during Licinius' quindecennial celebrations (November 322). In early Spring 324 (after January 23) the Emperor moves south, making Thessalonica his military headquarters. War breaks out and in July we have the first encounter between the Emperors at Hadrianople.

Let us now view the coinage against the background of Constantine's itinerary. Constantine and, with him, the central administration stayed in Aquileia during all 318, in so far as the Emperor's movements during this year can be traced. Very likely gold coining started in Aquileia the same year, logically with the Adventus (2) type (the type Victoria Constantini aug (Plate IV, 236) constitutes the link to the preceding gold issues of Thessalonica) and comes to an end with the consular coin and the multiple honoring Licinius II, both issued for the New Year. Very soon after the celebrations of the New Year, Constantine left Italy for Sirmium. The central administration, except for the officials travelling in the Imperial suite, remained behind, and actually were transferred to Ticinum. It is quite likely that no gold was issued for a while; the new types to be struck during the following bout of coining had to be planned, designed and executed. Towards the end of the year (319) an impressive series of gold, solidi as well as multiples, was issued. Felix processus cos vi aug n (75) gives the date, New Year 320. The friendly relations with the fellow-ruler Licinius are eloquently illustrated by the multiple Iovi conservatori (114) with Licinian obverse, by the type Concordia augg nn with obverses of both Augusti and Crispus (Plate III, 8; 9–10), and by the plurality of Augusti stressed in reverse legends such as Virtus augg nn (288), Victorib augg et caess nn/ vot xx (PLATE III, 279). The vota coins recording the vota vicennalia suscepta (277) give us the terminus post for the end of this gold issue, July 25, 320. It may be noted that Sol still appears to be the tutelary god of the Emperor (the multiple Soli invicto comiti, 197, the solidus Soli comiti aug n, Plate IV, 189).

However, few older types were represented in the Ticinese issue of 320/321. The Principi iuventutis (Prince hold. spear and globe, 135) and the two Sol types both represent old type stock; the new types, however, were distributed farther from Ticinum. The Felix processus cos vi aug n was struck both in Aquileia (76) and Sirmium (74, 77), where the processus consularis of the year must have taken place. In addition, Aquileia received from Ticinum the types Soli comiti aug n (190), 1 Principi iuventutis (Prince hold. globe and spear, 136) and Victoriae perpetuae/vot xx (278). The Aquileian multiple Securitas perpetuae (178) has no actual counterpart in Ticinum as far as we know. In all probability it too was struck in Ticinum, although no specimen has survived to our days. 2

Constantine spent almost the whole year 320 in Serdica, probably reorganizing the frontier defense and preparing the Sarmatian War. To have a central administration in faraway Ticinum must have been a serious disadvantage, and in the later part of 320 the court moved north and made Sirmium its residence. According to a preconceived plan Constantine joined the court in Sirmium in early 321. About the same time, if not earlier for the consular procession at New Year, his sons, the Caesars Crispus and Constantine II had arrived for their quinquennial celebrations. Hence the medallion Felix adventus caess nn (65). The center of gravity of the Sirmian gold issues belongs to the year 321/322 when the Caesars celebrated their consulates, the Augustus his expiring quindecennalia (Plate V, 280) and the Caesars their quinquennalia (Plate V, 266; 267; 263; V, 264). The mint had, however, been opened well in advance of this festival year. The necessary arrangements had probably been made during Constantine's stay in Sirmium in 320. The first Sirmian gold coins are from this year (m. m. · SIRM ·) and reflect their Ticinese prototype Felix processus cos vi (74). 1 The next series of gold (m. m. SIRM) reflects considerable Ticinese influence with its types Victorib augg et caess nn/ vot xx (PLATE V, 280) Soli comiti aug n (Plate IV, 191) and Soli invicto comiti (196). 2

The gold series of 321 demonstrate convincingly that Sirmium now had become the center of the Constantinian Empire. Here, and here alone, the quinquennalia of the Caesars were celebrated, here the consulships, here the arrival of the Caesars were commemorated on the gold coins. The types are, generally speaking, new. However, Constantine himself retains the Sol symbolism though in a fairly watered down version. That we now have reached the first year of tension between East and West can be seen in the wording of the reverses, Victoria (218) or Virtus aug et caess (287) instead of the plurality of Augusti in Victorib augg et caess nn/vot xx. To a certain extent Ticinum makes its imprint felt (237, 238).

Sirmium was to remain the administrative center until Civil War II ; no mint is known to have issued coins in precious metal until 324. 1 The earliest gold issues of 324 are the consular types of Sirmium Felix processus cos iii (· SIRM ·, 70) hailing the consulships of the young Western Caesars. The series of gold initiated with this consular type is of exceptional interest. The first series marked · SIRM · (comprising also two varieties of Gloria Romanorum, 103–4, and the type Perpetua felicitas, 123), must have been struck to the New Year, whereas the second marked SIRM in addition comprised the first type for Constantius II (145) nominated Caesar on November 8, 324, and the Augustae Helena (182) and Fausta (198), and further the first diademed upward-looking portrait of Constantine (Constantinus aug, Vict. seated on throne, 19). 2

From Sirmium instructions were sent to the other mints required for the gold coining. Obviously Aquileia by now played a very modest part in the economic life of the Empire. Two types only are known from the last postwar issue, the outdated Concordia augg nn (11) and the Principia iuventutis (163) with an obverse of Constantius II. Ticinum still served the needs of the central part of the Empire and struck a more varied series with both multiples and solidi of the types of the Augustae, the Constantinus aug (20) and two varieties of Principi iuventutis (one with an obverse of Constantius II, 142, 146). Finally we have Thessalonica, the operational headquarters of Constantine, where no coins had been struck since 318. Here we can attest two new types, not known from any other contemporaneous issue, Victor omnium gentium (Plate IV, 208) and Virtus Caesarin (293).

End Notes

1 I believe that the actual Ticinese prototype was a Soli invicto comiti, although no specimen has survived to our day. The imagery of the Soli comiti Constantini aug is different from the Sol coins of Siscia.
2 The Ticinese counterpart of the latter was Virtus augusti n.
3 Cf. p. 43 above.
1 He had advanced as far as Philippopolis at an early stage of the war.
2 Habicht, p. 370, also note 2, contrary to e.g. Stein, Geschichte I, p. 145 and Vogt, Constantin der Grosse, p. 176.
3 CTh VIII 7,1, cf. Arelate, pp. 18f., particularly note 5, and Habicht, pp. 368, 370, note 1. The court probably travelled straight to Thessalonica from Siscia, although the Emperor went by way of Serdica.
1 Corresponds to the Siscian Soli invicto comiti. The prototype might therefore be of Ticinese origin although not preserved to our day.
2 In "Trier" = Victoribus augg nn votis/x/xx (285), of later date in Ticinum.
1 Consular coins may have been struck by other mints as well, but the Adventus piece must have a purely local bearing.
2 Toynbee, Roman Medallions, p. 90.
3 Cf. pp. 44f. above.
1 But we cannot exclude the possibility that the Aquileian rev. was older, deriving its origin from the Ticinese Soli comiti Constantini aug.
2 This imposes upon us the question whether Ticinum or Sirmium was the central mint at this point (the type Securitas perpetuae being known from Sirmium (179), but not from Ticinum). The wider range of reverse types and the obverses of Licinius, without parallels in other Constantinian mints, conclusively prove the Ticinese series to have started earlier than the Sirmian one, and to have been issued by the central mint.
1 An inaccurate repetition of the original Felix processus cos vi aug n.
2 The imagery is slightly different from that of Ticinum.

3. From Chrysopolis to the Inauguration of Constantinople

Only a study of the first Nicomedian gold series explains the exact mechanics of gold minting during the first postwar years. The relevant series are two : As there is no difference in the mintmarks, the obverses must decide the grouping of the individual coins. 1 The first series of coins basically comprised types employed earlier at Sirmium and Thessalonica. This applies particularly to the solidi. Among the multiples, however, certain modifications of earlier conceptions were introduced.

Although one new type was created, the 2-solidi piece Pietas augusti n (125) — with an obvious reference to the capture of Byzantium and with a suggestion of the great future in store for that city 2 — the common denominator of the first Nicomedian gold coins struck by Constantine is the fact that all reverses were "old" types. I believe that two different groups can be distinguished representing successive stages of coining, the first comprising slightly anachronistic types created several years earlier, the second comprising comparatively recent inventions. To the first group I would refer the multiples Crispus et Constantinus iun nobb caess (48), Felix adventus aug n (63), Securitas perpetuae (180), Victoria aug et caess nn (219), and the solidi Adventus augusti n (3), 3 Concordia augg nn (13), Victorib augg et caess nn (Plate V, 281), and Vota publica (309–10). The second group would have the multiples Crispus et Constantius nobb caess (51), Pietas augusti n (125), and the solidi Constantinus aug (Victory seat. on throne Plate V, 22), Principi iuventutis (148), Solus reipublicae (174), Securitas reipublice (185), Victor omnium gentium (Plate IV, 209) and Virtus Caesarin (294). Now the latter group of solidi is almost identical with the one struck at Thessalonica in 324 and consisting mainly of types created at Sirmium,4 whereas the former was composed of old stock, partly outdated. I therefore suggest that Constantine's expeditionary force had been equipped far in advance with coin models in order that coining could start as soon as the mint was captured. On the other hand, the civilian administration followed slowly in the tracks of the Emperor, and its arrival at Nicomedia changed the types brought forth by the army into new up-to-date ones. Thus it seems that the financial administration, having created and distributed the types of 324 followed the Emperor to Thessalonica and from there to Nicomedia.1

The following bout of gold coining at Nicomedia bears the stamp of independence. A series of new types is created, the multiples paying homage to the senatorial (Plate VI, 186) and equestrian estates (54) and to Roman glory (107), the solidi celebrating the vicennalia (248) of the Emperor and the courage of the Caesars (295, 300). But before studying their echo in the West we should follow the way of the older types of the East, to Cyzicus and Antioch.

Cyzicus struck no gold under Licinius, but immediately after Constantine's conquest of the East a short gold series marked SMKE was issued with (as far as our knowledge goes) the types Concordia augg nn (14), Principi iuventutis (obv. of Constantius II, 149) and Victorib augg et caess nn/vot xx (Plate V, 282), all of them of old stock, the first and the third clearly "anachronistic." This suggests that the coins were issued already in 324. The second gold issue of Cyzicus (marked SMK) cannot be of much later date, as the Principi iuventutis type (Prince hold. two standards, 150) seems to have disappeared finally about 326. The second reverse is the Constantinus aug (Victory advancing l. holding wreath, 28) known previously from a Ticinese mule (26).

The first small gold issue of Cyzicus gains in importance when compared with the three first Constantinian issues of Antioch.

The dominating characteristic of the gold coins of the three first Constantinian series of Antioch (marked SMAN, SMAN ·, and SMAN*) is that all reverse types are "old"; 1 one single type had been created as late as 324. The vast majority goes back to Sirmian prototypes of 321. It should also be noted that only one obverse of Constantius II is known.

The reverses of the SMAN series were:

Crispus et Constantinus nobb caess coss ii (2-solidus, Plate V, 49), Securitas perpetuae (2-solidus, 181), Felix processus cos vi aug n (aureus, 78, and solidus, Plate V, 79), Felix processus cos ii (obv. of Crispus, Plate V, 66), Principi iuventutis (Prince with spear and globe, 137, and Prince standing with three standards, Plate V, 143).

The SMAN · reverses were :

Felix prooessus (sic!) cos ii (67), 2 Principi iuventutis (spear and globe, Plate V, 138).

The SMAN* reverses were:

Adventus augusti n (Plate V, 4), Soli comiti aug n (192), Concordia augg nn (Plate V, 15), Principi iuventutis (144).

We may note that these coins correspond with their prototypes even in significant details such as unusual obverse legends (Principi iuventutis, 3-standards type: DN CONSTANTINVS IVN NOB CAES, 141, Plate V, 143) or busts (Concordia augg nn, Crispus with spear and shield, 12, Plate V, 15) or reverse variety (Principi iuventutis, 3-standards, 141, Plate V, 143, common to Sirmium and Antioch, whereas all later varieties of this theme depict the Prince as holding a vexillum). 1

The frequent "anachronisms" such as the commemoration of the second consulships of Crispus and Constantine II in the year the Caesars actually held their third consulates and the Concordia augg nn, show that gold coining was resumed as soon as possible after Licinius' defeat. 2 The Adventus coin assists greatly in explaining this phenomenon; I think it is almost certain that Constantine himself advanced as far as Antioch after the victory at Chrysopolis, and that the scanty gold issue of Cyzicus should be regarded as evidence for his travelling either to or from the East. Antioch was, after all, the capital of the East. A Church council convened in Antioch in December–January 3 (324–25) to prepare the council of the following summer. Possibly owing to interference on the part of the Emperor, plans were altered and, instead of the council of the Eastern church scheduled to meet at Ancyra, an ecumenical council was summoned at Nicaea. We have ample evidence of the interest the Emperor at this time took in matters concerning the Church. 4 Already for this reason a visit to Antioch might have appeared imperative to him. As a point of curiosity we may note that at approximately this time, the time of the Antiochene council, the last Sol type was issued, Soli corniti aug n. Subsequently, until the last years of Constantine's life no gold was struck at Antioch.

We now return to the regions of the Bosporus, to Constantine's initial vicennial celebrations on July 25, 325. For this occasion the mint of Nicomedia had created a number of special types, the majority of which, however, stressed less the Imperial anniversary than things Roman in general. But the vicennalia were celebrated twice, for the second time a year later in Rome. The Roman gold of 326 (marked SMR) is very scarce, an indication of the short duration of the Emperor's stay in the old capital. Only two gold types are known, namely Senatus (multiple, Plate VI, 187) and Victoria Constantini aug/vot xx (solidus, 247, and submultiple).

Now the execution of these Roman gold pieces is truly remarkable. We note particularly the heavenward gazing portraits and the unusual arrangement of the obverse legend with Constantinus along the left edge and aug under the chin. 1 One other mint had previously employed exactly the same technique, Thessalonica, on the Equis Romanus (Plate VI, 55) medallions, 2 which are closely connected with the Senatus type. Also in point of style the resemblance is striking, and I regard it as quite possible that Constantine when visiting Rome in 326 brought ready-made dies with him from Thessalonica. Again, the Nicomedian and Thessalonican multiples cannot be exactly contemporaneous; the Senatus aurei of Nicomedia (Plate VI, 186) have a plain diademed upward looking head, the corresponding 3-solidi piece of Thessalonica a consular bust (Plate VI, 188).

Thus Nicomedia struck for the vicennial year 325, Thessalonica for the consular year 326. The conclusion is evident ; on his way from Asia Minor to Rome Constantine and, with him, the travelling mint, i.e. the travelling officials of the central mint administration stayed for a while in Thessalonica in early 326.

Before we follow Constantine on his journey to the West we should deal with a short but important Thessalonican issue, exceptionally marked THES. A submultiple of the reverse Victoria Constantini aug/ vot xx (246) gives the general date, the vicennial year. Two other THES marked coins are known, the splendid multiples of the type Principi iuventutis (Prince hold. standard, behind another standard, 151). The fact that no obverse of Crispus is known must be accidental. Thus we must conclude that the THES issue was struck in the Winter 325/326 before the SMTS issue with the Senatus medallion.

Nicomedia for the year 326 created an entirely new set of types depicting all rulers with diademed upward looking portraits. The reverse simply records the name of the ruler in question, the obverses have no legend at all. The type struck for Constantine shows two interlaced wreaths (Plate VI, 23), the type designed for the Caesars, Crispus (44), Constantine II (35) and Constantius II (40), Victory advancing left, holding wreath and palm branch. This set of gold coins was probably issued at New Year at Nicomedia and, it seems, at Thessalonica (Plate VI, 36; 41) during the Emperor's short stay in the Macedonian capital. 1 With the Emperor the types travelled West and were issued first at Sirmium (Plate VI, 24; 45, 37, 42), 2 then at Ticinum (Plate VI, 25 ; 46, 38, 43) 3 necessarily shortly before the tragedy occurred which ended in Crispus' execution. The mint of Rome confined itself to the special types designed for the Imperial anniversary.

Constantine's journey to Italy in 326 brought about a concentration of the gold coining in the East, or at least, in the mints close to the Bosporus. The mints of Sirmium and Ticinum were closed, while coining at Aquileia had been suspended slightly earlier.

The Emperor returned from Italy by way of Pannonia. On the last of December 326 he was in Sirmium; a short issue, by the vota type Principi iuventutis/vot xx (solidus, 162) dateable as 326–327 and including the medallions Gloria Constantini aug (Plate VII, 96) and Virtus DN Constantini aug (301), 4 must have been struck a little earlier in Siscia.

Acting on instructions previously received, Nicomedia during the Emperor's absence issued a special type to the decennalia of the Caesars, Votis x caess nn (315). The type was subsequently forwarded to Thessalonica, but at the time of issue Crispus had been put to death. The wording was accordingly altered to Votis x caes n (314), and the medallion issued together with the special type Votis decenn DN Constantini caes (311), the multiple Gloria Constantini aug (Plate VII, 97), a slight variation of the similarly worded Siscian type, and the Constantinus aug (solidus, 29). 1

The opening of the mint of Constantinople displays some unusual features. We know that the first bronze coins were struck in 326 at Constantinople, and one gold piece unequivocally belongs to the same year, namely the Gloria Romanorum with an obverse depicting Crispus as consul (105). The disturbing feature is that we know of no precedent of this type, although the Crispus coin points to a prototype struck during Crispus' third consulship in 324. Connected with the vicennial celebrations of Constantine is the uninscribed reverse with the Emperor standing in a quadriga (318). The first Constantinopolitan gold series was probably issued before the mint had been wholly established; the personnel of Aquileia appears to have been moved to the capital-to-be before the mints of Ticinum and Sirmium were closed and transferred to the Bosporus.

The first Constantinopolitan issue probably also comprising the reverses Constantinus aug (27) and Constantinus caes (34) 2 with diademed heads reminds us once more of the types created for 326 and brought by the Emperor to Thessalonica, Sirmium and Ticinum during his journey to Rome. True, the types survive for several years in Constantinople, but they lose their festival character, the obverses receive the ordinary legends, the busts become draped and cuirassed as was customary, Constantine wearing a rosette diadem, the Caesars laurel wreaths (e.g. 33).

During the busy years 327 (latter part)–330 (early part) Nicomedia was the residence of the court and its mint the central mint of the Empire. Before the inauguration of Constantinople on May 11, 330 Nicomedia issued three gold series, the first reflecting the series struck slightly earlier at Thessalonica. The multiples were Gloria Romanorum (108), Gloria Constantini aug (Plate VII, 98), Virtus Constantini caes (298) and Virtus Constanti caes (Emperor advancing r., holding trophy and spear between two captives, 296), the only solidus type being the Constantinus aug (Victory advancing l., holding trophy, 30) as in Thessalonica. Not much later a multiple varying the same theme was issued but with a different mintmark image namely, Gloria Constantini aug (Emperor standing between captives, Plate VII, 99) and finally, on the eve of the dedication of the new capital, the rejoicing of the Empire was expressed in terms of the Gaudium augusti nostri (Plate VII, 83) and Pietas augusti nostri (the latter struck both as multiple and solidus, 126). Here the creative phase of the Nicomedian mint ends. This issue appears to have been continued to the twenty-fifth anniversary of Constantine's ascent to the throne, when a series of Victoria Constantini aug/vot xxx (Plate VII, 255) was issued. The characteristic truncation of the obverse head as well as the short hair on the nape of the neck connects these coins with the previous issues and distinguishes them from the subsequent vota issues of 335–336 (Plate VIII, 256–7).

Before the responsibility of the central mint was conferred upon Constantinople, Heraclea struck a group of gold. The 3-solidi piece Salus et spes reipublicae (170) was shortly to be repeated in the new capital. The solidus type Victor omnium gentium had been struck after Civil War II in Nicomedia, but the rosette diademed bust of Constantine indicates a later time of issue (210). The third type was Victoria caesar nn (Victory advancing l., holding wreath, 221). 1

End Notes

1 Trier in Gaul is an exception.
2 These types can be singled out as belonging to the same issue because, of the remaining gold coins of Sirmium marked SIRM, all have counterparts in the short Nicomedian series marked N (Constantinus aug, Plate VI, 23, Crispus caesar, 44, Constantinus caesar, 35, Constantius caesar, 40). That the types Salus reipublicae and Securitas reipublice belong to this issue is also shown by a comparison with Nicomedia, where the issue preceding the N- marked coins comprised both the Constantinus aug (Vict. seat. on throne, Plate V, 22) and the types of the Augustae (174, 185).
1 All obv. busts are laureate, the multiples use the wording DN CONSTANTINVS MAX AVG for the Augustus. At first the elder sons of Constantine were given the long legends DN CRISPVS NOBILISS CAES and DN CONSTANTINVS IVN NOB CAES. Constantius II, however, is never named D(ominus) N(oster), a fact which suggests that the obv. leg. of the elder sons were conceived in advance of the capture of Nicomedia (similar leg. had earlier been used in Sirmium) but very soon curtailed.
2 Miss Toynbee, J RS 1947, p. 137, note 18 and Roman Medallions, p. 196, and Alföldi, J RS 1947, p. 12.
3 Recorded by Maurice (III, p. 57, rev. IX) without ref. to any coll. or sale cat.
4 Salus reipublicae was probably struck also at Thessalonica though no specimen has survived.
1 This solution implies that Constantine before his final victory over Licinius had planned to underline the importance of his dynasty by elevating his mother and wife to the rank of Augustae and his third son to the rank of Caesar, and to assume the oriental sign of sovereignty, the diadem. In actual fact, he might have done all this before Chrysopolis except for conferring princely honors on Constantius.
1 The fact that all the coins are solidi or multiples of the solidus (except the heavy Felix processus cos vi aug n) proves the coins to have been struck after Constantine's conquest. Strictly speaking this is not correct of the Felix processus cos ii. We know only a Felix processus cos iii of Crispus marked · SIRM · (in BM) and another marked SIRM (Hirsch XXIV 2613). In the Hunter coll. there is an irreg. Felix processus cos ii aug (68) of Const. II. This coin may, however, be a barb. copy (of distorted rev. leg.) of a regular Sirmian original. Because of the date clearly indicated on the consular coins they have caused confusion among scholars, and so Maurice was moved to declare them false (Num. Const. III, pp. 198f.). On formal grounds there would be no objection to a legend recording a certain consulate (n) until the person in question acquired his following consulate (n+1.). Thus e.g. the consular legend COS VI for Constantine would be valid throughout 320–325 (incl.) until he in 326 entered his seventh consulship. In practice, however, the consulships recorded on the coins of the Constantinian epoch appear to have been struck in the actual year of office. Be that as it may, the consular coins recording the second consulships of the Caesars were outdated in the year 324. On the other hand, we know that Constantine's appointing of his sons as consuls for the year 321 actually broke off the friendly relations between Constantine and Licinius (Seeck, Regesten, p. 172, my Arelate, p. 33, also note 3) ; the latter did not accept the Western consuls, appointed himself and his son opposition consuls for this year, and started a consular reckoning according to the formula consulibus quos iusserint dd nn (I, II, III and IV), cf. Kase, P RollPrinc, p. 34. Clearly, the two consular coins cannot have been struck by Licinius — and on no account at solidus standards.
2 Second O an engraver's slip.
1 No Crispus obv. of the latter variety known, but obv. of all other Caesars For Constantine II cf. e.g. for m.m. SIS Naples coin, Fiorelli cat. no. 14297, m.m. CONS Jameson cat. IV, 362, m.m. TS Hirsch XXVI 824, m.m. TR Maria Alföldi, "Trier ," Pl. XI, 5 (cf. 152–161).
2 It is possible that the second consulships of the sons of Constantine were noted according to the Antiochene Fasti, where only one previous consulship must have been recorded, since Licinius had refused to accept the consuls of 321 (Crispus and Constantine II). Even so the reverse types illustrate the rush to provide the army with cash.
3 Seeck, Regesten, p. 174.
4 Stein, Geschichte I, p. 160.
1 A similar obv. head on the silver piece Genium PR, cf. J RS 1947, Pl. I, 4.
2 Cf. Elmer, NZ 1930, Pl. II, 11.
1 The reverses Constantinus caesar and Constantius cæsar are the only ones preserved to our day. It is highly probable that the Constantinus aug and Crispus caesar were issued at the same time, as was the type for Helena, Securitas reipublice (184), struck there two years earlier.
2 Constantine is known to have stayed there at least from April 18 to May 22.
3 The Ticinese type Constantinus aug, Vict. adv. 1. (26) probably has to be regarded as a mule.
4 The prototypes were the Nicomedian Virtus Constantini caus (sic, 300) and Virtus Constanti caes (295) of 325 (these in their turn originating in the Virtus aug et caess nn, 287, first struck at Sirmium). A Virtus DN Constantini aug might well have been struck at Nicomedia at the same time.
1 The time of issue could possibly have been early 327 when Constantine on his return journey to Nicomedia spent some time in Thessalonica.
2 There is no corresponding coin in Paris for Constantius II, M. Lafaurie informs me. The later stages of the development of the Constantius II type are, however, abundantly represented in the material. The type Constantiniana Dafne (18) appears to have been a "gold strike" from a bronze die. The probable date is 328, cf. Alföldi, ZN 1926, p. 164.
1 Constantine was at least twice in Heraclea in 329, in August and in October.

4. From the Inauguration of Constantinople to the Death of Constantine

In connection with the inauguration of the new capital the central mint moves to Constantinople, a fact signalled by the reverse legend Adventus augusti n (5). An impressive array of magnificent multiples is issued, the 30-solidi piece Gaudium Romanorum (90), the 9-solidi Salus et spes reipublicae (171) and the 3-solidi Gaudium augusti nostri (84), a link with the preceding issue of Nicomedia. The ordinary solidi continue the series issued in 327 with Victory advancing l., holding wreath and palm branch. The reverses include individual legends for each ruler (33, 39). At this juncture the first Constantinopolitan vot xxx (suscepta) appears to have been struck; the only solidus known (252) has an anachronistic consular bust, a reminder of Constantine's eighth consulship in 329. The corresponding sub-multiple (253), the first one to be struck, employed the obverse legend CONSTANTINVS AVG, whereas all others regardless of mint subsequently used CONSTANTINVS MAX AVG (258).

For the following years we have but scanty information on the Emperor's whereabouts. Most of the notices put him in Constantinople. The gaps are considerable and give almost any amplitude for travels over large parts of his Empire. I record the most conspicuous ones below (according to Seeck's Regesten) :

a April 17, 331 Constantinople
b April 12, 332 Marcianople
October 17, 332 Constantinople
c October 26, 332 Constantinople
April 18, 333 Constantinople
d May 4, 333 Constantinople
October, 333 Aquae
e November 11, 333 Aquae
June 17, 334 Constantinople
f August 25, 334 Naissus
March 22, 335 Constantinople
g November 7, 335 Constantinople
August 22, 336 Constantinople
h February 4, 337 Constantinople

Although the coin reverses now grow increasingly stereotyped, we occasionally get indisputable evidence of the Emperor's travels. The first after the celebrations in Constantinople — if we except some tricennial vota suscepta both of Nicomedia (255) and Constantinople (252) struck July 25, 330/331 — is a Thessalonican issue (mintmark MTS) with a submultiple Victoria Constantini aug/vot xxx (258). 1 As important innovations in the same issue we may note the introduction of the Victoria Constantini aug (Victory advancing l., holding trophy, Plate VII, 239), and Principi iuventutis (Prince standing, holding vexillum, behind two standards, Plate VII, 154). The third solidus type is the Victoria caesarum (Victory advancing l., holding wreath, 223). The date of the following issue of Thessalonica (marked TS), repeating the two former types (240, 155), is in doubt. A sojourn of Constantine in Thessalonica any of the years 332–334 is possible; the absence of Constans, however, in this rich issue, excludes the year 334, and this, I believe is as close as we can get the date.

In the same years, i.e. in connection with Constans' elevation to the rank of Caesar and with Constantius' decennalia we get the first gold issue of Constantinople after the inauguration of the city and after the first vota issue of 330–331. The former event is attested by the 9-solidi piece Securitas perpetua (176), the latter by the solidus Victoria Constanti caes/vot xv (225). That the sons of Constantine were in the foreground is further attested by the heavy medallions Principi iuventutis (152, no obverse of Constans known). With the exception of the Gaudium augusti nostri (Plate VII, 85) 2 no medallion obverse of Constantine is known. The solidi of this gold series were of the traditional type, Victory advancing l., holding wreath.

Before we arrive at the special issues of the tricennial year we should mention a gold series struck at Siscia in 334, 3 comprising two types introduced 331 at Thessalonica, Victoria Constantini aug (Victory advancing l., holding trophy, Plate VIII, 242), and Principi iuventutis (Prince holding vexillum, 157). The latter type was shared by Constans also.

The tricennial issues are not easy to grasp. Rich series were minted at Siscia (Plate VIII, 249–50, 260), Nicomedia (Plate VIII, 256–7) and Antioch and a very small one at Aquileia (259). Thessalonica struck a special type of medallion (316, Plate VIII, 317)> but not the ordinary tricennial solidus. Finally Constantinople coined the unique Victoria Constantini aug/vot xxxx (Plate VIII, 262) together with the ordinary Victoria Constantini aug/vot xxx (Plate VIII, 254). We have, however, two different series of gold coins struck at Constantinople during the two last years of Constantine's life, the first continuing the traditional solidus type (Victory advancing l. with wreath) and introducing special types both for Constans (17) and Delmatius (53). This takes us past September 18, 335. The only vota type that possibly could belong to this series is the Victoria Constantini aug/ vot xxx (Plate VIII, 254), but both stylistic criteria and the scarcity of the type refer it to the later vota issue of the summer 336. It would certainly have been unworthy of the occasion for Constantinople to coin so sparsely for the Augustus, had the first tricennial celebrations taken place in the Eastern capital.

It now appears highly likely that he did not celebrate in Constantinople in 335 but in Nicomedia. This we can deduce from the enormous output of tricennial vota solidi, multiples, and special types of the latter mint.

The Nicomedian gold of 335 was probably struck in two phases, the first using the upward looking portraits (Plate VIII, 256), the second the diademed draped busts (Plate VIII, 257). Together with the vota type Victoria Constantini aug/vot xxx we find the magnificent multiple Felicitas perpetua aug et caess nn (58), and the solidi Virtus Constantini caes (299) and Virtus Constanti caes (297).

The belief that Constantine celebrated his initial tricennalia at Constantinople has been deeply rooted and ultimately goes back to Eusebius' assertion that he read his eulogy to the Emperor in the Imperial Palace on this occasion. Already Valesius thought this highly unlikely, 1 and Heikel is of the same opinion; the eulogy was not a speech but a treatise, probably written in Jerusalem. 2 Even if, in its entirety or in part, it had been pronounced in the presence of the Emperor, this could have happened in 336, when we know that Constantine celebrated his Imperial anniversary at Constantinople.

After the celebrations at Nicomedia in 335, 3 the Emperor travelled west. On October 23, 335 we find him in Nicopolis, 4 not far from Thessalonica. The dates of the Constantinian constitutions of this year have caused some bewilderment. According to the accepted chronology Constantine was in his capital on October 21 or 22. 5 Now the subscription giving both datum and propositum is unclear and earlier commentators have been in doubt whether the consulships should be referred to the datum or to the propositum. 6 Every difficulty is solved if we accept the latter solution. This implies that there is a gap in the Emperor's timetable between July 25 and October 23. On the strength of the Siscian gold coins we are entitled to assume that Constantine in August–September 335 visited Siscia, and two mintmarks, the first (SIS) with three varieties of the reverse Victoria Constantino aug/vot xxx (PLATE VIII, 249–50, 260), the second (·SIS·) 7 with one variety only (251), attest the correctness of the assumption. Contemporaneously a Principi iuventutis (Prince standing, holding spear and globe, 139) reverse and the type Victoria caesarum (224) were minted.

Returning from Siscia, obviously towards the end of October, 8 Constantine stayed for a short while in Thessalonica. On this occasion the Votis xxx reverse (aureus) both with (Plate VIII, 317) and without mintmark (316) was issued together with the customary solidus types Victoria Constantini aug (advancing l., holding trophy), 1 Principi iuventutis (Prince holding vexillum, 156), the dynastic multiple Solus et spes reipublicae (172), and, as a new creation, Gloria exercitus (Emperor standing, holding trophy, 100). This concludes the Constantinian issues in gold in this mint.

The Emperor then returned to his capital. The gold series issued in this context included the previous Constantinopolitan solidus reverse, varieties for the Augustus and all four Caesars, and the dynastic multiple (some months earlier struck at Thessalonica) Salus et spes reipublicae (173), and the new creation Securitas perpetua (177).

The final gold coining at Constantinople followed in the summer of 336 for the second tricennial celebrations, comprising not only the Victoria Constantini aug with both vot xxx (Plate VIII, 254) and vot xxxx (Plate VIII, 262), but also the Victoria Constantini caesar vot xx (265) for the vicennial celebrations due on March 1, 336/37. Other types were Principi iuventutis (Prince holding vexillum, 159, 160), Victoria Constantini aug (243) and Victoria Constanti caesar (226, both: Victory advancing l., holding trophy, a new reverse for Constantinople), the multiple Gloria Romanorum (106) and the aureus, without legend, showing the Emperor in a quadriga scattering coins (319).

It remains to account for two mints more or less out of touch — at least as far as gold coining goes — with the happenings in the rest of the Empire, namely, Antioch and Rome. Antioch ceased to issue gold after the coining of the winter and spring 324–325 and of early 326. During the tricennial year we find a series of solidi coined at the Syrian mint, Victoria Constantini aug/vot xxx (Plate VIII, 261) followed by some solidi of the reverses Victoria Constantini aug (Plate VIII, 245) and Victoria caesar nn (222, both: Victory advancing l., holding trophy). The solidi without vota must have been struck in the last year of Constantine, and both the series mentioned should be viewed in the light of the Sassanian war planned by Constantine (possibly also Calocaerus' rebellion) at the end of his reign. The tricennial series was preceded by a small issue of the reverse type "Victory advancing l., holding trophy," at first with the reverse legend Con- stantinus aug (31), subsequently altered to Victoria Constantini aug (Plate VIII, 244). In addition some Principi iuventutis solidi of the type "Prince holding vexillum" (153) were struck. It is natural to assume that this series, also connected with the preparations for the war, was struck immediately before the tricennial year. These issues must have been ordered to be struck by the central mint. 1

The decreasing importance of the old capital, Rome — at least financially — is demonstrated by the conspicuous lack of gold coins struck by the Roman mint. We had a meager issue during the vicennial celebrations. Later only two marks are known, R and PR, both, it seems, from the end of the Constantinian epoch. 2 The first PR coins are of the type Victoria nob caess showing Victory advancing l., holding a standard in each hand (268). Later the same wording appears with the type Victory advancing l., holding trophy and palm branch (mintmark PR, 269). One R-marked coin only is known of the same type as the latter PR coin and of the reverse legend Victoria aug (217). 3 The style of the portraits points to a late date. The reverses demonstrate the isolation — rather than independence — of the mint of Rome.

End Notes

1 The progressive enlargement of the portraits suggests the following internal order of the last Thessalonican marks: MTS, TS, TSE. The TSE issue is connected with the Constantinian tricennalia. The date 326 suggested by Mr. Carson (British Museum Quarterly XXI, 1957, P. 45) for the MTS mark is impossible on account of the mintmark.
2 Possibly originally struck for the inauguration of the city in 330. Constantine was in Constantinople (?) on June 17, in Singidunum on the Pannonian border on July 5, in Viminacium on August 4 (?) and in Naissus on August 25. The Emperor must therefore have travelled very fast and visited Siscia before Singidunum.
3 This, of course, was the year of the Sarmatian and Gothic campaigns.
1 Cf. Heikel`s edition of Vita Constantini, pp. CIVf., particularly note 1.
2 Heikel, ibid.
3 The Consularia Constantinopolitana (Chron. I, p. 235) record: Constantino et Albino: 1. His conss tricennalia edidit Constantinus aug die VIII k aug. 2. et levatus est Dalmatius caes etc. Thus the text gives no indication of the place.
4 CIust I 40,4.
5 Sirm. 4, CTh XVI 8,5; 9,1.
6 Cf. the apparatus criticus, CTh XVI 8,5. There is, however, also the possibility that the Sirmondian constitution, the best preserved of the three texts, is correct in leaving out the place of issue.
7 Submultiple marked SIS•.
8 On November 7, the Emperor was back in Constantinople.
1 The corresponding medallion (in the BM) weighing 7. 59 grm. is probably a "gold strike" from a die for silver coining.
1 The assumption that Constantine visited Antioch in the Spring 336 is tempting but probably too adventurous. Seeck's Regesten show the Emperor to have been in Constantinople on November 7, 335 and on August 22, 336. We have previously concluded that the final tricennial celebrations took place in his capital on July 25, 336. This would still leave room for a journey to the East, but as the literary sources are silent, it is prudent to decline this possibility.
2 The solidus Spes reipublicae (200) marked image of Fausta (in BM) is irregular in appearance and probably a "gold strike" from a bronze die.
3 The medallion Constantinus aug, Emp. stg. holding standard and scepter (32) weighing 7. 50 grm. is probably a "gold strike" from a silver die.


To express in a few words the conclusions of the present study is quite impossible. In a way the preceding analysis has been a conclusion in itself, based on material collected and studied over a considerable time. Those trying to follow the arguments have probably been irritated by the summary account of the basic material, the coins, and by the scanty references to the pertinent collections, publications and sale catalogues. Nevertheless, scarcely any coin mentioned is unpublished. 1

Summing up the results of the study of the relations between Constantine and Licinius as reflected in the coinage, we note that the bronze coinage is far more explicit than the gold coinage. The main reason is that bronze was struck continuously in most of the mints, whereas gold was issued sporadically and, generally speaking, only in one mint at a time. Of the two chronological problems discussed, the first regarding the date of Constantine's conquest of Italy is solved mainly with the aid of the gold coinage; one of the fundamental dates in the history of Constantine, the year of the "victory of the cross," of the battle of the Milvian Bridge, has to be altered to 311 from 312. The coins prove to be more accurate than the written tradition.

The second problem is solved with the aid of the bronze coinage. The conclusions previously drawn from my study of the coinage of Arelate are strongly supported by the survey of the bronze coinages of other mints. True, the smaller mints yield but meager results — they were obviously out of touch with the politics of the day—but all major mints eloquently show how Licinius was eliminated from the Constantinian bronze coinage on the eve of the nomination of the Caesars, and how the Caesars actually were appointed before the settlement of Serdica on March 1, 317. The Licinian mints are equally outspoken and demonstrate how Constantine disappears from the obverses and how Licinius stresses his sole-rulership, how in one case in the course of a single mintmark the reverse legend subsequently to this is altered to the plural augg and Valens appears as second Augustus, and how, finally, with the same mintmark, Constantine reappears at the same time as the sons of the Augusti are introduced as Caesars. This mintmark certainly gives us the history of Civil War I in a nutshell.

The revision of the dates of the Constantinian Civil Wars is the starting point for an investigation of the Constantinian financial administration.

The main theory developed in this study is that gold coining 1 in the Constantinian Empire 2 was closely connected with the court and with the Emperor. As a rule gold was struck at the temporary residence of the Emperor. This implies that the administrator of the gold coinage, the organizer, was a comitiva in the true sense of the word. Only in special cases, e.g., when the Emperor was campaigning and would have been hampered by bringing with him also a part of the civil administration, was the gold coining concentrated at the residence of the court left behind. Some instances of filial issues are known. In the years 320–321, e.g., we have parallel striking at three mints in the same area, Ticinum, Aquileia and Sirmium, obviously necessitated by the military preparations of these years. The last Constantinian gold series of Antioch should be explained in the same way — if our conclusion that the Emperor did not visit Syria in 336 is correct.

The fact that in almost every instance we may assume that the Emperor and the court directed the issues and almost supervised the execution in person gives added significance to the imagery, to the reverse legends, and to the slogans distributed by means of the coins. I have, however, refrained from attempting a new interpretation of the religious and political program of the Emperor in the light of the gold coinage. This important but arduous task is something for the future.

The iconography, too, is likely to appear in new light when studied against the background of the mobile gold minting. This concerns particularly such details as the development of the diadem, and the successive changes in the arrangement of the hair.

No doubt the new insight we have won of the working of the financial administration is likely to increase our understanding of the economic life of the later Roman Empire. It is remarkable that we can trace no tendency to cater equally to all parts of the Empire: Spain and Africa struck no gold in Constantinian times, Egypt nothing (under Licinian rule a single issue was struck at Alexandria), Antioch nothing during the years 327–335, Italy after the close of the mint of Ticinum nothing except some odd pieces in the last years of Constantine. One would have expected the central mint to have organized transports to far-away parts of the Empire, but it is obvious that the mobility of the gold minting increased the difficulties of segular service if, indeed, such service existed during the reign of Constantine. It appears that the local need of currency and cash was satisfied mainly by the bronze coinages.

The main results of this study, however, are, I believe of chronological character, partly as regards the readjusted itinerary of the Emperor and his suite, 1 partly in the discovery of a new numismatic method of checking the chronological evidence supplied by other sources. May the appended summary 2 speak for itself.

End Notes

1 E. g. submultiples have been brought into the discussion only when they have contributed something to the understanding of the chronological problems involved.
1 Obviously silver minting too.
2 The organization of Licinius appears to have worked differently.
1 I have not ventured to evaluate the intrinsic significance of the redating of the Constantinian constitutions discussed, nor to draw any conclusions with regard to the possible constitutions that now should be reconsidered, e.g., from the legal and prosopographical point of view.
2 Cf. p. 102.


Late imperial Delos rests in the obscurity of oblivion unworthy of its great past. Never wholly recovering from the catastrophic destructions in the first century b.c., the Roman (Hellenistic) way of life after the partial reconstruction nevertheless appears to have continued without noticeable break into Byzantine times. The shrine of Apollo, though within a relatively modest frame, retained its fame 1 well into the fourth century, and Julian is said to have asked for its advice before starting his ill-fated Persian campaign. 2 Later the existence of six Christian churches bears witness to the transition from Hellenistic paganism to the Christian faith. 3

No particular incident during the Late Empire is attached to the name of Delos. Life just went on, one hopes peacefully. Something of the mist of anonymity, however, is shattered by an early fourth century bronze hoard from Delos. Its very existence invites comments, its highly unusual composition seems to explain something of the circumstances surrounding its burial. Moreover, the hoard as such both sheds some light on Constantinian numismatic chronology, and is suggestive of a new methodological approach to coin hoards.

I. The Composition of the Hoard

The large Delos hoard comprising 3797 mainly Constantinian bronze coins was catalogued by Svoronos in 1910, 4 unfortunately at a time when only the first volume of Maurice's Numismatique Constantinienne had appeared (even if the majority of the mints had been described earlier in various learned journals). Thus an under- standable lack of insight into the working of the early fourth century mints in an unfortunate way impaired the value of the work of the great Greek numismatist ; e. g. bust and reverse varieties were not recorded in sufficient detail according to present day standards. Printer's errors (?) in some cases had disastrous effects through the mechanical way of using repetitions as often as possible.

The single coins of the hoard can therefore only hesitatingly be accepted as numismatic evidence, whereas the main groups, the general reverse types and the issues distinguished by mintmarks can be expected to serve some useful purpose. Immediately after the publication of the hoard, Kubitschek 1 drew attention to the fact that the hoard was an unusually closed one, the bulk of the coins having been struck during the years 308–318. In fact, a very negligible quantity of the coins exceeds these limits.

The composition of the hoard presents some exceptional and interesting features. Kubitschek pointed out that the hoard, regardless of its site in the Greek archipelago 2 , was formed in the West. The latest coins are the two Nicomedian ones marked image belonging to an issue initiated before Civil War II and struck during the war also. These coins are, however, the only ones struck by a Licinian mint after Civil War I, and they seem to be somewhat later than the rest of the coins.

The character of the hoard emerges most clearly out of a study of the last coins of the Constantinian mints integrated with the hoard. Arranging the coins in chronologically significant groups we get the following subdivision:

(a) marks and series with obverses of Constantine and Licinius, i. e. coins struck before the rupture between the two Augusti interpreted as a prelude to the Civil War (all the earliest coins inclusive of those from the early part of 316).

(b) marks and series with obverses of Constantine alone. These coins show that the relations between the Augusti had been severed; the date is the later part of 316.

(c) marks and series with obverses of Constantine and his two sons. These coins were struck in the course of the War, when Constantine reacted to Licinius' appointment of Valens as Augustus by elevating his sons to princely rank. This happened in the Autumn of 316. 1

(d) marks and series with obverses of Constantine, Licinius and their sons. These coins date from the time after the peace treaty and reconciliation at Serdica on 1 March, 317, a date confirmed by our literary sources.

A survey of all the coins belonging to groups (c–d) gives the following results:

Londinium 32
Lugdunum (closed 317–320) 88
Trier 54 3 75 72
Arelate 61 2 17 239 25.52
Rome 9 1 2011 0.4
Ticinum 28 4 269 3 10.41
Aquileia 44 17 57 77.19
Siscia 12 6 70 17.14
Thessalonica 14 4 106 13.20
Heraclea 18
Nicomedia 3 1 40
Cyzicus 116
Antioch 13
Alexandria 24
225 53 3158

In the entire hoard (groups a–d) other emperors were also represented, among them Maximin Daza by 168 obverses, Maxentius by 213 obverses. These have not been included in the table. If they had been included, the percentages in the last column would have been slightly lower except for Arelate (the mint was opened after the death of these emperors).

A further analysis of the 225 coins of the groups (c–d) shows that all the three last issues of the Treveran Sol coinage are represented, the latest coin being of Crispus and marked image 1 All other coins belong to the issues image and image

Two issues struck at Arelate after the Imperial reconciliation at Serdica are represented, namely image (two coins with Crispus obverses are marked image 2 with 36 coins, and image with 23 coins; on two coins the letter to left in the field is worn away.

Among the coins from Rome, 7 belong to the first "post-Serdican" issue image one Crispus coin has the rare mintmark image and the last is the consecration coin mentioned above.

From Ticinum we have 11 coins of the PT issue, 16 of image and one Constantine II marked image The 44 Aquileian coins all belong to the only postwar issue of the Sol coinage struck at that mint, marked AQP. The same applies to Siscia with six coins with obverses of Constantine marked image and six other coins for the Caesars marked ASIS. Thessalonica is represented by ten coins with obverses of Constantine and four coins with obverses of Constantine II, and Nicomedia, finally, by one Crispus mintmarked ΣΜΝΓ (no. 847) and two Licinius I (nos. 341–2) marked with image one each of officinae A and B (the latter of these is described as belonging to Nicomedia although the letters in the exergue are recorded as SMHB, the H probably being a printer's error).

Endeavouring to date these issues, we perceive, according to the present writer's table referred to above 1 that the coins of Rome, Aquileia, Siscia and Thessalonica scarcely had been struck later than 317, whereas the terminus ante quem for Trier would be early 319, for Arelate 320 and Ticinum 318. The latest Nicomedian coins, belonging to the same issue that was continued up to and throughout the Civil War II, must be dated to about 321. Of the two major issues struck at Nicomedia between the appointment of the Caesars in 317 and the Civil War II, Kent 2 assigns the first issue with its left facing consular effigies with mappa and sceptre to the years 318320 "in which years the five colleagues all assumed the consulship." The year 320 would therefore fit in very well for the beginning of the following issue, in all Eastern mints marked with image to right in the field. The Nicomedian coins give the approximate burial date 3 .

End Notes

1 Pausanias VIII, 33, 2 records that the Athenians in his time used to send a garrison to guard the shrine of Apollo. At that time, he states, the island was deserted by the Delians, "Δηλίων γε image νεκα image ρημός ἐστιν ἀνθρώπων," clearly a pessimistic exaggeration.
2 Theodoretos, Historia ecclesiastica III, 21, 1–2.
3 W. Déonna, La vie privée des Déliens, pp. 22f. Déonna dates the first Christian community to the end of the third or the beginning of the fourth century with reference to Inscriptions de Délos, nos. 2582ff.
4 J. Svoronos, Journal international d'archeologie numismatique XII (1910), pp. 153–193.
1 W. Kubitschek, "Ein Münzschatz auf Delos," NZ 43 (1910), pp. 50–53.
2 Ibid., p. 53.
1 As to the question of Constantine appointing his sons Caesars in advance of Serdica, where the appointment subsequently was confirmed, cf. Chr. Habicht, Hermes 86, 3 (1958), pp. 366f.
2 Among these the only specimen of a commemorative consecration coin, a REQVIES OPTIMOR MERIT with an obverse of Maximian.
3 Kubitschek, op. cit., p. 53 counts 123 coins for Licinius from Aquileia. The correct number is 15.
1 See the present writer's ''The Disappearance of Sol," Arctos, N. S. II (1958), p. 34 for summary.
2 Svoronos, op. cit., no. 857 erroneously (?) gives the mintmark as image Printer's error?

II. The Factual Background

The most interesting feature of the hoard is probably its exceptional composition with its preponderance of coins struck in the West. Normally a hoard "stands in a certain ratio to the amount of the coinage during the period covered by the find… and… this proportion reaches increasing agreement the larger the find is numerically." 4 "But," Thordeman continues, "this rule only applies to hoards which have been collected within the actual area in which the coins were currency." Now it is quite clear that the Delos hoard cannot reflect the coinage in circulation in Delos in the years after Civil War I. The circulation of the bronze coins was ordinarily restricted to the area surrounding the mint in question. This has been demonstrated in numerous cases, most recently probably for Gaul. 1 By common standards we would expect many more coins from the mints of Nicomedia, Cyzicus and Heraclea to have belonged to the hoard, and particularly one would have expected the latest coins to be from the mints located nearest the island.

An analysis of the hoard reveals that, except for the coins of Nicomedia (nos. 341–2, 847) which are according to the present writer the latest coins of the whole hoard, all coins struck during or after Civil War I had been issued in the Constantinian part of the Empire. 2 Before attempting an interpretation of this fact the position of Delos within the imperial administration should be clarified.

Diocletian's reform of the provincial administration transformed Asia into seven provinces, one of which was Insulae, the province of νήσων Κυκλάδων comprising 53 islands altogether. 3 Despite the fact that the surrounding islands were referred to Insulae, Delos together with Skyros, Lemnos and Imbros were referred to the Moesian diocese. 4 It now appears that the province of Achaia was created (recreated) by Constantine after Diocletian's reform; there is no Achaia in the Verona list. 1 The Constantinian Achaia, as distinct from the Achaia of the late Principate, was a proconsular province and C. Vettius Cossinius Rufinus seems to have been one of its first governors if not the first. 2 Now the Moesian diocese, comprising also the province-to-be of Achaia was ceded to Constantine at the reconciliation at Serdica; the creation of the new province cannot have taken place before that date. The exact time of the reform is uncertain, but Groag (p. 21) is probably correct in assuming that it took some time to complete the necessary arrangements. 1

In view of Delos' exceptional position in earlier times, it appears very likely that the island in the armistice of 317 was considered part of the Greek mainland and therefore in Constantine's provincial reform referred to Achaia. On the other hand it should be remembered that Delos was practically surrounded by parts of the province Insulae belonging to the Licinian diocese Asia.

Now the hoard we are discussing was found buried in an earthenware jar. It therefore represents a treasure, the ready cash of an individual, a body or an institution. 2 The fact that the hoard is an exceptionally closed one suggests the bulk to be, basically, a kind of cash payment from a stock of bronze coins rather than a great number of coins gathered during a period of time. A further inference from the composition of the hoard is that the main part of it was formed outside Delos. Accepting the year 321, when the last Nicomedian coins were incorporated with the bronze treasure, as the approximate burial date, 3 we realize that it would have been abso lutely impossible for so many Constantinian coins of Roman origin to flow to Delos through the ordinary channels. And if there had been regular connections with Italy (Rome) rather than with the Greek mainland (and the nearest Constantinian mint, Thessalonica), why was the percentage of post-316 (groups c–d) bronzes struck at Rome so much lower than the corresponding figure of Thessalonica? (Cf. the table above).

The composition of the hoard seems to be most easily interpreted in the following way:

The bulk of the hoard was received as a single sum somewhere in Italy in 316 and this lot of bronze coins can be assumed to have mirrored faithfully the bronze coinage in circulation in Italy at that time—the closer to the mint of Rome the sum was paid, the greater would be the proportion of recently struck coins issued by that mint, and the later the date. 1 In gradually decreasing proportions we would find coins of Ostia, Ticinum, Aquileia, Siscia, the Gallic, the Balkan and the Eastern mints (probably not more than stray pieces of the latter).

The recipient (presumably an army contingent) then was transferred to Gaul; the percentage of post-316 (groups c–d) Arelatensian bronze coins together with the remarkably high number of coins of Arelate, points to the Via Aurelia as the travelling route up to Arelate. Here additional cash was supplied—alternatively the coins of Arelate naturally poured in by way of exchange. 1 The small total number of Treveran coins (75 against 239 struck at Arelate) is particularly interesting in view of the fact that 54 pieces (72%) were post-316 (groups c–d). Two explanations appear to be possible, either that the person or body in question travelled to Trier but was forced to leave almost instantly, or that the army contingent staying in Arelate received reinforcements from Trier and that some of the cash brought forth from Trier trickled into the official treasure. At any rate, at a time suggested by the last date of the Gallic coins (probably 318) the contingent was dispatched to Aquileia (57 coins, 44 of which are post-316 from groups c–d), where mainly local coins trickled into the treasure, but also, in decreasing proportions, coins from Thessalonica, Siscia, Ticinum, Rome etc. This accounts for the post-316 coins from these mints (Rome only 9 coins out of 2011 in the table above, Ticinum 28 of 269–10.41%, Siscia 12 of 70–17.14%, Thessalonica 14 of 106–13.20%. The corresponding percentage of Aquileia is 77.19, the highest of any mint represented in the hoard).

To this comes a lot of almost 300 bronze pieces (group b, cf. p. 79 above) struck at Rome at a time when the bulk of the coins of this mint had already been incorporated in the treasure. The almost 300 coins in question were in my opinion struck during Civil War I fought between Autumn 316 and March 1, 317 and seem to have been shipped to Delos directly from Italy (Brundisium, cf. p. 96 below). The bulk of the coins had left Rome on the eve or in the early days of the war, while these coins were struck at Rome later, but before the end of the war. The 9 post-316 (groups c–d) coins are of later date.

In Aquileia the contingent embarked and headed for Delos. In Delos we would expect the coins in circulation to influence the composition of the hoard, and clear traces of the money market in Licinian times can be found. If we in the table above extract the coins struck by Licinian mints, we get a total of 361, less than 10% of the entire hoard. Some of these coins may well have slipped in at earlier stages of the travel of the treasure, particularly at Aquileia, but the Cyzicene quota (116 coins) cannot be explained in this way. Again the fact that only three post-316 (groups c–d) coins struck in the East were found in the hoard, shows that peaceful intercourse between Constantinian Delos and the surrounding islands of the Licinian province Insulae was negligible. 1 The majority of the 361 Licinian coins mentioned above must therefore have been in circulation on the island when the Constantinian garrison arrived. 2

Finally, we must assume that the Greek mainland kept in touch with Delos, and through this channel coins representative of the bronze coinage circulating in this area (mainly struck at Thessalonica, but necessarily also to some extent of old Licinian stock) oozed forth into Delos to be added to the hoard. The isolation of Delos must, however, have been fairly complete in view of the chronological gap between the latest Constantinian coins and the latest coins of the hoard, the two Nicomedian coins mentioned above. Remarkably enough no Thessalonican coins of later date than the Sol coinage belonged to the hoard. 3

It remains to explain the burial of the hoard. Kubitschek (p. 52) suggested the Spring of 324 and connected the burial with Crispus' moving the Constantinian fleet from Achaia northwards. Although the relative isolation of Delos, stressed in the preceding paragraph, might account for the absence of any coins later than 320–321, and thus explain such a late burial date, the very reason for burying the coins remains obscure if Kubitschek's views are to be accepted. The present writer therefore prefers to regard the burial against the background of the tension between Constantine and Licinius gradually growing from late 320 onwards. Our knowledge of the prelude to Civil War II is scanty. We know of Constantine's Sarma- tian War and of his violations of Licinian territory in the years preceding the Civil War. It would therefore be little surprising to find traces of similar actions along the sea frontier. Delos with its garrison must have been a thorn in the flesh to Licinius, a threat to the Licinian communications in the Aegaean, otherwise dominated by the East. It is feasible that a sudden raid finished the garrison on Delos; depending on our estimate of the burial date this raid could either be regarded as Licinian reprisals for Constantine's disregard of the division of the Empire as agreed upon at Serdica, or as the reason (or one of the reasons), for Constantine to operate with his army in Licinian Thracia.

Again, Constantine's decision to detach a contingent of his army to Delos 1 points to early planning of the second and decisive war against Licinius, and to an endeavour to use his bridgehead in the Licinian archipelago as well as possible. 2 The Licinian raid was certainly justified—from the strategic point of view.

End Notes

1 "The Disappearance of Sol," p. 34.
2 J. P. C. Kent, "The Pattern of Bronze Coinage under Constantine I," NC 1957, p. 31.
3 Kubitschek (p. 52) suggests that the hoard was buried shortly before Civil War II, not later than the spring of 324 when the Constantinian fleet with Crispus as nominal commander set out against Licinius from Piraeus. This date is probably too late.
4 B. Thordeman, "The Lohe Hoard," NC 1948, p. 201.
1 See e.g. the diagram showing the composition of Gallic (and Britannic) hoards in Carson-Kent, "Constantinian Hoards," NC 1956, p. 86.
2 This is strictly speaking not true of the IOVI CONSERVATORI of Thessalonica with the obverse legend IMP CONSTANTINVS PF INV AVG (5 coins in the present hoard, nos. 518–520). These coins were struck by the city of Thessalonica on its own initiative while cut off from its sovereign Licinius by the Constantinian army during Civil War I. The city obviously foresaw the victory of Constantine, but was reluctant to break its relations with Licinius before the armistice. Cf. p. 19 above.
3 J. Marquardt, Römische Staatsverwaltung I, pp. 190ff. Disagreement prevails as to some of these islands. Hierocles, p. 686 refers Amorgos and Astypalaia to Insulae; Brandis, RE I, col. 194 (s. v. Achaia), to Asia.
4 Hierocles, p. 648, 4, 5 and p. 649, 1, 2. See also E. Groag, Die Reichs- beamten von Achaia, Diss. Pannonicae I, 14, p. 23. The Atlas of the Early Christian World refers Delos to Insulae (map 19) and the other three islands mentioned, to Achaia.
1 See Notitia dignitatum, pp. 248f. E. Komemann, RE V, cols. 729–730 (s. V. Dioecesis) adds Achaia, and thus assumes the province of Achaia to have been preserved in Diocletian's reform. A. H. M. Jones, "The Date and Value of the Verona List," J RS 44 (1954), p. 28 believes that Constantine gained control of Achaia as early as 313 (sic) and that C. Vettius Cossinius Rufinus (cf. below) was accorded the proconsulate of Achaia at that juncture.
2 The main source recording Rufinus' career is CIL X 5061 : "praefecto urbi, comiti Augg nn, corr. Camp., corr. Tusciae et Umbriae, corr. Venitiae (sic) et Histriae, cur. alvei Tiberis et Cloacarum sacrae urbis, cur. viae Flaminiae, proconsuli provinciae Achaiae sortito, pontifici dei Solis, auguri, salio Palatino…" Groag, op. cit., pp. 17ff., however, connects a passage in Firmicus Matemus' Mathesis (II 29,10, pp. 81ff., ed. Kroll-Skutsch) with Rufinus and his father: "eius geniturae pater post geminum ordinarium consulatum in exilium datus est, sed et ipse ob adulterii crimen in exilium datus et de exilio raptus in administrationem Campaniae primum destinatus est, deinde <ad> Achaiae proconsulatum, post vero ad Asiae proconsulatum et praefecturam urbi Romae." Groag appears to be correct in his interpretation of the father as the consul of 316 and 323 (despite the lacking iteration in the Fasti) and the first praetorian prefect of Gaul, and the son as the urban prefect mentioned in the inscription quoted above. The details of the career of the younger Rufinus are, however, obscure. The fact that Atina, a Lucanian township, clearly hails him as its patron, its corrector, but nevertheless omits the correctura Lucaniae, entitles us, with Groag, to supply this office in the list and assign it to the time of Maxentius' usurpation. The wide range of offices held in Italy excludes the possibility that he was in office in Achaia in Diocletianic times (as suggested by C. Schuurmans, De Samenstelling van den Romeinschen Senaat, no. 776, in typescript only). The chronological order of the highest offices is hard to ascertain. Firmicus Matemus cannot be right when saying that the urban prefecture (315–16) followed after the proconsulate of Africa, the earliest date of which is 324 (note Groag, p. 20, n. 1 "Ich wieder- hole, daß das Horoskop eines Astrologen kein staatsrechtlich maßgebendes Dokument darstellt"). It should therefore be enough to accept Firmicus' statement that he held these offices without paying attention to the seeming chronological arrangement. Again, the CIL X 5061 appears to enumerate the offices in groups without attempting to record their chronological order (Groag, p. 19, n. 1). The younger Rufinus could therefore very well have been proconsul Achaiae after his tenure of the urban prefecture, and the present writer suggests that this actually was the case. (Groag, p. 19, n. 2, records a similar case as such an unusual occurrence in the cursus honorum).
1 Though accepting Groag's contention in this respect the author cannot agree with his reasons. He regards the Symmachus mentioned in Codex Theodosianus II, 15,1 as holding the office of vicarius (of the Moesian diocese). But above p. 40 the present writer has suggested that this constitution should be dated to July 25, 329 instead of 319. This, of course, has no further consequences in this context other than that we cannot automatically supply the title of vicarius in the Codex Theodosianus II, 4, 1 of February 4, 318, also mentioning Symmachus as addressee. At that time he may well have been proconsul Achaiae: whether he preceded or succeeded Rufinus the younger is impossible to determine. At any rate, it is very significant that two of the first officials charged with the administration of Achaia were trusted followers of Constantine, and Rufinus, one of the two, was an extremely experienced administrator. A third proconsul of the same category was Ianuarinus (Groag, pp. 21f.). The importance of the proconsular province of Achaia is also reflected in the rank and repute accorded to its governors within the bureaucracy (Groag, p. 22).
2 Svoronos (p. 154) suggests a small merchant dealing on the local market. This is impossible in view of the composition of the hoard as the author proposes to demonstrate below.
3 The question of the burial date is more complex than what would appear at first sight. The absolute terminus post quem is the turn of the year 320/321. Chronologically there is a puzzling gap of about two years between the latest Gallic coins and the three Nicomedian coins mentioned above. This indicates that Delos must have been more or less isolated after the arrival of the garrison. The Nicomedian coins show that the isolation was broken some time in 321, but nothing prevents us from assuming that the isolation continued until the time of the incident, which forced the individual in charge of the treasure to bury it. This may well have happened a couple of years after the 321 incident. It would therefore be more correct to give the burial date as 321–324 (inclusive). Our estimate of the burial date does not, however, affect the interpretation of the composition of the hoard.
1 Theoretically we could even play with the thought of interpolating the exact spot in a case, in which we could establish
  • the relative numerical strength of the output of two mints, and
  • the exact dates of all the coins of the hoard, or at least of the latest coins of both mints, if normal (peaceful) conditions had prevailed in the geographical area concerned up to the time of burial.
Assuming the latest coins of both mints to be exactly contemporary, and having worked out, on the basis of an independent and a very much larger group of material than that supplied by the hoard, that the mint A had an output twice that of the mint B, by equal representation of both mints in the hoard, the exact spot would be located so that the distance from A would be twice the distance from B.
1 And, of course, not only coins of Arelate, but also in decreasing proportions coins of Lugdunum, Trier, Londinium, Ticinum, etc. The fact that no post-316 Londinian coins were found in the hoard is suggestive with regard to the length of the stay in Arelate.
1 With regard to the presumable travelling slowness of the bronze coins we may possibly except the first year after the reconciliation of Serdica.
2 On the other hand the paucity of Licinian coins mixed with the hoard during the three or more years that passed until the coins were buried shows that economic life in Delos was very quiet indeed.
3 As to the date, see the author's "The Disappearance of Sol," p. 34 and, as regards the last IOVI CONSERVATORI of this mint, see above, p. 18. The presence of the latest Nicomedian coins might be due e.g. to a chance visit of a smaller vessel from Asia Minor.

III. The Numismatic Chronology

According to views expressed earlier by the present writer, the Sol coinage dominated the years 317–319 in Trier, and 317–320 in Arelate (where it continued to be struck contemporarily with other types during the years 321–322 although at that time forming a decreasing proportion of the entire output of the mint). 3 In Siscia and Ticinum the Sol coinage was followed by the VICTORIAE LAETAE PRINC PERP issues in 317 and 318, respectively. No single coin of this type was included in the hoard. But the author's contention that the plain vota coins (DN CONSTANTINI MAX AVG-VOT XX, CAESARVM NOSTRORVM-VOT V etc.) at Arelate were struck in consecutive issues during the years 317–321, 1 contemporarily with the Sol issues, receives no corroboration through the Delos hoard. On the contrary, although the Sol issues were by far the more numerous of the two, the natural selection of the current issues would have also included vota coins had both coinages been contemporary, and had the later Sol issue of the two represented in the hoard image really been struck as late as 319–320. Therefore, it seems necessary to squeeze the Gallic Sol issues within a shorter span of time, both in Arelate and in Trier, and assume that no major vota issues had been struck contemporarily (conceivably in 317–318; this would seem to support Mr. Carson's and Dr. Kent's view that the issues were "consecutive and exclusive, not contemporary." 2 Even so that does not mean that the Sol coinage ceased to be struck simultaneously in all the Western mints). 3 As regards the controversial date of the commemorative consecration coins, the composition of the hoard does not seem to speak against the date (317) suggested by the author. 4

In the present hoard the coins struck at Rome after New Year 317 were very scarce (9) and among them one was of the type REQVIES OPTIMOR MERIT- understandable with regard to the very big issues struck of these series. The corresponding REQVIES issues of the other mints, Arelate, Trier, Aquileia, Siscia and Thessalonica were lacking. Again, had these coins been issued before 317, the lack of more coins from Rome would have been very surprising indeed. At any rate, Kent's notion that these issues dated from the time immediately preceding Civil War II (recorded as belonging to the period 320–324, or 321–324), appears impossible. 1

Finally a few words remain to be said with regard to the extremely rich material from Rome. Svoronos did not attempt any chronological grouping, and the flan sizes recorded by him are suggestive but not conclusive (in some cases even controversial when e. g. all the coins mintmarked image with obverses of Licinius have the diameter 21 mm. whereas the corresponding coins with obverses of Constantine measure 22 mm.).

Now, if we assume that all the coins preserved to our days are a fair representation of the coins struck in antiquity, and if we take a fair selection of collections and calculate that, despite the haphazard way in which many of these collections have been brought together, the various factors tending to give any one collection a particular character, are likely to counterbalance one another, something might be gained by comparing the coins of a certain hoard with the number of the corresponding coins found in the collections. Taking the coins of the Delos hoard struck at Rome and comparing them with the material collected by the present writer, we get the following result: 2

RP (all large coins) 168 487 34.70
RP (SPQR POPVLI ROMANI) 1 86 165 52.12
image 106 77 137.66
image 416 156 266.67
image 586 147 398.64
image 231 91 253.85
image 20 83 24.10
image 26 57 45.62
RP (small size) 294 100 294.0
image 8 2 99 8.08

Before an interpretation of these figures is attempted, the sequence of mintmarks should be discussed briefly. The first mark employed after the battle of the Milvian Bridge is RP using the long obverse legend IMP C CONSTANTIN VS PF AVG. 3 This mark was followed by image and image using the same long obverse legend, and subsequently by a second RP series employing exclusively short obverse legends. 1 The latter series, of reduced standards, is the last one in which coins of Daza appear. The order of the subsequent marks is more difficult to establish. The decrease in module is almost imperceptible, 2 but nevertheless the image coins, which logically should form one group despite the varying serial marks, seem to be slightly larger and therefore earlier than the others. The correctness of this assertion is shown by the fact that the type MARTI CONSERVATORI was struck with the two former image marks, and the striking of this type was discontinued well in advance of Soli invicto comiti in other mints, 3 and, therefore, probably in Rome as well. Assuming a development of mint- marks from plain marks to compound ones, we get the sequence image after mid-313, the war against Daza. The crucial question is now the internal order of the remaining marks. The decrease in module suggests the image mark to be the earliest (21 mm. according to Svoronos), the image the last (19/21 mm.) with the image as the second. 4

We now return to the table above and to the percentages of the entire material constituted by the Delos coins. In actual fact we get a frequency curve, even if irregular on the right hand side of the diagram, representing the chronologically last mark of the series.


One important sequence of coins has, however, been omitted in these discussions, the third and obviously late RP series, the diameter of which was recorded as 19/21 mm. by Svoronos. The series comprises two types of Soli invicto comiti, the ordinary "Sol standing left, holding globe," and the exceptional "Sol with captive." Both types were struck exclusively with obverses of Constantine.

The difficulty in determining the position of the RP series in the sequence of mintmarks is the absence of other chronological criteria except for the measurements of Svoronos. Minor inaccuracies in Svoronos' lists would not have affected the treatment of the other series, all distinguished by clear serial marks. Now the type "Sol stg. l." is common to all three RP issues. Here all the coins of a diameter of 24 and 23/24 mm. have been referred to the earlier issues, whereas the coins of 19/21 mm. have been regarded as belonging to the late issue. However, in view of the difficulty of classifying coins on the evidence of measurement alone, Svoronos' figures should be treated with caution. On the other hand, the figures concerning the type SPQR populi Romani suggest that the maximum error cannot be very significant.

On the evidence of flan size the RP series should either precede or follow after the image mark; the probability of this contention is strengthened by the common feature of the two series: the absence of Licinian obverses. Theoretically, however, with the frequency curve as a point of departure, the RP issue with its percentage 294 could fit in between the image (266.67%) and the image (398.64%). In practice it appears very unlikely that the three RF's should not have been consecutive. 1 On the declining (right) side of the curve the RP percentage 2 could possibly fit in between the peak and the image(253.85%). The curve, interpreted in terms of political history would in this case give us the unexpected picture of two conflicts with Licinius (an idea previously suggested by Seston, and later elaborated by Andreotti) 1 with a peaceful period marked by the image and image series in between. Two facts speak against this solution, the flan sizes and the short time shown to have passed e. g. by the bronze coins of Trier between the rupture between Constantine and Licinius on the one hand and the official nomination of the Caesars at Serdica on the other. Thus it remains to place the RP series as immediately preceding the image mark. This gives us the following curve—with two peaks (cf. p. 97):

How can this curve be explained ?

One would, naturally, expect the issue current at the time when the hoard was collected to show the highest percentage. We have here two such instances. The bulk of the hoard would have been collected towards the close of the image issue. Later an addition was made in the very beginning of the image issue, comprising mostly late RP coins (possibly delivered directly from the mint as is suggested by the extraordinary high figure for the "Sol stg. l." type). Some earlier coins could very well have flowed into the treasure at the same time; in view of the few coins of the image and image (Sol with captive) series little importance need be attached to the variations of the percentage figures of these series. It is feasible that this part of the hoard never travelled very widely before it was shipped to Delos. It could have been shipped directly from Brundisium as cash for the garrison to be sent to Delos, possibly together with some troops. (With regard to the chronological discrepancy between this part of the hoard and the one shipped from Aquileia, one might even advocate that the consignment from Rome-Brundisium constituted the primary part, and the consignment from Gaul-Aquileia the secondary part, the reinforcement).


We may add that some of the earlier coins possibly were incorporated with the treasure at a later stage in Arelate, or still later, in Aquileia. The possible additions made when the army contingent was in Aquileia would chronologically be slightly later than the expiry of the image issue in view of the natural time lag.

With regard to all the facts known concerning the Delos hoard, it seems possible to explain its structure in detail. As to the general validity of the method employed, many objections can be raised. The basic material must be considerable and collected equally in the whole geographical area covered by the hoard. 1 It would obviously be wrong to compare an Eastern hoard with the coin material of the Western collections. Further, assuming that the hoard contained coins of the same weight and module, no unequivocal conclusions as to the chronological sequence of marks could be drawn. Considering the percentages only, the columns of the frequency curve could be grouped in many different ways. It must be assumed that a simple hoard, collected in the geographical area of the burial, would give us a straightforward curve similar to that on p. 94, if we exclude columns 7–9. In this case, however, it would be easy to switch nos. 4 and 6, but nos. 1(2) and 3 must be consecutive, no. 4 or 6 must precede no. 5. 2 In the case of a composite hoard such as the Delos one, only certain clues as to the internal order of marks can be found, but without other criteria it would be all too easy to draw a misleading curve with a single peak.

End Notes

1 The composition of the hoard, clearly indicating the travels of the onetime treasure, excludes the possibility that the coins had been brought to Delos through ordinary channels of trade, or by an individual merchant. But even if that had been the case, the factual explanation of the hoard would scarcely differ from the one offered.
2 The recreation of the province of Achaia and the decision to make Achaia a proconsular province, and further, the appointment of Rufinus, Symmachus, ( ? ), and Ianuarinus as proconsules Achaiae also shows the importance attached to the province. Administrative views may well coincide with military reasons.
3 "The Disappearance of Sol," summary on p. 34.
1 P. Bruun, The Constantinian Coinage of Arelate, pp. 78–82.
2 Kent, "Pattern," p. 32.
3 Sol in actual fact disappears much later from the coinage. Sirmium, opened according to different views in 320 (Maurice, Numismatique Constantinienne II, p. 387) or 324 (Voetter, NZ 1913, p. 133), struck four different reverses of Sol, and Sol appears on gold coins of Aquileia, Ticinum, and Siscia as well. The last Constantinian Sol coins, SOLI COMITI AVG N was actually struck at Antioch in 324, after Civil War II (cf. above, pp. 63f.).
4 Cf. Arelate, pp. 39–43 and also "Some Dynastic Bronze Coins," Eranos LIII, pp. 193–198. Quite independently of the Delos hoard, however, one might advocate a slight readjustment of the possible dates (317–320) suggested on the basis of an examination of the Langwith, Luxor, and Nagytétény hoards (Arelate, p. 43) and further pinned down to 317 (ibid., p. 41). The hoard evidence should in the author's opinion be accepted but the consecration coins should be dated as struck after the Sol coinage, at least in some mints. All Sol issues of Trier use A and B as officina letters whereas the officina letters of the Treveran consecration coins are P and S. The consecration coins, therefore, are of later date.
1 Kent, "Pattern," nos. 218–223, 545–565, 699–701, 770–772, 837–842, 906–908.
2 The collections recorded in Arelate, pp. ix f., and in addition, the Museo Nazionale, Museo Capitolino, and Bibliotheca Vaticana in Rome and the Munich collection. The hoards mentioned in Arelate do not affect the material from the mint of Rome.
1 This particular type was chosen as a means of checking the result. As no coins of this type with the early long obverse legend are known, the type belongs to the later part of the issue; its percentage should therefore be slightly higher than the figure for the entire issue. The table proves this assumption to be correct. Moreover, this also seems to show that Svoronos' grouping of the less easily classified Sol coins of the mark RP on the whole is correct. If we include the coins of Daza in the figures above the percentages are slightly higher, as could be expected, 35.96 and 53.09, respectively.
2 The consecration coins have been excluded. Of these series the writer's material comprises some 200 specimens.
3 Note the same phenomenon in Ostia and Arelate. Cf. Arelate, p. 15, C. King, ''The Maxentian Mints," NC 1959, pp. 65, 74, and Kent, ''Pattern," p. 50, recording this mark as the earliest. Kent, however, does not discuss the significance of the long and short obverse legends.
1 Kent, ibid. See also King, op. cit., p. 74 who, however, speaks of a single RP issue.
2 Svoronos records: image 21 mm. (Constantine), 21 mm. (Licinius); image 22 mm. (Constantine), 21 and 21/23 mm. (Licinius); image 19/21 mm. and 21 mm. respectively; image 21 mm. (both); image 19/21 mm. (Constantine), 20 and 21 mm. (Licinius) ; image and RP 19/21 mm.
3 See Arelate, pp. 23, 65ff. and for Trier, "The Disappearance of Sol," pp. 18ff.
4 The internal order of the image mark on the one hand and the image mark and the RP series on the other is of some historical significance. No obverse of Licinius was struck in the course of the two latter, which therefore can be assumed to have been struck during Civil War I. If the image series is considered to be later, that would imply that Constantine after the war issued bronze coins depicting both him and his fellow-ruler in the East before the nomination of the Caesars, confirmed at Serdica on March 1, 317. The Civil War would in this case have been fought in 314/315 and not in 316/317.
1 Though too much heed should not be paid to the short series "Sol with captive," the numerous coins (far too many to be overlooked) of the regular Sol type would constitute a serious difficulty; if treated separately from the "Sol with captive" type, this series would be the peak of the curve.
2 If we regard the two types separately we get the following result: "Sol with captive" 18 Delos coins, 53 in the material, i.e. 33.96% "Sol standing 1." 276 Delos coins, 47 in the material, i.e. 589.36%.
1 W. Seston, "Relazioni," Int. Hist. Congr. VI (1955), p. 426, n. 3. R. Andreotti, "Licinius" in Ruggiero, Diz. Epigr., cols. 1004–IIII.

IV. Conclusions

The foregoing notes on the Delos hoard have shown Greece and in particular the recreated province of Achaia, now raised to proconsular status, to have been of particular importance to Constantine, no doubt on strategic grounds. In Delos, Constantine got a bridgehead in the Aegaean archipelago, otherwise dominated by Licinius, and it is permissible to assume that the burial of the hoard was caused by a Licinian raid on the island. In Civil War II the navies of the two adversaries played an important part; Athens and Piraeus seem to have been one of the foremost bases of Crispus, and Licinius' later concern with his possibilities to master the sea is mirrored in an Egyptian papyrus. 1

The extraordinary composition of the hoard, in view of the burial place, allowed us to trace in detail the history of its formation, and at the same time gave us some clues as to the date of certain of the coin series included in (or absent from) the hoard:

  • The commemorative consecration coins should probably be accorded a slightly later date than the one previously suggested by the present writer. The year 320 should nevertheless be regarded as a terminus ante quem.
  • The dates of the Sol coinages of the Western mints should probably be readjusted. It is feasible that they, generally speaking, disappeared as early as 318–319.
  • The earliest so called plain vota coins were scarcely contemporary with the Sol coinage. The hoard suggests that the bronze coinages were consecutive and exclusive rather than contemporary and over-lapping.

Finally an attempt was made to compare the numerical strength of the different issues of the mint of Rome with the large material collected by the author, assuming this material to be a fair representation of the bronze coins in circulation at the time covered by the hoard. This method, if used with caution, and with due regard to the individual characteristics of each hoard, seems to give some clues as to the internal order of mintmarks.

End Notes

1 The coins struck in the East were of no concern in this context; the subject of examination was the coins struck in the mint of Rome. These coins mainly circulated in the West, and the author's material comprised the coins of the major Western collections.
2 Had the coins which were paid in Italy (first batch) been buried immediately after having been received, we would have had no columns 7–9.
1 P. Vindob. Boswinkel 14 of a.d. 323; cf. C. H. Roberts, "A Footnote to the Civil War of a.d. 324," J EA 31 (1945), p. 113: "… Chrysopolis, a victory itself only made possible by the destruction of a large part of Licinius' fleet in the Dardanelles. Both sides had built up powerful fleets and Egypt contributed no less than 130 ships to the losing side. If much of Licinius' fleet consisted of such old tubs as are described (l. 3 πολύκωπα παλαιωθέντα) in the Vienna papyrus (which I should assign to a.d. 323 when it became apparent that war was inevitable), it is hardly surprising that it proved no match for the numerically inferior squadrons of Constantine."In view of the military activities involving Delos we may assign an even earlier date to the papyrus.



  • CTh I 1, 1, pp. 27, note 2, 45
  • CTh I 2, 1, p. 32
  • CTh I 2, 2, p. 35
  • CTh I 2, 5, p. 45
  • CTh I 4, 1, p. 44
  • CTh I 4, 2, p. 46
  • CTh I 12, 1, pp. 32, 37, notes 6 and 7, 43
  • CTh I 12, 3, p. 32
  • CTh I 15, 1, p. 45
  • CTh I 16, 1, pp. 35, 43
  • CTh I 16, 3, pp. 38, 40, 44
  • CTh I 16, 4, p. 46
  • CTh I 22, 1, p. 43
  • CTh II 4, 1, pp. 39, 85, note 1
  • CTh II 4, 2, p. 44
  • CTh II 6, 1, p. 43
  • CTh II 6, 2, pp. 36, 38, 44
  • CTh II 10, 4, p. 45
  • CTh II 15, 1, pp. 39f., 46, 85, note 1
  • CTh II 16, 1, pp. 39, 46
  • CTh II 16, 2, pp. 39, 46
  • CTh II 17, 1, p. 45
  • CTh II 18, 1, p. 44
  • CTh II 18, 2, p. 44
  • CTh II 18, 3, p. 45
  • CTh II 19, 1, p. 44
  • CTh II 19, 2, p. 44
  • CTh II 24, 2, p. 46
  • CTh II 30, 1, pp. 33, 44
  • CTh III 2, 1, p. 44
  • CTh III 17, 1, pp. 36, 39, 44
  • CTh III 30, 1, pp. 32, 43
  • CTh III 30, 3, p. 46
  • CTh III 32, 1, p. 45
  • CTh III 32, 2, p. 46
  • CTh IV 8, 4, p. 44
  • CTh IV 8, 5, p. 45
  • CTh IV 8, 6, p. 45
  • CTh IV 12, 3, p. 44
  • CTh IV 22, 1, p. 46
  • CTh V 2, 1, pp. 34, note 2, 36, 38, 40, 44
  • CTh V 10, I, pp. 40, 46
  • CTh VI 1, 4, pp. 43f.
  • CTh VI 4, 1, p. 46
  • CTh VI 22, 1, p. 45
  • CTh VI 35, 1, pp. 32, 43
  • CTh VI 35, 2, p. 38
  • CTh VI 35, 3, p. 44
  • CTh VI 35, 5, p. 46
  • CTh VII 4, 1, p. 45
  • CTh VII 20, 1, p. 45
  • CTh VII 20, 3, p. 46
  • CTh VII 20, 5, p. 46
  • CTh VII 22, 1, p. 38, note 1
  • CTh VII 22, 2, p. 44
  • CTh VIII 7, 1, pp. 33, 43, 55, note 3
  • CTh VIII 10, 1, pp. 32, 37, 43
  • CTh VIII 12, 2, p. 44
  • CTh VIII 16, 1, p. 44
  • CTh VIII 17, 2, p. 36
  • CTh VIII 18, 1, pp. 33, 44
  • CTh VIII 18, 2, p. 44
  • CTh IX 1, 1, p. 43
  • CTh IX 1, 5, pp. 40, 42, 45
  • CTh IX 3, 1, p. 44
  • CTh IX 3, 2, p. 45
  • CTh IX 7, 1, p. 45
  • CTh IX 8, 1, pp. 42, 44.
  • CTh IX 9, 1, p. 46
  • CTh IX 10, 1, p. 44
  • CTh IX 12, 2, pp. 43, 45
  • CTh IX 24, 1, pp. 42, 44
  • CTh IX 34, 3, p. 40
  • CTh IX 37, 1, pp. 40, 44
  • CTh IX 40, 1, pp. 32, 37
  • CTh IX 40, 2, pp. 35, 43
  • CTh IX 42, 1, p. 44
  • CTh IX 43, 1, p. 44
  • CTh X 1, 1, pp. 35, 43
  • CTh X 1, 4, p. 44
  • CTh X 1, 5, p. 46
  • CTh X 4, 1, p. 45
  • CTh X 8, 2, p. 44
  • CTh X 15, 1, pp. 32, 37, 43
  • CTh XI 1, 2, pp. 32, 37, 43
  • CTh XI 3, 2, p. 46
  • CTh XI 7, 1, pp. 32, 37, 43
  • CTh XI 7, 3, p. 44
  • CTh XI 7, 4, pp. 43, 46
  • CTh XI 16, 1, p. 38
  • CTh XI 19, 1, p. 44
  • CTh XI 27, 1, pp. 40, 46
  • CTh XI 27, 2, p. 45
  • CTh XI 29, 1, p. 32
  • CTh XI 29, 2, p. 36
  • CTh XI 30, 1, p. 32
  • CTh XI 30, 2, pp. 32, 37
  • CTh XI 30, 3, pp. 35, 37, 43
  • CTh XI 30, 4, p. 43
  • CTh XI 30, 5, p. 43
  • CTh XI 30, 6, p. 43
  • CTh XI 30, 7, p. 44
  • CTh XI 30, 9, pp. 36, 44
  • CTh XI 30, 11, p. 44
  • CTh XI 30, 12, p. 45
  • CTh XI 30, 13, p. 46
  • CTh XI 36, 1, pp. 32, 37
  • CTh XI 39, 1, p. 45
  • CTh XII 1, 6, pp. 36, 44
  • CTh XII 1, 8, p. 45
  • CTh XII 1, 16, pp. 40, 46
  • CTh XII 1, 17, p. 46
  • CTh XII 5, 1, pp. 43, 46
  • CTh XII 17, p. 45
  • CTh XIII 5, 2, pp. 35, 43
  • CTh XIII 5, 4, p. 45
  • CTh XIV 24, p. 46
  • CTh XIV 25, 1, pp. 36f., 43
  • CTh XV 1, 2, p. 44
  • CTh XV 1, 4, pp. 40, 42
  • CTh XV 3, 1, p. 38
  • CTh XV 14, 3, p. 46
  • CTh XVI 2, 5, pp. 45, 57
  • CTh XVI 5, 2, p. 46
  • CTh XVI 8, 5, p. 72, note 5–6
  • CTh XVI 9, 1, P. 72, note 5
  • CTh XVI 10, 1, p. 44
  • CIust I 40, 4, p. 72, note 4
  • CIust III 11, 3, p. 36
  • CIust IV 32, 25, p. 46
  • CIust V 71, 18, p. 46
  • CIust V 72, 43, p. 46
  • CIust VI 9, 9, p. 44
  • CIust VI 21, 15, p. 45
  • CIust VI 23, 15, p. 44
  • CIust VI 37, 21, p. 44
  • CIust VI 56, 3, p. 38
  • CIust VII 32, 10, p. 32, note 7
  • CIust VII 62, 17, p. 46
  • CIust VIII 10, 6, p. 44
  • CIust X 1, 7, p. 46
  • CIust XI 62, 2, p. 44
  • Frg. Vat. 33, p. 34
  • Frg. Vat. 273, p. 43
  • Frg. Vat. 274, p. 34
  • Frg. Vat. 290, p. 43
  • Sirm. 4, p. 72, note 5


Trier 2 literary TR–PTR
Spring-Oct. 28, 311 Trier campaigning literary PTR
New Year 312- Rome Rome rev. type 3 PR POST, POST*
Late 312- early 313 Ticinum Ticinum lit. ; rev. type SMT
313 Ticinum-Trier Arelate minting PARL
Late 313- July 25, 315 Trier Trier CTh; rev. type PTR
Autumn 315 Ticinum Rome CTh SMT
October 315 Ticinum Milan CTh SMT
End of 315 Ticinum Italy CTh SMT; PT; ·Τ·
Early 316 Ticinum Trier CTh SMT
Spring-Summer 316 Ticinum Gaul CTh SMT PTR 4
Autumn 316 Ticinum campaigning literary SMT
Early 317 Siscia Siscia rev. type PS; SIS; ·SIS· PTR
March 317–318 Thessalonica Thessalonica, E. Europe CTh; minting ·SM·TS·
Aquileia CTh; rev. type AQ
319 Ticinum Sirmium CTh SMT
320 Sirmium Serdica CTh SIRM ·AQ·; AQ
Early 320 Sirmium Serdica CTh SIRM
321 Sirmium Sirmium CTh; rev. type SIRM; ·SIRM·
322 Sirmium Sirmium, campaigning CTh; lit. SIRM
323 Sirmium Sirmium, campaigning CTh; lit. SIRM
Early 324 Sirmium Sirmium CTh; rev. type SIRM; ·SIRM·
Spring 324 Thessalonica ? Thessalonica CTh; minting SMTSA
Summer-Sept. 18, 324 Thessalonica campaigning literary SMTSA SIRM
Autumn 324 Nicomedia Nicomedia rev. type SMN SMT; SMAQ
Winter 324– 325 Nicomedia Cyzicus minting SMKE
Winter 324– 325 Nicomedia Antioch rev. type SMAN; SMAN; SMAN*
Spring- Summer 325 Nicomedia Nicomedia CTh; minting SMN
Autumn 325 Nicomedia Thessalonica 1 minting THES
New Year 326 Nicomedia Nicomedia minting N SMAN
Early 326 Nicomedia Thessalonica rev. type SMTS
April-May 326 Nicomedia Sirmium CTh; minting SIRM MNA 2
Early summer 326 Nicomedia Ticinum minting SMT
July-August 326 Nicomedia Rome CTh; lit.; rev. type SMR TR
Early 327 Nicomedia Siscia rev. type SIS CONS 3
March-April 327 Nicomedia Thessalonica CTh; minting SMTS
Autumn-Winter 327–28 Nicomedia Nicomedia CTh; minting SMN
End of 328 Nicomedia Trier CTh; minting TR image
End of 329 Nicomedia Heraclea CTh; minting SMHER ; SMH
Winter 329–330 Nicomedia Nicomedia minting SMN
Spring- Summer 330 Constantinople Constantinople CTh; lit.; rev. type CONS
July 330 Constantinople Nicomedia rev. type SMN
Winter 330–331 Constantinople Thessalonica rev. type MTS
Winter 332–333 Constantinople Thessalonica minting TS
Winter 333–334 Constantinople Constantinople CTh; minting CONS
Summer 334 Constantinople Siscia lit.; minting SIS
Winter 334–335 Constantinople ? ? SMAN 1
July 335 Constantinople Nicomedia rev. type SMN TR
Autumn 335 Constantinople Siscia rev. type SIS; ·SIS· SMAQ 2
October 335 Constantinople Thessalonica rev. type TSE
Winter 335–336 Constantinople Constantinople CTh; minting CONS SMAN; TR
Spring 336 Constantinople Antioch 3 minting image
July 336–337 Constantinople Constantinople CTh;rev. type CONS image

End Notes

1 Instead of the name of the mint the mintmark used for the issue in question has been recorded in this column.
2 During this period Trier was the permanent residence of the Emperor, and the Treveran mint the only one to strike gold. The limited area of Constantine's realms made it unnecessary to move the goldminting when the Emperor left his capital.
3 This entry is used when a reverse type unequivocally indicates either the time of issue or the temporary residence of the Emperor (e.g. consular or Adventus type).
4 Trier has been entered in this column only when the gold coining of the mint appears to have been influenced from the outside.
1 There is some difficulty in believing that Constantine should have been in Thessalonica both in the Autumn 325 and early in 326.
2 Exact date of issue in doubt. The question is whether the anniversaries of the Caesars were celebrated on March 1 or on July 25, co-ordinated with the anniversaries of Constantine.
3 This filial issue actually struck in the Autumn of 326.
1 Struck before the beginning of the tricennial year. We have no indication as to the Emperor's whereabouts.
2 One coin only is known. It is possible that Constantine returned from Siscia to Thessalonica by way of Aquileia.
3 In the text, p. 74, n. 1 above, I have not ventured to suggest that Constantine visited Antioch in 336, although it is quite feasible that he did.


This index of reverse legends does not aim at recording the gold coinage completely, but at illustrating the coins and numismatic landmarks discussed in the text (where e.g. submultiples have been included only when they have contributed something to the chronological discussion). Occasionally a cursory rev. description has been added to keep apart types of identical legg. struck with the same m. m. The point has been to illustrate the type, not to mention all possible obverse varieties. When gold was struck in several off., one off. only has been indicated. Collections referred to have been abbreviated as follows :

  • ANSM (American Numismatic Society Museum)
  • Aqu(ileia)
  • B(erlin)
  • Bel(grade, ex Weifert coll.)
  • Bu(dapest)
  • C(openhagen)
  • DO (Dumbarton Oaks)
  • H (The Hague)
  • Hu(nterian Museum, Glasgow) L(ondon, British Museum)
  • Len (ingrad)
  • M(unich)
  • Mi(lan)
  • Ox(ford)
  • P(aris)
  • Pa (Paris, ancien catalogue)
  • RC (Rome, Museo Capitolino)
  • RT (Rome, Museo delle Terme)
  • St(ockholm)
  • V(ienna).

The roman numbers preceding the no. of the rev. leg. refer to the plates. Rev. legg. recorded in this index have not necessarily been mentioned by legend in the text.

1 ADLOCVTIO AVG, S.M.T, Hess April 1936 no. 2780, p. 53
3 ΣΜΝГ, d'Ennery p. 189 no. 11 1 , p. 61
V 4 SMAN*, Hirsch XXIX 1375, p. 63
5 CONS, d'Ennery p. 189 no. 10, p. 69
6 CLARITAS REIPVBLICAE (Sol stg. l), ·SM·TS·, Ox, p. 56
7 (Sol stg. l, kneel. capt.), ·SM·TS·, Hirsch XXII 154, p. 56
III 8–10 CONCORDIA AVGG NN, SMT, Cahn 75 no. 1493 (Const.), Mi (Lic.), V (Crispus), p. 58
11 SMAQ, B, p. 60
12 SIRM, C, p. 64
13 SMN, ANSM, p. 61
14 SMKE, DO, p. 62
V 15 image Hirsch XXII, pp. 63f.
16 CONSECRATIO, PTR, Maurice I Pl. XXII 8, p. 47, note 1
18 CONSTANTINIANA DAFNE, image C, p. 67, note 2
19 CONSTANTINVS AVG (Vict. seat. on throne), SIRM, P, p. 60
20 SMT, V, p. 60
21 SMTSB, Rollin & Feuardent 1909 no. 320, p. 60
V 22 SMN, Hirsch XXXIV 1559, pp. 60, note 2, 61
VI 23 (interlaced wreaths), N, Ars Classica XVI 2016, pp. 60, note 2, 66
VI 24 SIRM, Canessa 1923 no. 554, pp. 66f.
VI 25 SMT, Rollin & Feuardent 1896 no. 799, pp. 66f.
26 CONSTANTINVS AVG (Vict. l, hold. wreath), SMT, P, pp. 62, 66, note 3
27 CONS, Canessa 1923 no. 555, pp. 67, 69
28 SMK, M, p. 62
29 (Vict. l, hold. trophy) SMTS, Jameson no. 355, p. 67
30 SMNC, Hirsch XXXIV 1558, 1 p. 68
31 SMAN, V, pp. 73f.
32 (Emp. stg., stand.), RT, V, p. 74, note 3
33 34 CONSTANTINVS CAES, CONS, Cahn 47 no. 1024 (laur.), Len (diad.), pp. 67f., 69
35 CONSTANTINVS CAESAR, N, Ars Cl. VIII 1499, pp. 60, note 2, 66
VI 36 SMTS, Merzbacher 1910 no. 2170, pp. 66f.
37 SIRM, M, pp. 66f.
38 SMT, Jameson no. 368, pp. 66f.
39 CONS, L, p. 69
40 CONSTANTIVS CAESAR, N, L, p. 60, note 2, 66
41 SMTS, Hirsch XXII 226, pp. 66f.
42 SIRM, ANSM, pp. 66f.
43 SMT, V, pp. 66f.
44 CRISPVS CAESAR, N, ANSM, pp. 60, note 2, 66
45 SIRM, Jameson no. 361, p. 66
46 SMT, DO, p. 66
V 48 SMN, V, p. 61
50 SIRM, Hess May 1935, no. 3970, p. 63
54 EQVIS ROMANVS, SMN, Ars Cl. III 184, p. 62
VI 55 SMTS, Bel, p. 65
56 FELICIA TEMPORA, ·Τ·, Ox, p. 53
57 MAQ, Glendining Nov. 1958, no. 239, p. 57
58 FELICITAS PERPETVA AVG ET CAESS NN, SMN, Rev. Num. 1906, Pl. VII 2, p. 71
59 FELICITAS PERPETVA SAECVLI, PARL, Rollin & Feuardent 1909, no. 321, p. 55
IV 60 ·SM·TS·, Ars Cl. XVII 1889, pp. 55f.
62 FELIX ADVENTVS AVG N, SIS, Schulman 1923, no. 2641, p. 54
63 SMN, V, p. 61
64 FELIX ADVENTVS AVGG NN, SMT, Toynbee, Rom. Med. Pl. XVII 11, p. 49
V 66 FELIX PROCESSVS COS II, SMAN, Jameson no. 361a, p. 63
67 FELIX PROOESSVS (sic) COS II, SMAN·, Egger XXXIX 1402, p. 63
68 FELIX PROCESSVS COS II AVG, SIRM, Hu, p. 63, note 1
69 FELIX PROCESSVS COS III, SIRM, Hirsch XXIV 2613, p. 63, note 1
70 ·SIRM·, L, pp. 60, 63, note 1
III 72 SMT, Ars Cl. III 185, pp. 8, 51f.
73 FELIX PROCESSVS COS IIIII AVG N, AQ, Hess May 1935, no. 3875, pp. 24, 56
74 FELIX PROCESSVS COS VI, ·SIRM·, Hirsch XXII 160, pp. 58f.
75 FELIX PROCESSVS COS VI AVG N, SMT, Caylus 1 , p. 58
76 ·AQ·, V, p. 58
77 SIRM, Aqu, p. 58
78 V 79 SMAN, Bu (aureus), A. R. Bellinger 2 (sol.), p. 63
III 81 SMT, V, pp. 51f.
84, VII 85 CONS, P (earlier), DO (later), pp. 69f.
87 PR, Ox, p. 48
89 SMT, J RS XXII Pl. III 16, pp. 52f.
90 MCONS, V, p. 69
93 GAVDIVM ROMANORVM/FRANC ET ALAM, PT, Hess 1951, no. 279, pp. 52f.
VII 97 SMTS, Jameson no. 352, p. 67
VII 98 SMN, L, p. 68
VII 99 imageCanessa 1923 no. 551, p. 68
102 GLORIA PERPETVA AVG N/MVL XX, ·SM·TS·, Hess May 1935, no. 3912, p. 56
103 GLORIA ROMANORVM (turr. female pres. Vict. on globe to Emp.), ·SIRM·, V, p. 60
104 (turr. female pres. wreath to Emp.), ·SIRM·, Bu, p. 60
105 (Roma seat. l), CONS, d'Ennery p. 235, no. 481 (Crispus), p. 67
106 CONS, P (Const.), p. 73
107 SMN, Len (325), p. 62
108 SMN, L (327), p. 68
109, 110 IOVI CONSERVATORI (hold. Vict. on globe), SMN, L, SMNГ, ANSM, p. 22
111 image ANSM, p. 22
112 image L, p. 16
113 SISC, C, p. 16
114 (Emp. crown. by Jup.), SMT, P, p. 58
115, 116 IOVI CONSERVATORI AVG (Jup. hold. thunderbolt),
image V ( Licinius pf aug), ANSM ( Licinius
augustus), pp. 16, 20
117 IOVI CONSERVATORI AVGG (Jup. seat., hold. thunderbolt), TR, L, p. 49, note 1
118 PR, B, p. 48
119 POST, ANSM, p. 48
III 122 SMT, Hess April 1936, no. 2781, p. 52
124 PERPETVA VIRTVS AVG, SMT, Ars Cl. III 178, p. 48
126 PIETAS AVGVSTI NOSTRI, SMN, Ratto June 1926, no. 2421, p. 68
127 PM TRIB P COS IIII PP PROCOS, PTR, Glendining 1929, no. 852, p. 52
128 SMT, Ox, p. 52
129 PT, Ox, pp. 52f.
130 PONT MAX PPP PROCS, TR, L, p. 47, note 1
131 PRINCIPI IVVENTVTIS (hold. spear, globe), PTR 1 , V, p. 47
132 PR, Maur. I pp. 205f., rev. VII, p. 48
133, 134 POST, L, POST*, L, pp. 48f.
135 SMT, Ars Cl. XVI 2022, p. 58
136 AQ, Bu, p. 58
137 SMAN, L, p. 63
138 SMAN·, Ars Cl. XII 3030, p. 63
139 SIS, C (Constans), p. 72
140, 141 PRINCIPI IVVENTVTIS (3 stand.), SIRM, Basel XV 865 (Crispus), Bu (Constantine II), p. 64
V 142 SMT, Bel, p. 60
V 143 SMAN, Jameson no. 365, pp. 63f.
144 SMAN*, L, p. 63
145 PRINCIPI IVVENTVTIS (2 stand.), SIRM, Hu, p. 60
146 SMT, Bu, p. 60
147 SMTSA, Rollin & Feuardent 1887, no. 716, p. 60
148 SMN, L, p. 61
149 SMKE, C, p. 62
150 SMK, Basel 159, no. 25, p. 62
151 PRINCIPI IVVENTVTIS (Prince hold. stand., sceptre, behind stand.), THES, Bel, pp. 65f.
152 PRINCIPI IWENTVTIS (hold. vexillum, behind 2 stand.), CONS, Toynbee, Rom. Med., Pl. XIX 8, p. 70
153 SMAN, Basel VII 666, p. 74
VII 154 MTS, Basel 159 no. 29, p. 70
155 TS, Hirsch XXVI 824, pp. 64, note 1, 70
156 TSE, P (Delmatius), p. 73
157, 158 SIS, St (Constans), p. 71, Naples (Constantine II), p. 64, note 1
159, 160 CONS, DO (Delmatius), p. 64, Jameson no. 362 (Constantine II), pp. 64, note 1, 73
161 TR, C, p. 64, note 1
167 PT, Bel, p. 53
169 SMT, Hu, p. 52
170 SALVS ET SPES REIPVBLICAE (Emp. seat. receiv. Vict. on globe), SMHER, V, p. 68
I71, 172 (Emp. enthroned, 4 princes), CONS, Hess April 1936, no. 2767, p. 69, TSE, Len, p. 73
173 (Emp. enthroned, 2 princes), CONS, H, p. 73
174 SALVS REIPVBLICAE, SMN, L, pp. 60, note 2, 61
III 175 SECVRITAS PERPETVA (Emp. hold. sceptre, erect. trophy), S·M·T, Hirsch XXIX 1385, p. 53
176 (Emp. stg, 3 princes), CONS, Maur. I Pl. XV 3, p. 70
177 (Emp. enthroned, 4 princes) CONS, Toynbee Rom. Med., Pl. VI 5, p. 73
179 SIRM, C, p. 58, note 2
180 SMN, Canessa 1923, no. 558, p. 61
181 SMAN, Ars Cl. XVI 2018, p. 63
183 SMTSA, Hirsch XXII 151, p. 60
184 SMTS, V, p. 66, note 1
185 SMN, L, pp. 60, note 2, 61
VI 186 SENATVS, SMN, Hirsch XXIX 1387, pp. 62, 65
VI 187 SMR, B, p. 65
VI 188 SMTS, L, pp. 65f.
IV 189 SOLI COMITI AVG N, SMT, Ars Cl. XVI 1901, p. 58
190 ·AQ·, L, p. 58
IV 191 SIRM, Schlessinger 1939, no. 580, p. 59
192 SMAN*, Len, pp. 63f.
193 SOLI COMITI CONSTANTINI AVG, SMT, J RS XXII Pl. III 21, pp. 53–56, 58, note 1
195 SOLI INVICTO COMITI, ·SIS·, V, pp. 54, 56, note 1
196 SIRM, Hirsch XXXIV 1564, p. 59
197 SMT, Hirsch XXIX 1388, p. 58
199 SMTSA, DO, p. 60
200 image L, p. 74, note 2
I 201–2 SPQR OPTIMO PRINCIPI, PTR, Sotheby July 1914, no. 184 (obv. break N–T), Ars Cl. XV 1920 (obv. break I–N), pp. 5ff., 48, 50, note 2
I 203 POST, L, p. 48
204 VBIQVE VICTORES, PTR, L, p. 49, note 1
I 205 PR, L, p. 48
206 VICTOR OMNIVM GENTIVM (Emp. hold. stand.), PTR, L, p. 52
207 SMT, L, p. 52
IV 208 SMTSE, Jameson no. 354, p. 60
IV 209 SMNГ, Ratto June 1926, no. 2430, p. 61
210 SMH, Basel XIII 765, p. 68
III 2II (Vict. crown. Emp.), SMT, Jameson no. 353, pp. 53, 56
I 212 VICTORE AVG N VOTIS/X MVL XX, PT, Jameson no. 352, pp. 8, 52f.
213 VICTORE AVG N VOTIS X/MVL XX, S·M·T, St, pp. 52f.
III 214 SMT, Schulman 1923 no. 2654, pp. 8, 52f.
215 VICTORE AVG N VOTIS X/XX, PTR, Ox, pp. 14, 52
216 VICTORE AVG N/X/XX, PTR, L, pp. 14, 52
217 VICTORIA AVG, R, V, p. 74
219 SMN, Jameson no. 357, p. 61
222 image L, p. 73
223 VICTORIA CAESARVM, MTS, Seligmann 1931, no. 45, p. 70
224 SIS, Hirsch XXII 222 (Constans), p. 72
227 VICTORIA CONSTANTINI AVG (Vict., no capt.), PTR, V, p. 48
I 228 PR, L, pp. 48f.
229 POST, V, p. 48
230 (Vict. l, capt.), PTR, L, p. 52
231 SMT, Dupriez Oct. 1934, no. 666, pp. 52, 56
232 SIS, Ratto June 1926, no. 2432, p. 54
233 ·SM·TS·, Bourgey March 1913, no. 663, p. 56
IV 234 (Vict. r, capt.) ·SM·TS·, Hu, p. 56
235 (Vict. crown. Emp.) ·SM·TS·, Bourgey March 1913, no. 664, pp. 19, 56
236 AQ, Canessa 1923, no. 561, p. 57
IV 237 ·SIRM·, ANSM, p. 59
IV 238 SIRM, B, p. 59
239 (Vict. l, trophy), MTS, Hirsch XXII 169, p. 70
240 TS, Canessa 1923, no. 562, p. 70
VII 241 TSE, Sotheby July 1914, no. 185, p. 73
242 SIS, Hirsch XXXIII 1470, p. 71
243 CONS, Santamaria 1938, no. 897, p. 73
VIII 244 SMAN, Ratto June 1926, no. 2434, p. 74
245 image Ratto Oct. 1934, no. 1105, p. 73
VIII 246 VICTORIA CONSTANTINI AVG/VOT XX, THES, Hess May 1935, no. 3917, p. 65
VIII 247 SMR, Rollin & Feuardent May 1909, no. 325, p. 65
248 SMN, Hu, p. 62
249 VICTORIA CONSTANTINI AVG/VOT XXX (Vict. w. vota shield), SIS, Ars Cl. III 195, pp. 71f.
250 (Vict. w. vota shield support. by genius), SIS, Hirsch XXII 190, pp. 71f.
VIII 251 ·SIS·, V, p. 72
VIII 252–4. VIII 254 CONS, Hess May 1935, no. 3885 (330), Bourgey Dec. 1913, no. 665 (330), Santamaria XVI 898 (335/336), pp. 69ff., 73
VII 255 SMN, Hirsch XXII 180 (330/331), pp. 68, 70
VIII 256 SMNC, Hirsch XXII 178 (335, diad. head), pp. 68, 71
VIII 257 SMNP, Hirsch XXII 189 (335, diad., drap. bust), pp. 68, 71
258 MTS, Dorotheum June 1955, no. 2235, pp. 69f.
259 SMAQ, Rollin & Feuardent 1896, no. 817, pp. 71, 104, note 2
VIII 260 (Vict. w. vota shield, push. capt.), SIS, Hirsch XXII 174, pp. 71f.
VIII 261 (adv. l, trophy), SMAN, Ratto June 1926, no. 2436, p. 73
V 264 ·SIRM·, Ars Cl. VIII 1503, pp. 24, 59
267 ·SIRM·, P, pp. 24, 59
268 VICTORIA NOB CAESS (Vict. l w. 2 stand.), PR, Ciani Oct. 1920, no. 208, p. 74
269 (Vict. w. trophy, branch), PR, Ball VI 2347, p. 74
IV 272 VICTORIAE LAETAE AVG N/VOT X/MVL XX, ·SM·TS·, Hirsch XXIX 1393, pp. 17, 20, 56f.
I 274 VICTORIAE LAETAE PRINC PERP/VOT X, PR, Ars Cl. VIII 1486, pp. 8, 48, 49, note 2
275 SMT, L, pp. 8, 48f., 50
I 276 VICTORIAE LAETAE PRINC PERP/VOT PR, SMT, Ars Cl. III 191, pp. 8, 48, 50
277 VICTORIAE PERPETVAE/VOT XX, SMT, Hess April 1936, no. 2782, pp. 56, 58
278 AQ, Len, p. 58
V 280 SIRM, Florange & Ciani 1924, no. 421, p. 59
V 281 SMN, Hirsch XXIX 1395, p. 61
V 282 SMKE, Platt, Coll. C, no. 94, p. 62
284 VICTORIBVS AVGG NN VOTIS /X/XX (Vict. stg.) PTR, V, p. 14
285 (Vict. seat. r), PTR, L, pp. 14, 56, note 2
287 VIRTVS AVG ET CAESS NN, SIRM, V, pp. 59, 66, note 4
288 VIRTVS AVGG NN, SMT, Pa, p. 58
290 VIRTVS AVGVSTI N (Emp. mounted), PTR, L, p. 52
291 SMT, J RS XXII Pl. III 6, p. 52
III 292 (Mars, capt.), SMT, Ars Cl. VIII 1487, pp. 52ff., 56
293 VIRTVS CAESARIN, SMTSA, Canessa 1923 no. 568, p. 60
294 SMNK, V, p. 61
295 VIRTVS CONSTANTI CAES (adv. r, hold. trophy, push. capt.), SMN, Baranowski 1931, no. 3178, pp. 62, 66, note 4
296 (adv. r, hold. trophy, 2 capt.), SMN, V, p. 68
297 (adv. l, trophy, dragg. capt. by hair), SMNP, Hirsch XXIX 1479, p. 71
298 VIRTVS CONSTANTINI CAES (adv. r, trophy, 2 capt.), SMN, Pa, p. 68
299 (adv. r, hold. trophy, push. capt.), SMN, P, p. 71
300 VIRTVS CONSTANTINI CAVS (sic) (adv. l, trophy, dragg. capt. by hair), SMN, M, pp. 62, 66, note 4
302 VIRTVS EXERCITVS GALL, (rev. without capt.), PTR, V, p. 48
303 (with capt.), PTR, L, pp. 52, 54
304 image Bu, p. 54
IV 305 SIS, Ars Cl. XII 3026, p. 54
306 ·SM·TS·, B, pp. 54, 56
308 VOTA PVBLICA, PTR, P, p. 52
309–10 SMNГ, L, SMN, Hirsch XXIX 1398, p. 61
I 312–3 VOTIS V MVLTIS X/VICTORIA AVG, PTR, Hirsch XXX 1259 (obv. break N-T), Rosenberg 1914, no, 462 (I-N), pp. 5, 48f.
314 VOTIS X CAES N, SMTS, Reo. Belge Num. 1958, Pl. I 6, pp. 24, 67
315 VOTIS X CAESS NN, MN·B, Hess April 1936, no. 2801, pp. 24, 67
316 VOTIS XXX, no m.m., Hess May 1935, no. 3889, pp. 71f.
VIII 317 TSE, Hirsch XXIX 1399, pp. 71f.
318 — (no leg.), Emp. in quadriga, no m. m., Canessa 1923, no. 566, p. 67
319 CONS, V, p. 73

End Notes

1 There is reason to suspect that the rev. leg. recorded as ADVEN TVS AVG N, is incorrect.
1 Gold despite silver mark on plate.
1 Cf. Cohen 154, Maur. II, p. 273, rev. VI.
2 I am greatly indebted to Prof. Bellinger for his kindness in drawing attention to this unique solidus.
1 A wide range of varieties both of denominations and of obv., all of Const., in different coll.