THE SUMMER OF BLOOD: The "Great Massacre" of 337 and the Promotion of the Sons of Constantine Author(s): R. W. BURGESS Reviewed work(s): Source: Dumbarton Oaks Papers, Vol. 62 (2008), pp. 5-51 Published by: Dumbarton Oaks, Trustees for Harvard University Stable URL: http://www.jstor.org/stable/20788042 . Accessed: 16/05/2012 13:25

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Dumbarton Oaks, Trustees for Harvard University is collaborating with JSTOR to digitize, preserve and extend access to Dumbarton Oaks Papers.

http://www.jstor.org THE SUMMER OF BLOOD The "GreatMassacre" of 337and thePromotion of theSons ofConstantine

R. W. BURGESS

in a indeed. To Tim Barnes, thefirstyear ofhis retirementfrom teaching: giant's shoulders,

I. Introduction

Constantine was the firstChristian fact that these would be considered amiracle," HA Although emperor, days massacre his reignwas marred by more familialbloodshed than that Claud. 2.6).1 This of themale descendants of was to most of any other Roman emperor: he himself involved Theodora,the half brothers and of thehalf nephews one or another in the deaths of his wife's of isone of themost degree father, Constantine, intriguing personal epi his wife's brother, his half sister'shusband, his eldest son, sodes in the history of theRoman emperors. The problem no source his wife, and another half sister'shusband and son (Max is that surviving ancient directly describes this we have tendentious imian,Maxentius, Bassianus, Crispus, Fausta, Licinius, event; hints, rumors, accusations, see statements that readers were and Licinius II, respectively; stemma, p. 6).Moreover, coverups, vague obviously accounts soon after his death most of the male descendants of meant to understand, and doctored forpolitical or Such source material has created Constantius I, his father, and Theodora, Constantius's religious purposes. s and half sister numerous formodern with the result second wife and Constantine stepmother problems scholars, were in a at least that of themassacre is contested: in-law, assassinated plot that involved virtually every aspect sons. names of those who met their the dates of their one of his The late-fourth-century author of the the end, was one massacre or HistoriaAugusta had thisbloody record in mind when he deaths, whether there two, who Claudius ancestor ofConstantine: the assassinations, and the behind eulogized II, supposed prompted reason(s) res them. there are almost as scenarios as there "[Claudius] amauit propinquos; nostris temporibus Indeed, many loved his a are scholarswho have theorized because the comparanda miraculo" ("Claudius relatives, them, usually is as an interlude between more episode treated merely narratives.2 important military, political, and religious

Pierre-Louis I should like to thank Tim Barnes, Cathy King, and Malosse for their comments on earlier drafts of this paper, aswell as the editors and was to have been between comments 1 Of course, since "Pollio" supposed writing the two anonymous referees for their detailed and helpful and and when Constantius I was still Claud. 1.1, 3.1, I alone am for what told 293 305, {HA suggestions. As always responsible keeping they events as as 9.9,10.7,13.2), he could not have known about late 337, but me to remove or fix. I should also like to thank Joel Kalvesmaki for help is the subtle and not so subtle text. this just another of many chronological with the formidable technical problems presented by my original the author. For a short but to to the slipsmade by anonymous late-fourth-century I should like thank the following for permission reproduce excellent introduction to this see A. His Numismatic , Harlan general problem, Chastagnol, photographs: Classical Group (nos. 2,3,4,8,14), J. toire Les romains des Ile et lile si?cles Lanz Auguste: empereurs (Paris, 1994), Berk Ltd (no. 5),Dr. Paul Rynearson (Vcoins) (no. 11), Numismatik = IX-XXXIV, C-CXXXI. LHS Numismatik (nos. 12,16,18), Dr. Busso Peus Nachfolger (nos. 9, AG Fritz 2 The studies are A. Olivetti, "Sulle di 15,17), H. D. Rauch GmbH (nos. 6,19, 21, 23), Tkalec (no. 7), major stragi Costantinopoli Marc ofAncient succedute alla morte di Costantino il RFIC 43 67-79; Rudolf K?nker M?nzenhandlung (no. 10), Breitsprecher grande," (1915): Coins & Artifacts X. Lucien-Brun, "Constance II et lemassacre des BullBud?, Imports Inc. (Vcoins) (no. 26), Imperial (Vcoins) (no. princes," M?nzen & Medaillen GmbH series, Lettres d'humanit? 32 R. Klein, "Die 20), Roman Lode (Vcoins) (nos. 22,24), and 4th (1973): 585-602; um nach dem Tode Constantins des (no.25). I shouldalso liketo thankMariana Reynolds for all her help with K?mpfe die Nachfolge Gro?en," in for 101-50 in idem, Roma versa aevum: Aus the intricacies involved preparing these photos publication. ByzF6 (1979): (reprinted per = = zur heidnischen und christlichen ed. R. von Conventions for dates: 337-38 337 to 338, 337 / 38 337 or 338. gew?hlte Schriften Sp?tantike,

DOP 62 6 R.W. Burgess

Stemma The Descendants of Constantius

= = ? (i) * (2) Eutropia

= = theodora Maxentius* (1) constantius (2) Fausta* = constantine i

fl. DALMATiust Julius CONSTANTiusf Hannibalianus Constantia Eutropia Anastasia

(i)=Galla =Liciniusr =Virius =Bassianus* I Nepotianus?tI

DALMATIUSf HANNIBALIANUSf Licinius II* = Constantina Nepotianus

gallus four other cousins of Juli?nt sont daughter = =Constantina

(2)=Basilina

Minervina=(i) constantine i (2)=Fausta* =Helena

constantine h constantius h Constantina Helena Crispus*

(1) =daughter (1)=hannibalianus =julian

(2) =Eusebia (2)=Gallus

*executed by Constantine tassassinated in 337

sources: Barnes, New Empire, 265-166, and PLRE 1:1129.

. andM. Di suivis de ed. A. Dihle Haehlingand Scherberich [Hildesheim, 1999], 1-49); Maio, Sept expos?s discussions, (Geneva, 1989), 120-25; T. Gr?newald, Constantinus Maximus in Jr., and D. W.-H. Arnold, "Per Vim, Per Caedem, Per Bellum: A Study Augustus: Herrschaftspropaganda der 153;P. Cara, ofMurder and Ecclesiastical Politics in the Year 337 A.D.," Byzantion 62 zeitgen?ssischen ?berlieferung (Stuttgart, 1990), "Aspetti see e e del conflitto per la successione di Costantino," RSCI (1992): 158-211 (on which, I. Tantillo, "Filostorgio la tradizione sul politici religiosi 47 39-5?; I?Tantino, La orazione di Giuliano a Costanzo: testamento di Costantino," Athenaeum 88 [2000]: 559-63). A selection (!993): prima recent can Introduzione, traduzione e commento 1997), 228-39; D. Hunt, of important interpretations be found (in chronological order) (Rome, , trans, "The Successors of Constantine," in The Ancient in E. Stein, Histoire du bas-empire, vol. and aug. J.-R. Palanque Cambridge History, vol. 13,The Late A.D. 337-42$, ed. Av. Cameron and P. (Paris, 1959), 131-32 and 484-85; A. H. M.Jones, The Later , Empire, Garnsey 3-4; B. Bleckmann, "Der zwischen Con 284-602: A Social, Economic, and Administrative Survey (Oxford, 1964), (Cambridge, 1998), B?rgerkrieg n. . A. stantin II. und Constans (340 Chr.)," Historia 52 (2003): 225-26 , 112;R. MacMullen, Constantine (New York, 1969), 224-25; Piganiol, 241-43; D. S. Potter, The Roman at AD 180-3?$ (New York, L'empire chr?tien (Paris, 1972), 82-83; R- Browning, The Emperor Julian Empire Bay, 460-63; and R. M. Frakes, "The of Constantine down (London, 1975), 34-35; J.W. Leedom, "Constantius II: Three Revisions," 2004), Dynasty . to 363," in The to the Constantine, ed. Byzantion 48 (1978): 132-36; RIC 8:4-7; G. W. Bowersock, Julian the Cambridge Companion Age of T. D. Constantine and Lenski 2006), 98-99. For lists of earlier accounts and discus Apostate (Cambridge, MA, 1978), 22-23; Barnes, (Cambridge, sions, seeOlivetti 76-77; Lucien-Brun 595-99; Di Maio Eusebius (Cambridge, MA, 1981), 261-62; E. G. Gonzalez, "Observaciones (above), (above), un cristiano: and Arnold 161-62 n. 24; H. Chantraine, Die sobre emperador Fl. Jul. Constante," Lucentum 3 (1984): (above), Nachfolgeordnung Constantins des 5-9. 268-70; C. Pietri, "La politique de Constance II: Un premier 'C?saro Gro?en (Mainz-Stuttgart, 1992), ou in et au IVe papisme' Vimitatio Constantini?" LEglise l'empire si?cle:

DOP 62 The Summer of Blood | 7

in matter is not two To make any headway this therefore wives, Helena and Theodora,3 bywhom he had had necessitate a one son sons easy, and the complexities of the evidence (Constantine), and three and three daugh is as two sons complex analysis. My approach follows. ters, respectively (see the stemma). The eldest context for of Constantine Constantine After establishing the general historical himself, Crispus and II, events summer caesar on the of the of 337 (section II and Appen had been proclaimed March 317, the former at dix 1),I beginwith theproblem of establishing"what perhaps around twenty years of age the time, the lat in of events and the ter less than a His next Constantius was happened," particular the sequence year.4 son, II, a caesar on 8 seven responsibilityfor them (section III). This beginswith proclaimed November 324, when only in or summer lengthyand detailed analysisof the survivingliterary years old.5 Shortly afterward, the spring of are as was memo sources, since they abundant and complex, well 326,Crispus executed and suffereddamnatio as sources The next subsection riae? on December his the epigraphical (III.i). Eight years later, 25 333, youngest presentsthe hints and cluesoffered by thecoinage of the son, Constans, then either ten or thirteen, was invested of Con with the rank of caesar aswell.7 Constantine had period immediately preceding the promotion By 332 stantine s sons This section reached or was about to reach his sixtieth (III.2 and figures). concludes probably just a his He knew his sons with summary and synthesis that establishes the rela birthday (i.e., sixty-firstyear). that were and and that he tive importance and reliability of the foregoing evidence very young inexperienced might a are not survive for more to them with (III.3). From this number of general conclusions many years provide on some were drawn that focus very much the question of instiga the experience they needed before of them tion: was it the army alone or was itConstantius? next most The section is concerned for the part with 3 PLRE 1:410-11, s.v. "Helena 3," and 1:895, s.v. "Theodora 1";T. D. summer Barnes, The New Diocletian and Constantine MA, thechronology of theevents of the of 337from Empire of (Cambridge, D. R?mische Kaisertabelle: einer the death of Constantine to the return of Constantius 1982), 33-34,36,37; Kienast, Grundz?ge r?mischen Kaiser 2nd ed. (Darmstadt, 1996), 281-82. to after the with his brothers. chronologie, meeting s-v- s.v. 4 PLRE 1:233, "Crispus 4," and 1:223, "Constantinus 3";Barnes, Since the date of the death ofConstantine iswell attested New Empire, 7, 44-45, 73; Kienast, Kaisertabelle, 305-6, 310. Crispus in I other known was son sons were the sources, begin with the only date, the of Constantines firstwife Minervina; his other sons to the of Fausta, his second wife and half sister of Theodora. Con that of the promotion of Constantine's augus offspring stantine II was not born on Barnes and The sole tus two sections 7 August (pace Kienast). (IV.i). Then follow short outlining source for this date, the calendar of Polemius Silvius numerous mid-fifth-century the of modern scholars for the A. suggestions (CIL i2:27i andInscriptionesItaliae 13.2, ed. Degrassi [Rome, 1963], of the massacre and the statements is error "Constantini" chronology (IV.2) 271), the result of scribal and hypercorrection: was written at some in the and then of the survivingliterary sources (IV.3). The legal,epi for "Constanti(i)" point tradition, "minoris" was added later to him from Constantine I,whose and evidence is considered next, distinguish graphic, papyrological was listed. 7 is the of Constantius II, or birthday already August birthday but with few exact conclusions I then as seen specific (IV. 4). can be from themid-fourth-century calendar of Filocalus (CIL on to not been , The pass other types of evidence that have i2:255 and 270 a Inscr. Ital. 13.2:253). See Aso Inscr. Ital. 13.2:492. a names "Constantinus" and "Constantius" are confused in considered before: victory title (IV.5), the coins (IV.6, frequently Greek and Latin texts. Indeed, one of the three manuscripts of Filo Appendix 2,and figures), and theitineraries (IV7, Appen calus's calendar (V) has a mistaken "Constantini" corrected to "Con dices 3 and 4, and the All the above map). description stantii" (see the photo in Inscr. Ital. 13.2:252) and both Polemius Silvius is then in a for "Constanti" and analysis brought together hypothetical and Filocalus have "Constantini" opposite 31March, a of the of Constantius I (CIL i2:26o-6i and Inscr. Ital. 13.2:243 reconstruction (V) and general conclusion major birthday in and 266). The "Natales caesarum" section Filocalus, though, has the points (VI). correct Con "Constanti" (CIL i2:255). Besides, the public celebration of never memoriae stantine I Is birthday would have survived his damnatio (CTh 11.12.1)into the middle of thefifth century.

5 PLRE 1:226, s.v. "Constantius 8"; Barnes, New Empire (n. 3 above), 8, 45; Kienast, Kaisertabelle, 314. 8 n. Kaiser In 6 Barnes, New Empire, 30,84; Kienast, tabelle, 306. gen 20 II. Prolegomena eral, see P. Guthrie, "The Execution ofCrispus," Phoenix (1966): 325-31; Historia H. A. Pohlsander, "Crispus: Brilliant Career and Tragic End," Constantinehad decided thatthe 2 By 332 clearly empire 33 (1984): 79-106; and Frakes, "Dynasty" (n. above), 94-95. and would be shared both branches s.v. New imperialpower by 7 PLRE 1:220, "Constans 3"; Barnes, Empire, 8, 45; Kienast, Is Kaiser tab of his father's family, the descendants of Constantius elle, 312.

DOP 62 8 R.W. Burgess I

to and full He needed Constantine's own was promoted augustus imperial power. daughter Constantina16 married away ofproviding them with theproper guidance and the to her half cousin Hannibalianus,17 the son of Flavius Dal in case he too soon. two even more empire with strong leadership died matius, thus linking the sides of the family For most of Constantine s sons 18 not two reign the surviving of closely (see stemma). On September 335, quite Theodora (hisfathers second wife) had been kept away years after the promotion of Constans, Hannibalianus from the center of power in virtual exile?Dalmatius and his elder brother, Dalmatius,18 were honored with to caesar inTolosa (modernToulouse) andJulius Constantius in imperial promotions, Dalmatius the rank of Corinth.8 This been to distancing has attributed Helena, with theadded titleof nobilissimus and Hannibalianus to seen s was who could onlyhave thechildren of her husband the nobilissimate.19 The latter also given the unique as to own son et second wife rivals her and grandsons.9 title of "rex regum gentium Ponticarum."20 in It seems certain at She leftfor Palestine 326,after the deaths ofCrispus almost thatConstantine planned, and and itwas in that while Constantine some future when were mature Fausta,10 year, date, they old enough and himself was in to two Italy celebrating the end of his vicennalia, enough, promote the eldest caesars, Constantine II that Constantius was to Constan to at once. Gallus born Julius and Constantius, augusti, probably both Thus, tius and Galla in not in Corinth. Helena or two Etruria, died upon Constantine's retirement death, augusti in and soon Constantine to two caesars in a early 329 afterward began and would succeed him recreated tetrar two bringhis surviving half brothersinto power. They chy,intimately linked by blood andmarriage. Itwould were in seem honoredwith consulships 333and 33$and with thatConstantine believed that dynastic succession ancient yet venerable titles, censor for Flavius Dalmatius (hencethe return of his halfbrothers and theirfamilies in or to (consul [cos.] 333)11 333 early 334,12 and patr?cius and favor and the appointment ofDalmatius) would solve nobilissimus for Julius Constantius (cos. 335),13 the first theinherent problems that had doomed theDiocletianic by 335and thelatter in September 335, when his nephews system.21 In addition the return of Theodora's children were to caesar Dalmatius and Hannibalianus promoted rex were 16 PLRE 1:222, s.v. "Constantina 2";Kienast, and respectively (see below).14 These swift and Kaisertabelle, 318. 17 PLRE s.v. high honors.A daughterof Julius Constantius and his 1:407, "Hannibalianus 2" Barnes, New Empire, 43; Kie nast, Kaisertabelle, 308. His name is "Hanniballianus" on the wife Galla was married to her half cousin Constantius II, spelled coinage. the son of Constantine, in and that same 335/3615 year 18 PLRE s.v. 1:241, "Dalmatius 7"; Barnes, New Empire, 8,45; Kienast, Kaiser His name is most on tabelle, 307. for the part spelled "Delmatius" the coinage (see Figs. 4, 24-25). 8 Barnes, Constantine and Eusebius . 2 251. The elder Hanni n. ( above), 19 Barnes, New Empire, 8, 28; Kienast, Kaisertabelle, 307 and 308. s.v. balianus (PERE 1:407, "Hannibalianus 1";Barnes, New Empire, 37), 20 PLRE 1:407, s.v. "Hannibalianus 2"; Kienast, Kaisertabelle, 308. Constantine's third half seems to have died before ca. brother, 333-35. For see the title, Anonymus Valesianus 6.35 and Polemius Silvius, Later 9 An connection between Helena and Constantius's min. He is on implicit Julius culus 1.63 (MGH,^4 9, Chron. 1:522). simply called "rex" time inCorinth ismade in a letter of to the Corinthians, the which was in Julian's quoted coinage, struck only Constantinople (RIC 7:584 and Libanius inwhich described S. Helena as his by {Or. 14.29-30), Julian 589-90, nos. 100 [silver] and 145-48 [bronze]). father's "wicked a stepmother" ( a). 21 For Constantine's tetrarchie see plans, Chantraine, Nachfolgeord 10 Barnes, Constantine and Eusebius, 221. . 2 nung ( above), 3-25. Most recently P. Cara ("La successione di Costan 11 of Domitius corrector Aevum has on the of Consul prior, ahead Zenophilus, former tino," 67 [1993]: 173-80) argued basis Constantine's consularis and of and favoritism toward his eldest son that Con Siciliae, Numidiae, proconsul Achaeae, Asiae, and Afri promotion apparent cae s.v. stantine intended for Constantine II to as {PLRE 1:993, "Zenophilus," and Barnes, New Empire, 106-7): only succeed him augustus, see S. et Consuls while the other caesars would remain as thus Roger Bagnali al., of theLaterRoman Empire (Atlanta, theywere, preserving the 200-201. as one 1987), imperial college ithad existed between 333 and 335with augustus caesars . 12 PLRE s.v. and three (on this, see Bleckmann, [ above]: 226 1:240-41, "Dalmatius 6"; Barnes, New Empire, 105. "B?rgerkrieg" . Caras from the as as one 3). hypothesis?first argued coinage long ago 13 Also consul prior ahead of who outranked him, Ceionius Rufius Cara n. 1949 (see [above], p. 173 2), before the publication of RIC 7 in Albinus, former consularis Campaniae, proconsul Achaeae and Asiae, obviated two 1966?is by fundamental problems. First, Constantine and from the very end of his consular year {PLRE 1:37, II had been caesar than his next eldest s.v. "Albinus and New see et longer colleague (Constantius) 14," Barnes, Empire, 108): Bagnali al., Con more than seven years. As a result he outranked Constantius and it suls, 204-5. by should therefore come as no was to a surprise that he able take victory 14 PLRE 1:226, s.v. "Constantius New 108. 7"; Barnes, Empire, title before of his caesarean He any colleagues. had after all resided in 15 PLRE 1:1037, s.v. 1"; New "Anonyma Barnes, Empire, 45; Kienast, Trier from 328 (after eleven years as caesar) and won the titleAlamanni Kaiser tab elle, 517. cus ca. defending his territory in 330,while his brothers remained with

DOP 62 The Summer of Blood | 9

not was to intothe fold would greatlyreduce if eliminateany steppedin before he able put thefinishing touches at on their on his problem of future attempts usurpation part. preparations. are 22 a Constantines plans forHannibalianus unknown, On May 337,while preparing for campaign to a to Constantine died in an but his title is clearly related Roman desire control against the Persians, imperial near an not far fromNicomedia the territoryof theArmenian kings.22 Constantines half villa Charax, emporion were no to an in A brothers doubt intended play important .25 fifty-year tetrarchie precedent clearly as a new role in the concilium senior statesmen, advisors, and prescribed that the proclamation of member of the even to since the or the of a caesar perhaps regents the young emperors, imperial college promotion required was his the of an or the active of the eldest surviving son, Constantine II, just shy of presence augustus approval senior situation inwhich either of these twenty-firstbirthday when Constantine did eventually augustus. Any summer two in the offend die inMay 337,having been born in the of 316. rules had been violated had resulted caesar or as a and In addition, Constantines trusted praetorian prefect, ing augustus's being regarded usurper was to Constantius after often also resulted in civilwar. When Constantine Flavius Ablabius, assigned open in as This Constantines death (probably Constantines will), died, theonly reigningaugustus died well. gave a in was intended to act as Constantine II and Constantius no constitutional means relationship which he clearly and Ablabiuss of from the earlier of guardian advisor; daughter, Olympias, becoming augustus, apart precedent to senate had earlier been betrothed Constans.23 proclamation by the army and acceptance by the was set: of Rome.26 No doubt each caesar worried For Constantine, the stage his legacy and and people on in an of Chris about this to because the policies would live unassailable college allowing happen unilaterally, it as In addi tian emperors, all related by blood and bymarriage, all other might regard attempted usurpation. was two caesars ablylooked after by elder and wiser counsel,and protected tion, there no guarantee that the young a and caesars. would remain or that their armies would allow by tetrarchie system of regional emperors content, territories been set aside as of them to remain as caesars in the con Even specific had spheres content, ensuing caesars more was the division of the activity for the four in 33s24 From Constantines fusion. Even problematic view was Fate certain territorial had point of the plan perfect. Unfortunately, empire. Although arrangements been made for the four caesars in 335,there was no reason to believe that thesewould continue after the as necessarily their father until 335 (Constantius's eleventh year caesar) and from were not not reside in areas that were death of Constantine.27 335,when they with him, they did Nor witness to hostilities (see Barnes, New Empire [n. 3 above], 84-86). come as coins were should it any surprise that special silver minted for s.v. New Constantine Us vicennalia, amilestone that none of the others reached 25 PLRE 1:224, "Constantinus 4"; Barnes, Empire, 8, 80; see within Constantines lifetime. Second, Cara has not considered all the Kienast, Kaisertabelle, 301. For Constantine's death, R. W. Burgess, available numismatic evidence. Even a brief of the and silver Studies inEusebian and Post-Eusebian Historia Ein study gold Chronography, " zelschrift and R. W. or coinsminted betweenthe end of 333and mid-337 (Appendix1 below) 135 (Stuttgart, 1999), 221-32, Burgess, II and Constantius were a e ?The Location and Circumstances of Constantine's Death," demonstrates conclusively that Constantine on were as even as a in n.s., 50 153-61. closely linked the coinage; treated equals, pair, JTS, (1999): of Constantine II's and were in that et romain spite seniority; together promoted 26 See J.-R. Palanque, "Coll?gialit? partages dans l'empire medium farmore than their This evidence, and that junior colleagues. aux IVe et Ve si?cles," REA 46 (1944): 54-55. See also N. Lenski, "The Chantraine (which somewhat), indicate that Con in 2 argued by overlaps Reign of Constantine," Lenski, Cambridge Companion (n. above), stantine was the and civil service, the audience for preparing army major 62 and Barnes, Constantine and Eusebius, 28-29 (with respect to Con the of these coins, for the eventual succession of both sons as his message stantine's accession). successors, Constantine II, as the member of the though longest serving and Or. 27 Barnes, New Empire, 198. Julian (Or. 1.19A-20A 2.94B-C) would nevertheless have been the senior and have college, augustus pre was most makes itquite clear that the division of the empire the important eminence over his brother as was the case among augustus, eventually matter when the sons did meet in Pann?nia. He twice that = finally says the three brothers Chantraine 19 and 24, 4 RIC 8 Sis (see [above], fig. a In the Constan they concluded "treaties" ( ; 19A, 2oB). event, cia nos. 18-18A and below n. 121). tine II, who would have controlled the entireWest with a subordinate 22 See G. Wirth, "Hannibalien: zur Geschichte eines Anmerkungen caesar under a tetrarchie system, lost the dioceses of Italia, Africa, and BJ190 (1990): 201-32. to one Dalmatius's two dioceses ?berfl?ssigen K?nigs," Pann?nia Constans, who also gained of s.v. Constantine and Eusebius Constan 23 PLRE 1:642, Olympias 1";Barnes, [n. (Moesia) from what would have been Constantius's territory. on eastern 2 above], 252; Bernes, New Empire, 45; Kienast, Kaisertabelle, 313. tius, focused as he had been the frontier since 335,would have been to have surrendered the difficult Danubian frontier to Con 24 Barnes, Constantine and Eusebius, 251-52 and idem,New Empire, happy stans. Constantine II, on the other hand, was less to have lost 198. happy Italy,

DOP 62 10 IR.W.Burgess

the obvious need for a end whose liveswere saved it is However, despite quick two, (so said) because of youth to the Constantine's more or interregnum following death, (Julian) expecteddeath fromillness (Gallus),35 all the sonswere than threemonths passed before Constantine's male descendants of Constantius I and Theodora had tomeet in finally able Pann?nia, accept promotion to been assassinated inwhat Libanius later called in one and establish their ter Or. a augustus together place, ("the great massacre," 18.10). Such slaughter ritorial on divisions and seniority. This then completed within thefamily of thereigning imperial family is unique two of the new returned to in 9 September, augusti their the annals of Roman history. area capitals, while Constans remained in the to continue themilitary activities begun thereby Constantius (see below, section IV.7). Unmentioned in any of the official event were proclamations of this happy Dalmatius, the fourth caesar, and his brotherHannibalianus, not because had been over for III. The Circumstancesand they passed promotion, but because they Responsibility had been assassinated. III.i. The and Evidence And these two were not ones to Literary Epigraphic the only die. Con stantine's two half Flavius Dalma No source an account of massacre surviving brothers, provides the and only tius and met as a Julius Constantius, also their deaths, did few sayanything specific about it at all:we have no Constantius s eldest whose name is no no no Julius son, unknown; chronology, context, causes, coherent narrative. four other cousins of whose identities are No one even states it some Dalmatius, where took place. In situations also unknown;28 Flavius and consul this is a result of certain authors' not Optatus, patr?cius simply knowing any Flavius of334?29 Ablabius, praetorian prefect of the East details; in other cases, however, our sources did know the and consul of 331;30 and "many nobles," who probably details and either assumed that their readers knew them as included Aemilius Arborius31 and Virius well orwere to Magnus possibly unwilling (or unable) provide them. Even and Flavius Felicianus.32 we an Nepotianus Julian, the closest have to eyewitness, avoids describ and the two sons of the actual events a line Julian, youngest Julius Constantius,33 ing by quoting from Euripides were not killed but were and raised from as spared apart {Orestes14): "Why should I now, thoughfrom a tragedy, the under the care of recount imperial family,Julian his maternal the unspeakable horrors?" (AdAth. 270D). We grandmother and the bishop Eusebius inNicomedia, and Gallus in were were sent Ephesus; when they older, both

furtherinto exile for six to an villa called in he was as in Rome years imperial 350 put up emperor against (for less in than a but ifhe was in not Macellum Cappadocia.34 With the exception of these month), born 337 he could have been more at a than twelve years old the time, fact that no source comments upon. He is on as a depicted his coinage bearded young man, but that means as he later the demonstrated (for division of the empire and its results, little in context since same the it is the portrait the mint used for Con see . Bleckmann, "B?rgerkrieg" [ above], 225-50). stantius. On see s.v. Nepotianus, PLRE 1:624, "Nepotianus 5" Kienast, 28 . Julian,To theAthenians (AdAthenienses[AdAth.]) 270D. Kaisertabelle ( 3 above), 321. 29 , New nova PLRE 18.10 History (Historia [Hist, nov]) 2.40.2; 35 Libanius, Or. (repeated by Socrates, HE 3.1.8). Itmay also be 1:650, s.v. 3." that the rescue of owed to "Optatus Julian (and Gallus, by extension) something the involvement of of was 30 , VS 6.^.9-iy, Zosimus, Hist. nov. 2.40.3; Jerome, Chron. Eusebius, bishop Nicomedia. Eusebius related can. to mother not to 234e; PLRE 1:3-4, s.v. "Ablabius 4." Julian's (though Gallus's), who died shortly after Julian's birth, and was under his inNicomedia afterEuse 31 with R. P. H. The Julian supervision (even Ausonius, Professores 16.9-16 Green, Works of bius was transferred to This could the state Ausonius (Oxford, 1991), 351-53; PLRE 1:98-99, s.v. "Arborius 4". Constantinople). explain ment that itwas a Mark of Arethusa bishop, (southwest of Amphipolis 32 PLRE 1:625 and 330-31? s-vv- 7" and "Felicianus 5"; "Nepotianus in who rescued as we Chalcidice), Julian, though, shall see, Constantius Barnes, Constantine and Eusebius, 261-62,398. later claimed the rescue for himself. Perhaps both Eusebius and Mark 33 PLRE and s.w. "Gallus and "Iulianus 1:224-25 477-78, 4" 29"; were in the for the funeral. See Ammianus capital 22.9.4; Gregory of and Kienast, Kaisertabelle, 318 and 323-24. Or. source Nazianzus, 4.91 (the ultimate for Theophanes AM 5853 [de 34 the son ofVirius and Boor and Nepotianus, Nepotianus (it seems) Eutropia, 48.8-11] Theophylact of Bulgaria, Martyrium 10 [see below, sister of Constantius and Flavius is not mentioned n. . Julius Dalmatius, by 45], PG 126:165c); Bowersock,y#//?? ( above), 23; Barnes, Con source as survived the not even . . any having massacre, by Julian, which stantine and Eusebius [ 2 above], 398 14; and T. D. Barnes, Athana would be most had he been alive. This that sius and Constantius: and in peculiar strongly suggests Theology Politics theConstantinian Empire was with him at the time of the massacre. Eutropia pregnant However, (Cambridge, MA, 1993), 105.

DOP 62 The Summer of 11 Blood |

must sources a , to begin by arranging the chronologically and 67.2), since the armies firsthad learn for them. ofConstantine's looking relationships among death throughmessengers (68.1) and they A of the sources shows that then had to communicate their decisions letters preliminary analysis through theyfall roughlyinto three main chronologicalgroups, ( a a ; 68.3). In spiteof the supernaturalelabora its own sources is to account each with Tendenz: early that either tion, Eusebius obviously trying for the gap the events or what is ofmore than three ignore altogether provide clearly months between May and September without its existence. theofficially sanctioned version, which laysthe blame actually admitting Nevertheless, on an mutinous writers over time uncontrollable, army; later who Eusebius then glosses the considerable lag he accuse mass e baldly Constantius of murder; and much has just described and calls Constantius ?a sources or at later that merely report reflections hints of the time of his father's funeral (70.2). But before this accounts in senate Rome sons the earlier sometimes fabricated and fanciful the and people of have proclaimed "his contexts aimed at certain or alone and no others as and supporting political religious emperors' augusti" (69.2: no a a a ... a a a a a viewpoints. Since modern scholar has submitted the a to a in to its we evidence chronological analysis order chart e?a ).Again have the constitutional legitimacy development,I shalldo sohere. This analysisis longand of the threebrothers stressed; not onlyhad theytwice it is in case complicated, but the clarity provides fundamental inherited the empire from their father,but anyone to or the final interpretation. found that insufficient suspect,they had been fully source events Our earliest for the surrounding the accepted by the army and the and people sons is as death of Constantine and the promotion of his well, these three groups being the traditional, and after Eusebius'sLife ofConstantine (VitaConstantini [VC] Constantine's death the only legitimate, bestowers of in 4.51.1,65-71), written the years immediately preceding imperial rights and powers. But, of course, their promo states at some sons not Eusebius's death inMay 339.Eusebius that tion of the three would have been necessary had as point after the end of his thirtieth year emperor and Constantine himself actually promoted them before his before his death (4.49 and 4.52.4), Constantine "divided death. Eusebius has mistakenly allowed the realityof a to the government of the whole Empire among his three proclamation by army, senate, and people intrude as a to he loved into his fiction of a uncontentious handover of sons, though disposing patrimony those smooth, best" (51.1; a a ?a e a a power. It is as ifneither Dalmatius nor Hannibalianus a e a , a a a a a a had ever existed. is in a ).Then on his deathbed,before the This absence particularly clear the portions of sons he bestowed Constantine to his sons before his assembled bishops and soldiers, "[o]n his empire that assigned as a estate fathers the inheritance of Empire, having death (VC 4.51.1),since this is infact the division of the as he desired" a a as itwas afterConstantine's not arranged everything (63.3: empire arranged death, e a a a e a a e before. It is even more obvious when one compares VC ' ?a , a a a a a a e ).36Then, 4.40.1-2 to a passage in his earlier oration on the occa in sion of thirtieth delivered afterConstantines death and the lying state, all the Constantine's anniversary, as "as if on soon afterDalmatius's accession Tri armies, throughout the empire, acting one, by 25 July 336, (his would rec cennial There Eusebius refers toConstantine's supernatural inspiration" declared that they Oration). sons alone of a caesar for each decade of his and ognize "no other than his [i.e.,Constantines] promotion reign as of theRomans" a e the of the fourth caesar (Dalmatius) for emperors f (68.2: proclamation caesars as a a a a a a a ) and then the fourth decade. He then describes the four e a caesars to e "soon" ( ) promoted them from yokedbefore the emperor's quadriga (?a ; must some consider is in VC augusti (68.3).Now this have taken Triac. 3.2and 4). Thispassage repeated the only in state a but now there are three decades able time afterConstantines lying (66-67, esP few years later, only caesars a a and the have become "like trinity, triple off All translations of Eusebius are from Av. Cameron and S. G. Hall: 36 of sons." see spring Eusebius, Constantine 1999). For this passage, the Life of (Oxford, stress on as an Furthermore, Eusebius's the notes on pp. 333-34. For the idea of the empire hereditary posses legitimacy sons in sion, see I. Tantillo, "'Come un bene ereditario': Costantino e la retorica of Constantine's alone rings hollow because his dell' tardive 6 251-64. no one to succession: impero-patrimonio," L'antiquit? (1998): version there is else challenge the

DOP 62 12 R.W.Burgess I

there isno or virtue in an to importance sayingwe'll take only with opportunity demonstrate their bravery and three and these three if three are on offer.The a a or a a alone, only ' courage ( a).37 e e a a e a over a winter and Just decade later,during the of 355-56, a in narrative in in to particular and the whole general Julian, his firstpanegyric Constantius, also mentions was to a ... therefore betray Eusebius's purpose. He knew thatmany that Constantius heir the empire ( ifnot most the existence , readers would know about and Or. 1.7D), and immediately mentions the removal of Dalmatius and he is for circumstances succession Caesar, providing involving the of Constantine I, an to them implicit explanation for his disappearance: he who, after his fathers death in 306, had succeeded was did not become augustus because it not thewish of the throneby the choice of his fatherand thevote of or senate Constantine, the army, the and people of Rome. all the armies. This entire narrative is to states a therefore intended explain the Later Julian thatConstantius had acted of sons to a legitimacy the promotion of Constantine's ("justlyand moderately") toward his broth and the absence of once a the s augustus Dalmatius, legitimate ers, citizens, his father friends, and the army, "except, heir selected without the dif ever not by Constantine, admitting if forcedby timesof crisis,you unwillinglydid ficulties involved in either. It is others e a e really quite ingenious prevent from doing wrong" ( ? in its a e e e deception. a a a e a , next sources are two The earliest panegyrics delivered Or. 1.16D-17A). Libanius and in also in the context by Julian 344/45and 355/56respectively. Julian mentions, of the begin Itmust be remembered the of the Persian throughout following analysis ning war, that military affairs had been were that both panegyrics delivered in the presence of thrown into great confusion in consequence of the Constantius himself, and this had consequences forwhat political change followingthe death ofConstantine, could and could not be said. In Or. 59.48-49 Libanius and thatthe soldiersshouted that they longed for their alludes to certain events that he fol to e implies immediately previous commander and theywished control (a ) lowed the death of Constantine and that he does not Constantius (18D). That Constantius's army mutinied mention when he his account of the summer of accession is a presents upon his surprising admission for his pan in same must comment 337slightly later the oration(?? 72-75),where he egyricist. It be that this appears for a very mentions at nothing untoward all. In ?? 48-49 he insists particular ulterior purpose. that in of themomentous notes s spite change that followed the Julian that afterhis father death Constantius death of "the of the was was surrounded Constantine, government empire by "dangers and manifold problems: not nor a disturbed, did anyof theevents affect the heirs of confusion, serious war, many raids, a revolt of allies, But while the in a a imperial power. government remained lack of discipline among the soldiers ( an itdid so not a a a orderly disposition, without degree of a), and other great difficulties at that time" (20B). trouble nor without the successors' tomake use of The war and the to having raids refer the Persian siege ofNisi violence to retainwhat had been to them" bis and Sarmatian incursions on theDanube in 337; the securely ' granted e e e a a e a e a a are ( a allies the Armenians (see 18D and 20D); the rest a e a e a e a a e must therefore ?e?a describe Constantius's other problems a a, He with the sons are 48). rejoiced because "they with the army. The parallels with Libanius's account received imperial power from their father and theyproved to e the concomitant tumult" 37 R. superior ( F?rster, ed., Libami opera 4 (Leipzig, 1908), 232.19 and 233.5, and a a a a e a , comment at 8. It has been tome that Libanius is s 49)?a he repeats suggested speaking ofConstantine II , revolt in this and Imust that I find this a attrac theend of 49, substitutinga e a for a a a a is passage, say particularly tive in aword thatmeans and can be interpretation, especially connection with what follows in sections politicalupheaval applied The is I can no reason 51-52. major problem that imagine why Liban to rebellion or civilwar. He thenmentions that faced they iusshould imply (more than once) thatthe revolt (spring 340) immedi some sort of "difficulty" ( ). Like Eusebius, Liba ately followed Constantine's death and the succession (337).Would not nius uses the sons as to a listener at the time have assumed that he was about events imagery of the the heirs imperial speaking at the time of the succession? See also the comments in favor of 337 power (oi a ,48, and , by R-L. Malosse, LibaniosDiscours (Bud? series; Paris, 2003), 4:192. As a are 13,where the third of Three result of this I have not on they generation heirs). uncertainty, placed great emphasis Liban times Libanius states crisis sons ius's comments here. that this provided the

DOP 62 The Summer of Blood | 13

name obvious. In this case, however, Julian says that before by inscriptions fromwhich his has been erased.41 to was as is Constantius returned Syria the mutiny ended and That Julius Constantius condemned well implied was on two order restored (20D). by his omission from Gallus's titles milestones no connection is as Although explicit made between fromGaul: he isdescribed only "diuiConstantii pii the army revolt and a time of crisis when Constantius Augustinepos" with no hintof whose "filius"he was (CIL not from and As we shall see both "unwillingly did prevent others doing wrong," 17.2:147 171). below, Eunapius it is is to Not is a clear that Julian referring the army. only andAmmianus implythat Julius Constantius played in to in causes massacre. the army the last the list, closest the exception, but leading role the of the on In states Julian immediatelygoes togive specificexamples of his panegyric Julian thatConstantius alone treatment his Constanti hastened toConstantine s sidewhile he was still alive Constantius's good of enemies, (Or No mention is summer nople, his brothers, and his friends. made 1.16D). But later,during the of 358,he extended as was of the army. The only situation described by Julian oi this claim (Or. 2.94A-B), saying that Constantius a at Constantines that as he Constan isthis crisis thebeginning of his reigninvolving favorite and laydying thesoldiers. When thearmy is finally discussed indetail tine summoned him alone; his brothers neither were sum nor came e e e (18C-D, 21B-22A),Julian is highly critical with respect moned (ol a e ). to its war in he Constantine then entrusted him with lack of preparedness for 337. When arrived in toCon e a This panegyrical account, delivered person supremepower ( a a a) and assigned accession was marred to stantius, openly admits that the him theappropriate portion of theempire govern(his from the to by crisis, confusion, and direct threats army own). OnlyZonaras (EpitomeHistori n, 13.4.28),writing the successors of Constantine. Libanius may be admit after 1118, repeats the claim that Constantine was still same if he over the alive when Constantius and he derives ting the thing, but not, then passes arrived, probably comment at it He notes difficulties of the succession without any from this panegyric.42 later the contradiction all. And as in Eusebius's account, Dalmatius Caesar is in his sources: some stated thatConstantine divided the sons missing. One could understand that the panegyricists empireamong his and othersthat they had divided to on but it themselves after his death would have been unwilling dwell his removal, amongst (13.15.1). accounts This insistence that Constantius was that hardly for his complete absence from both promoted by was a was main works. Constantine II is also missing from Libanius's his father before his death falsehood that it is Constan tained even the Constantius panegyric, wherein explicitly said that beyond panegyrical sphere: ever one It is as if celebrated an accession in of while tius only had the brother, Constans. anniversary May 357 never was s Constantine II too had existed. This is because he he inRome, exactly twenty years afterConstantine war on in even was not due until declared Constans early spring 340, and after death, though the anniversary his death in battle he suffered damnatio memoriae?* 8November 3s8.43 was to In a in honor We know that Eusebius always keen follow the later panegyric delivered of Eusebia, new who had in officially sanctioned version of history,whatever thatmay Constantius's wife,44 354 persuaded s was have involved. Crispus, Constantine eldest son,who executed in the or summer of never no. spring 326, appears 41 Ibid., 307. See, e.g., TituliAsiae minoris 3.1 (Vienna, 1941), 944; or and CIL in Eusebius's Tricennial Oration Life ofConstantine, AEpigr 1934.158; 1948.50; 6.40776. him hisHistoria 42 See Studies . 15 225 . 132. For as Zonaras's and Eusebius had earlier expunged from Burgess, ( above), Julian see II" 2 Note that ecclesiastica and doronici ca?ones, inwhich he had once source, also Lucien-Brun, "Constance (n. above): 595. Malalas says that Constantine I made Constantine II (sic) emperor of He too had suffered damnatio memoriae.40 was in appeared.39 Rome while he still alive the year 338 (Chronographia 13.15;Thurn, These instances that Dalmatius's non same account. strongly suggest 249), a confused version of the existence in and is also the the Eusebius, Libanius, Julian 43 See R. W. Burgess, "Quinquennial Vota and Imperial Consul NC Note that the result of damnatio memoriae, and this is indeed proved ship, 337-511," 148 (1988): 83-84. takes the hint and calls the celebration his vicennalia (Bonn, 542.19-20), even coins of time and the consulum though the Descriptio (s.a. 357.2; 38 CTh . 2. . R. and the Consularia Con W. Burgess, The Chronicle ofHydatius mark the 39 See T. D. Barnes, "The Editions of Eusebius' EcclesiasticalHistory\ stantinopolitana [Oxford, 1993], 238) correctly anniversary zi as his GRBS (1980): 197-98, and Burgess, Studies (n. 25 above), 66-74. thirty-fifth. . s.v. 40 Kienast, Kaisertabelle ( 3 above), 306. 44 PLRE 1:300-301, "Eusebia"; Kienast, Kaisertabelle, 317.

DOP 62 14 R.W.Burgess I

Constantius to allow to to to But Julian go Athens study, theKG also provided other important infor mation. was Julian praised Constantius for having saved him from It said that Dalmatius killed "factione so that no one without divine assistance militari dangers great [et]46Constantio, patrueli suo, sinente potius could have He also states thatwhen escaped (Or. 3.117D). quam iubente" (Eutropius).47 This claim that Constan "one a tius allowed or an assassination hishouse had been seizedby of thepowerful" ( a permitted instigated by a ),Constantius recovered it and made themilitary clearly parallels commentsmade byJulian two itwealthy again (118A).Echoes of these claimswill inOr. 1.16D-17A (see above) andAdAth. 271B and by be seen we later. Gregory ofNazianzus inOr. 4.22 (see below). As shall The connection to between the problems with the army see, by themid- late 340s this had become the official described above and the ofDalmatius was in disappearance explanation, when outright denial (as Eusebius and made forthe first time in the as Libanius Or. was no a or historicalrecord (as far $9.72-75) longer viable useful itcan be reconstructed)by theKaisergeschichte (KG), option. a set now-lost of imperial biographies that formed the This portion of theKG appears to have been writ a ten conclusion of large epitome history of Rome from about twenty years after the event, when this new times. was was current mythological Although thework updated explanation and the force of the damnatio the exact dates of had to thus a to periodically, making composition begun relax, allowing greater freedom the narrative of the events of was mention events as was to uncertain, 337 prob these (just Julian able mention written in Later witnesses show in ably 358. that theKG Constantine II his panegyric, whereas Libanius had stated that itwas the that had assassinated But is not an explicitly army not). this just mindless parroting of offi Dalmatius Caesar in themidst of a not as mutiny.45 cial explanation. The author does say, Julian and that was to Gregory do, Constantius unwillingly unable 45 For theKG, see A. Enmann, "Eine verlorene Geschichte der r?mi prevent Dalmatius's death; rather, he says Constantius schen Kaiser und das Buch de viris illustribus urbis Romae: it to viz. Quellenstu allowed happen: he wanted Dalmatius dead. dien," 4 Handbuch der lateinischen Philologus, suppl. (1883): 335-501; The KGs mention of an alternative forDal Literatur R. explanation der Antike, vol. 5, ed. Herzog and R L. Schmidt (Munich, = matius's death that Constantius ordered shows 1989), 196-98 Nouvelle histoire de la litt?rature latine (Turnhout, 1993), (i.e., it) and R. W. and the Historia that this claim was made at the and 5:226-28; Burgess, "Jerome Kaisergeschichte" clearly being time, 44 (J995): 349-69. For the date, see R. W. "On the Date of the the not towrite more Burgess, author, being able while Constan CPh and "A Common Source Kaisergeschichte" 90 (1995): 111-28, idem, tius still combines lived, the official explanation and the for Jerome, Eutropius, Festus, Ammianus, and the de caesaribus Epitome to create a between on accusation version between 358 and 378, along with Further Thoughts theDate and Nature private halfway of the CPh 100 see now the two. The of the statement in the form Kaisergeschichte" (2005): 187-90 (for pp. 185-87, framing of G. "Adrien the in Kelly, de Valois and Chapter Headings Ammianus Mar an alternative in the of = ("potius quam iubente"), place cellinus," CPh 104 The KG [2009]: 233-42). Eutropius 10.9.1 ("Dalma either a or rather no alternative at ... simple negative all, tius Caesar oppressus est factione militari"), Aurelius Victor 41.22 weakens what at first to be a defense of ("Dalmatius... mterf?cit\ir"),Epitomede caesaribus 41.18 ("Delmatius mili greatly appears tum ui necatur"), and Jerome 234e ("Dalmatius Caesar... tumultu militari interimitur"). For the importance of Jerome as awitness to theKG, see Bur Modern scholars fail to realize that all these works HE in Jerome. with their gess, "Jerome" (above). Socrates, 2.25.3 (repeated 3.1.8with material common account derive from a common from other such as see ultimately single source, sources, Libanius, Or. 18.10-11) derives from Eutropius; theKG. F. Die Geppert, Quellen des Kirchenhistorikers Socrates Scholasticus 46 The "et" is difficult and does not seem to I have followed (Leipzig, 1898), 67-68 and 119-20, and Barnes, Athanasius and Con belong. stantius n. HE F. Riihl text and F. L. M?ller (n. 35 above), 304 4. , 5.2.7; Theophanes AM 5829 (Teubner [Leipzig,1887], 73) (Eutropii Brevi?rium ab Urbe condita. Kurze Geschichte Roms seit (de Boor, 35.7-10 fromHE 2.25.3) and AM 5830 (de Boor, 35.12-16, from Eutropius, Gr?nd HE see v. n. in 3.1.8, also p. 48.11); and the late-eleventh-century Martyrium ung(7S3 Chr-364 Chr.), Palingenesia 56 [Stuttgart, 1995], 144) ss. illustrium of it. quindecim martyrum Theophylact, the archbishop of bracketing PG all derive from the latter also to Bulgaria (7; i26:i6iB) Socrates, employ 47 Eutropius is the only witness theKG who states this in thisman of Nazianzus Or. C. and R. Scott ner. was a ing Gregory 4.21 (see below). Mango Aurelius Victor, who writing just few years earlier than Eutro The Chronicle and Near Eastern in most ( ofTheophanes Confessor: Byzantine pius (finishing 361/62), for the part while Constantius was still AD attribute Theo felt itwiser to s History, 284-813 [Oxford, 1997], $6h) incorrectly alive, clearly avoid theKG specifics and says that Dal account to his matches matius was but "incertum phanes' directly Eutropius: wording exactly killed, quo suasore." Jerome, writing many Socrates' came via and the ecclesiastical epitome of Theodorus Lector. years later in 380-81, when theArian Constantius's direct involvement Cedrenus derives from Orosius and was as was (521.9) Theophanes. 7.29.1 Prosper accepted fact by all, baldly says he killed "factione Constan s.a. as well as other later Latin et ?1051, 338, many accounts, derive from tii patruelis tumultu militari."

DOP 62 The Summer of Blood | 15

Constantius accusations that he was in to against directly having often praised Constantius comparison Julian, combined with theKG's reference criticizes a responsible. This, nevertheless him for having saved Julian to two an a a e a the close kinship between the caesars, points a , at a a , a a a a a a accusing finger Constantius: "while Constantius, his ? own not so it as it to e a arms cousin, did much order allow hap ?a ("when the army took up against those offers a different from like out pen" quite picture something who held power, rebelling of fearof rebellion,and "and Constantius was unable to it." affairswere new prevent imperial being managed by rulers," Or. This accusation is context in was was made stronger by the 4.21).49 This something thatConstantius claimed which it ismade. The authorof theKG his control: a a a thoughtvery beyond [ ] ' of in of the circumstances e e e a a highly Dalmatius, and spite a ) ?a his death and the damnatio he was not e e e surrounding ("Constantiusdefended to so: Caesar at afraid say "Dalmatius prosperrima indole himself against those who had rebelled the begin ?eque patruo absimilis" ("Dalmatius Caesar, a man of ning of his reign,claiming that theyhad undertaken their action Here we can exceptionallypromising talent and very likehis uncle daring against his will," 4.22). even went so see new [i.e.,Constantine]"; Eutropius 10.9.1).He quite clearly the official version noted above? as to a in new far make very clear statement, direct opposi the rebellion of the army against the rulers and tion to was Constantius's to a the official version, that Dalmatius indeed inability control them?along with a and intended successor the death of to legitimate upon claimknown only fromJulians panegyric Eusebia: was one Constantine: "[Dalmatium Caesarem] patruus Con Constantius the responsible for saving the only consortem two survivors massacre stantinus regni filiis derelinquerat" ("His uncle of the (see also 4.3). As noted Caesar as co-ruler with his n. in same at Constantine leftDalmatius above (see 35), later the speech, 4.91, Gregory successores sons"; Jerome, 234e) / "[Constantinus] filios includes Mark of Arethusa, bishop and later author of tres unum reliquit atque fratris filium" ("Constantine the so-called "Dated Creed" of 359, among those who sons son as suc time entire lefthis three and the of this brother his rescued Julian "at the when [Julians] family was in a a a e a cessors"; Eutropius 10.9.1). danger" ( e), text Vic in The of the standard edition of Aurelius claim later found Theophanes and Theophylact, both we no tor (who know reliedupon theKG) adds,however, doubt from Gregory. no source It ismost to see what other notes, that the army had vigorously surprising the official explanation to accession in "obsisten in so as we objected Dalmatius's 335 (41.15, resurfacing Gregory many years later,when, a to its no one was in invec tibus ualide militaribus"), thus providing clue shall see, else mincing words. But his a a tive tries to revolt year and half later.Unfortunately, this reading against Julian, Gregory present Constantius an not in a so it suits to return to is emendation byMommsen, undertaken for any positive light and his purpose or but to the Constantian version of events. grammatical palaeographical reason, merely are sources to make the text conform to the army's supposed later upris The KG and Gregory the latest reflect in or the official version themassacre. The earliest ing against Dalmatius. The manuscripts fact indicate present of Dalmatius's acces account of the events that does not follow an that the soldiers strongly supported surviving comes sion ("assistentibus ualide militaribus").48 This makes official version from the pen ofAthanasius, bishop a him in of in his Ari the possibility of spontaneous mutiny against Alexandria, History oftheArians (Historia even anorum the of a 337 less likely and surreptitiously undermines the [Hist. Ar.]), "perhaps... surviving part excuse a no never or to for his death offered later, result the author work which Athanasius completed intended in its written was in doubt intended. publish present form," while he exile his in late around the time the relevant recension of the Gregory of Nazianzus, who finished writing 357, in was written was fourth oration just after the death of Jovian 364, KG being and Constantius celebrating

two at o "absistentibus" of n. similar 48 The manuscripts split this point, offering 49 Theophylact Bulgaria (see 45) says something very e e e a a 6 and p, "assistentibus." "Absisto" makes no sense with "ualide" and since about the army, ( ; must "a kind of desire for revolution fell the derived no doubt the initial "a-" is clearly part of the paradosis, p's "assistentibus" be upon soldiers"), correct was no from (the original doubt "adsistentibus"). Gregory.

DOP 62 16 R.W. Burgess I

in In it his anniversary Rome. he "states outright much Gallus's mother (Galla, the firstwife of Julius Constan that [he]deemed itpolitic to suppressor veilwhen he tius).These complaintsare laterechoed by Eunapius (VS was to or to a or a , writing defend justify himself neutral 7.1.6'. "Julians family hostile audience."50 he states that Constantius did was source Here, stripped bare"), though the ultimate for this not spare even his own kindred, but murdered his uncles may be Julian. We can thus see that as a result of these and cousins and did not commiserate with the assassinations Constantius sufferings benefited financially from ofhis wife sfather (see n. 15,above) or his otherkinsmen the estatesof both JuliusConstantius and Galla (and He about a of the soldiers. rest (69.1). says nothing mutiny probablythose of the of thefamily in addition). And not was in eastern s Now, only Athanasius Trier when the first since he obtained Thrace, the half ofDalmatius news Constantine's was in as of death arrived, he close prox territory,he benefited territorially well.51 to Constantine II and to In accusations us an imity anymessengers heading themidst of these Julian gives to Trier at that time and then to Constantius at assassinations were to shortly inside look how these justified afterward section a few two survivors. in court (see IV.7, below); moreover, only the young Those the imperial months after the massacre he himself passed through and those around his brother repeatedly commented where he no doubt would have heard e e Constantinople, ( ) thatConstantius had repented( e firsthand reports from those closest to the events. Itmust a ) andwas stungby greatremorse ( e ). was a also be remembered thatAthanasius writing private They saidhe attributedhis childlessnessand his failures was never to massacre document?it intended for publication?and against the Persians his part in the {AdAth. so he could saywhat others, even the author of theKG, 270D-271A). While Julianand his brotherwere inexile could not while Constantius still lived. at those who watched over them say Interestingly, Macellum, kept telling he does not his it is a accusation to e e e argue point here; simple themand tried persuade them(e , ) that without introduction or context that he Constantius clearly expects had done what he had done partly because all his readers to know and understand. he had been deceived (a a e ) and partlybecause We now must turn to in to an the only participant these he "had given way the violence and confusion of the who was a a a a a ... events, newly proclaimed augustus Julian, undisciplined and rebellious army" (? to summer or autumn a a a writing the Athenians in the of a a a e a ,271 ). his current break is to note two 361, justifying with Constantius. Julian It important that quite different excuses are a herehas droppedhis earlierpanegyrical pretense and, like provided here, each from different time Athanasius four years earlier, blames Constantius alone and source. In the second account that Julian narrates, forthe massacre, calling him e a ,a e , attributed to the time that he and his brother were at a e e ,a a e e a Macellum (ca. 342-48), we see the official version of a e e a ? murderer of Libanius's s ("the my father, and Julian panegyrics, theKG, and Gregory: not brothers, and cousins, the executioner of practically all Constantius had been able to control themutinous our common family and kinsmen," AdAth. 281B) and army that had assassinated Dalmatius and the others. him of six of their cousins to Even mirrors accusing having put death, the vocabulary that of the panegyrics. He we not along with Julian's father (Constantius's uncle), another adds what do find in the earlier sources, that the common uncle, and Julian's eldest brother, all without army had somehow deceived Constantius before or dur trial He to their rebellion. However in one (a ; 270D). wished put Julianand Gal ing the first explanation, to as in was at court was caesar lus death well, but theend justexiled them.In that offered by those when Gallus notes that Constantius had confiscated was addition, Julian (351-54), Julian implies that Constantius involved the and wealth of his Constantius in some in the since property father,Julius actively way deaths, this explanation carries on (273B),leaving Julian only his mother's house (probably directly from his earlier comment that Con the house he mentioned in Or that was taken stantius a 3.118A had executed his family without trial (270D, and then He also took the entire estate of returned). quoted above).What Constantius repented of and blamed

50 Both from Athanasius and Constantius New A quotations Barnes, (n. 35 51 Barnes, Empire (n. 3 above), 199. general indictment against 126. For the context of this see and for the Constantine and his sons at above), work, pp. 126-32 pas for shedding the blood of relatives is found discussed below see sage pp. 127 and 131-32. Caesars 336B and similar accusations are found in Or. 7.230A.

DOP 62 The Summer of Blood 17 | forhis laterfailures was themurder justdescribed. That A finalhint of theevents may be drawnfrom Julians to the courtiers admitted this and then tried explain his parable inhis ToHeracleius theCynic (227C-234C) of at states actions and describe his contrition suggest that by this 362,which, 228A-B, that the brothers wished time was an no one even to that his guilt undeniable fact that possess the empire for themselves, and after tried to conceal.52 Constantine's death the relatives of Constantine's sons was in over Clearly Julian, likeAthanasius, hostile toward became involved this squabble the division of sowe must sowere as in Constantius, be cautious when analyzing his the empire and assassinated, if fulfillment a reason emotion is a curse. The evidence. However, likely for Julians of tragic deliberately vague mythological context it to thathe honestlybelieved what he said:Constantius had unfortunately makes difficult go beyond as massacredhis familyand kepthim in a stateof virtual thebasic factsof thepassage, but itdoes look though rest were some and actual exile from the of his family for almost the murders precipitated by controversy themassacre over the division of in the immediate aftermath fifteen years.He describes his feelings about empire out It is to note inTo Heraclius theCynic 230A: when he found what of Constantine's death. interesting that was not at comments are had happened (theimpression is thathe told Julian's about his relatives here decidedly a e a a a first)he was sooverwhelmed that he felthe had to throw negative: they shared the a a ("stupidity e a e a a sons the himself into Tartarus (e and ignorance") of Constantine's concerning a e a e a e see no on ). These governing of the empire.We attempt Julian's are we can still in the letter to theAthe towhitewash or exonerate his in of feelings detect part relatives, spite nians. in of the of their fate.He condemns them all. And, strangely enough, spite enormity as these crimes, he writes about them just he does about After the death ofConstantius, Libanius, freed of the Constantius and his in the other complaints he has against necessityof followingthe official line, Julian'sfuneral treatment statements cannot be oration described the deaths as "the massacre" by him. Julians therefore great (0 as one out to a went entire dismissed merely the invention of blacken ), swordthat throughalmost the to reasons sons he his opponent. He is trying explain the for his family, fathers and alike (Or. 18.10).Constantius, as causes was themurderer of father and and rebellion against Constantius (he also gives said, Julian's brothers, as a a Constantius's removal of his Gallic troops and the petty result Julian suspected plot when Constantius, the But man who had him the offered tomake accusations Constantius kept making against him). wronged most, stresses even in him a thatwas overall Julian his loyalty and service, the caesar, suspicion prompted by previously faceof everything that had happenedto him. His attitude spilledblood (Or. 18.31-2). can seen most in ver be clearly his discussion of his brothers A clearerview of thereality behind theofficial no to is written in execution by Constantius. He makes attempt deny sion providedby theHistory ofEunapius, or as indeed soon mitigate his brother's cruelty caesar, and itsfirst edition after378, which (via thehistory of as his states the death of agrees that he deserved execution punishment for Zosimus) that, following Constantine, s to two as deeds. He does attribute Gallus evil behavior his early Julius Constantius and his nephews acted col ofConstantine's sons and that Constantius imprisonment by Constantius and blames Constantius leagues Julius not a chance to was the first to be killed the then Dalma in particular for having given Gallus by soldiers, not defend himself (271D-272D), but he does exagger tius (the caesar), Optatus, and finally Hannibalianus case to Hist. nov. and He states that ate or try to twist the facts of the his advantage (Zosimus, 2.39.2 40.2).53 are we see in (and these facts that know). I nothing his discussion of the massacre that would suggest that he 53 at leasttwice in his Lives the statesthat was or false accusations to Eunapius of Sophists(VS) dissembling making justify events e more in his account he described these a ? ("in detail") his revolt Constantius. Indeed, his account of the an overview see against of Julian, i.e., his history (6.3.8 and 7.1.5). For of the VS, excuses when he was D. F. "Lives the A 62 different given younger strengthens Buck, of Sophists: Literary Study," Byzantion (1992): a to his see R. C. The the trustworthiness of his account. 141-57. For general introduction history, Blockley, greatly Historians theLater Roman Euna Fragmentary Classicising of Empire: ARCA 6 pius, Olympiodorus, andMalchus, (Liverpool, 1981), 52 also mentions Constantius's 5853, de Decline of Theophanes repentance (AM 1-26, andW. Liebeschuetz, "Pagan Historiography and the Boor, a section that derives from ofNazian in Roman in Late 47.2-3), ultimately Gregory the Empire," Greek and Historiography Antiquity: zus, Or. 21.26 to (see below). Fourth Sixth Century A.D., ed. G. Marasco (Leiden, 2003), 177-201.

DOP 62 18 R.W.Burgess I

e a a is Is [ a ] ? a identifiable. After naming Constantius children a a e a a by his second wife, Theodora, the anonymous author a a ("[Constantius] commanded the soldiers adds, e e a a a a to out content no e a of whom was to cry that they would be with other ?a ("none considered worthy sons a successor is ruler than the of Constantine," 2.40.3). Eunapius be of their fathers rule," p. 312.8-9). This names as a victim as amore a same Ablabius well, but he provides clearly reflection of the propaganda reported by account in it nor detailed of his death another, laterwork (the Eunapius, though derives from neither his work not is no mention in Lives ofthe Sophists, discussed below) and does saythat Zosimus's. There thisLife of the deaths hewas killed by thearmy. Eunapius (viaZosimus) thus in 337 or of Dalmatius as caesar. Let us return to even the provides the link between Constantius and the army, and Eunapius, though follow two versions massacre. is out Because thus the link between the of the ing analysis of chronological order. of the in not an s Eunapius, writing favor of Julian, is unbiased survivalof Eunapius Lives ofthe Sophists (VS), written in source, but the clear links with the official version and 399,we know some details about the death ofAblabius, in account is he that of Eusebius argue strongly favor of his accuracy. though Eunapius's quite hostile, because In was of his sources fact, Eunapius confident enough blamedAblabius forengineering the downfall of the was even to cite in main that he able the order which the philosopher Sopater {VS 6.2.12,3.7,13; Zosimus 2.40.3). states was victimshad been killed. Eunapius thatAblabius dismissed byConstan it suits own tius Because his anti-Constantinian bias, immediately after the death of Constantine, who also an bit Constantius to no Eunapius unwittingly preserves additional had "bequeathed" Ablabius, doubt of about at the same time as so the latter could act as propaganda, likely spread praetorian prefect, guardian, accounts massacre. He states to the official of the that after and even, perhaps, regent. Ablabius then retired his Constantius I in estates in no to a the death of 306 the praetorian guard Bithynia, having wish be emperor, fact selected as his successor Constantine, whom he pres that amazed everyone ( a a a a ents as a "none e e e a bastard,54 because they thought that of ?a ? ), the implicationbeing thatit was was to [Constantius's] legitimate children worthy of the expected that he would attempt retain his almost e a a a ... a over empire" ( emperor-like position Constantius. But Constantius a a e a a e ?a a , wanted to be ridof him and thatpossibility for good Zosimus Constantius's children" are so senta to 2.9.1). "legitimate and groupof assassins( ) dispatch the sons of Theodora. This sounds much like very part him. Taking advantage ofAblabius's natural arrogance, of a excuse to were Constantius tricked him into secondary explain why they again declaring himself emperor, denied a share of the in was cut power 337. and Ablabius down when he demanded thepurple The excuse in cen are wide currency of this the fourth (VS 6.3.8-13). The other assassinations onlymentioned is its in an at tury demonstrated by reappearance indepen later 7.1.5 ( a a a ; dent source of the ninth or tenth a "all century, hagiographie those assassinated byConstantius"). Although Euna life of Constantine, referred to as the "Guidi does not mention it,Constantius then confiscated usually " pius or s as Life," after its only editor, the di Costantino," Ablabius property, he did that of the other victims s account no s Guidi's title.55Much of the.Life of Constantine (noted above).56 Zosimus makes mention ofAblabius is fanciful or derived from a sources few surviving like supposed usurpation (2.40.3). sources Theophanes, but information from other, earlier AlthoughGregory of Nazianzus had earlierutilized the official version of the massacre when it suited his For the date of the see T. D. history, Barnes, The Sources of theHisto ria in his oration in of Athanasius in Augusta (Brussels, 1978), 114-23; Blockley (above), 2-5; and Liebe purposes, praise 380 schuetz he at (above), 179-91. stated that the end of his lifeConstantius repented 54 SeeW. "The of Constantine and the Birth of three crimes were Leadbetter, Illegitimacy that unworthy of his reign. The first of the in Constantine: ," History, Historiography and Legend, . ed. S. C. Lieu and D. Montserrat 74-8$. in was (London, 1998), 56 Ablabius's house Constantinople still in the possession of the M. state in the and when 55 Guidi, ed., "Un di Costantino," RendLinc, 5th series, 390s early 400s Galla Piacidia lived there (Syne 16 = see R. (1907): 304-40, 637-62 BHG164.. See also S. Lieu, "Constantine sius,Ep. 61; Janin, Constantinople Byzantine, 2nd ed. [Paris, 1964], The and S. I. "Some Problems in the of Galla Byzantinus: Anonymous Life ofConstantine (BHG 364)," in Lieu 304, Oost, History ," and Montserrat, Constantine, 97-146, esp. 97-106. CPh 60 [1965]:3 and 9 n. 14).

DOP 62 Summer The of Blood | 19

In was ("themurder of his family";Or. the fifth century, Socrates derived his description a excuses s 2i.26).57 This is reflection of the later made ofDalmatius death from Eutropius {HE 2.25.3 and 3? ?8; are see n. in in a during the reignof Gallus that reportedby Julian 45, above), but book three, passage that in his letter to theAthenians. derives from a of sources, he states that when variety' a a trans Jerome,writing in 380-81,employed the KG forhis Constantius's (perhaps best nn. lated in the context as "hostile attitude toward description of the death of Dalmatius (see above 45 them," two and 47), but he also had a source that mentioned the i.e., his half cousins) had abated, he allowed them death ofAblabius.58 He statesthat Ablabius was killed to be educated (3.1.9), a comment that has no known and he source but is more than an inference along with "multi nobilium" ("many nobles"), perhaps nothing accession ofConstantine s on Socrates' And also on places the entry right after the part. finally , relying three sons and before themurder of Dalmatius. an unknown source, says that "Constantius... killed his was a ... Ammianus Marcellinus, who writing between relatives because he feared usurpations" ( ca. 385 and 390/91,59had no doubts about Constantiuss e a a e e a a a a s involvementin the murder of therest of his family.Like so [HE 3.2]).This may be relatedto Gregory claim {Or seems to many others,Ammianus had no love forConstantius, but 4.21), but it not be. Note that in the VS Euna we not or comments also mentions with toAblabius. do know what details he provided pius usurpation, respect hemade when he discussedthe events of 337in one of the These hintswill be discussedbelow. to be no earlier now-lost books of his history.60 In what survives After Theodoret there appears further inde even cannot he states that Constantiuss cruelty surpassed that pendent evidence. My analysis stop here, however, in imitation for I have omitted two accounts that are often considered of Caligula, Domitian, and Commodus, cune tos et factual to one or another modern scholars and ofwhom "inter imperandi exordia sanguine degree by se at the that omission must be and defended. These genere contingentes stirpitus interemit" ("right explained of his he eradicated who was accounts concern Constantine's will and derive beginning reign everyone ultimately The from the official claim thatConstantine his mind connected with him by blood and birth," 21.16.8). changed is since Gallus about the succession on his deathbed. In Eusebius he exaggeration patent, Julian and survived, s at intentions to witnesses. Consti as did Constantius brothers. Later 25.3.23 he explicitly proclaims his orally many mind was describes the cause of the assassinations: "[Iulium Con tutionally and pragmatically this change of open excessum com to or so in stantium] post fratrisConstantini inter doubt interpretation, later pro-Constantian successorum" accounts of his Constantine makes his wishes plures alios turba consumpsit imperii ("after death, a written Two related versions of thedeath of his brother Constantine [Julius Constantius] known through will. was others in an involv in thefirst half of thefifth killed along with many upheaval thisstory appeared century, successors to "Turba are in much earlier. ing the imperial power"). imperii though they origin wrote successorum": itwas a conflict that involved Constan The earliest version appears inRufinus, who his translation and continuation of Eusebius's Ecclesiastical tine's successors, particularly Julius Constantius (whom notes as to not his around His account is followed Eunapius being the first die), simply History 400 (10.12). by Hannibalianus. Ammianus HE and Sozomen, HE nephews Dalmatius and says Socrates, 1.25.1-5, 39.3 2.2.2-3; an HE nothing about attempted military coup. 2.27.2-4,34.2; 3.1.3; Theodoret, 2.3.1-7; Ps-Gelasius, itwas HE 3.12; and many others. No doubt originally is source of Chron. AM Boor 57 This the Theophanes, 5853 (de 47.2-4), similar accounts made to bolster the inspired by legiti missed and Scott, Chronicle (n. 45 above), 80. byMango in macy of the succession of Constantine 306 following 58 Chronici ca?ones 234e. The source was a kind ofDe uiris Jerome, to men of see R. W. the death of his father, Constantius I.61 illustrihus of literary of the reign Constantine; Burgess, According An Introduction to his Chronicle and a Guide to its this before he died Constantine "Jerome Explained: account, just composed 16 28. For Ablabius's tal Msz" Ancient History Bulletin (2002): literary hiswill, inwhich he leftonly his sonsas his successors, ents, see PLRE s.v. "Ablabius also 2, s.v. "Ablabius 1:4, 4" (see p. 3"). as in the earlier no mention is though, again accounts, Ammianus 59 J.Matthews, The Roman Empire of (London, 1989), made of any other heirs. This will he entrusted 20-27 and 31. possible see 60 These books were more substantial than is usually supposed: on the of theLost Books ofAmmi 61 See Constantine and Eusebius . 27. Cf. Lactan R. M. Frakes, "Some Thoughts Length Barnes, ( above), De and Or anus," The Ancient World 31 (2000): 48-53 (with full bibliography). tius, mortibuspersecutorum 24.8 Julian 1.7D.

DOP 62 20 R.W. Burgess I

to a to one priest, who had been recommended him by his made by theninth-century Byzantine bibliophile the ex-wife of Licinius. This out to as a sister, priest turned Photius, the other appearing excerpts within ninth be an Arian who had wormed his way into Constantia's century saint s life called thePassio Artemii.6* confidences. He thewill safe and when Constantius to set out kept According , Constantine from arrived afterConstantine's death he entrusted it toCon as Constantinople against the Persians but only got far who "in his desire for the realm was on account as forwhile was was stantius, Nicomedia, he there he poisoned of this favour so bound to him ... that he cheerfully by his brothers. Constantine realized too latewhat had to was an allowedhimself be governedby him" (Rufinus,HE happened, but before he died able towrite account It is this who was then a 10.12).62 priest solely responsible ofhis brothers'deeds and append to it requestthat the for Constantius's descent into because of Con sons heresy firstof his toobtain thedocument should avenge his stantius's debt to him. death, lest all three sons suffer the same fate. This docu In its this was to ment was to origin story clearly designed entrusted Bishop Eusebius of Nicomedia, the of the sons' a it out strengthen legitimacy promotion, result who kept of the hands of Constantine's brothers must ofdoubts that have existedin theyears following until the return of Constantius, who read the account as news Constantine's death, especially detailed about and fulfilledhis dyingfathers last command.64 the massacre As it it has been This account as spread. stands, though, again depicts Constantius alone the modified to absolve Constantine of the blame for murderer of his but itwas having relatives, cleverly designed an son: raised Arian the earlier pro-Constantian version to absolvehim of theblame forhis role in themurders has been to a later use. a put pro-Constantinian through the employment of version of thewill story seen states Rufinus thatafter the will had been entrusted above.That blame iscleverly shifted to Constantine s half to theArian now to priest, brothers, revealed be poisoners and traitors, and Constantius ispresented simply as thedutiful and pious etiam erant in son his fathers to [Constantio] quod eunuchi,qui palatio, following orders avenge his untimely arte morte fauebant, indicio de imperatoris subpresso death (andprevent his own and thoseof his brothers). ad Constanti multis noua This is usque praesentiam, temp reworking quite ingenious,65 but it isfictitious and, tantibus res tutae mansere. in the context of obpressis integraeque the development of the various accounts described its are above, purpose and origin clear. Since the palace eunuchs were also on [Constan This version takes no of, and fails to news was cognizance rely tius's] side, of the emperors death skillfully upon, the official version of events. This indicates that untilConstantius' who tried suppressed arrival,many itwas not as was early, the version of the story to seize were original power put down, and the state remained employed by Rufinus; itmust then have arisen at a time safe and undamaged. excuses when these had been abandoned and everyone, or s secretly openly, accepted Constantius role in themas As in ofNazianzus Euna Gregory (and Theophylact), sacre and so was no a outright denial longer plausible the and we have a reference to pius (in VS), Theodoret, Ifwe can option. judge from Julian, Athanasius, and noua but it is so usurpation ("multi temptantes"), vague, theKG, the earlier excuses were the source so being dropped by the late, and the story so fabricated that patently mid to late350s and so thisstory has its littlecan bemade of it theobvious probably origins beyond attempt(left then or afterward. over from to the original version) shelter Constantius In spite of this, too many modern scholars have been fromany accusationsby layingthe blame with palace eunuchs who thus the state kept "tuta integraque." 63 i6 and i6a. For the Passio see The most Philostorgius Artemii, R. W. Bur interesting version of the story told above gess, "The Passio Artemii, and the Dates of the Inven is that theArian writer inhis Philostorgius, reportedby Philostorgius tion and Translations of the Relics of Sts Andrew and Luke," AB 121 Ecclesiastical was History, which probably written in the (2003): 13-23. The remains two 64 This is theGuidi see 440s. of his history appear in epitomes, story repeated by Zonaras, life, and Cedrenus: Burgess, Studies (n. 25 above), 226 and n. 13$. 65 For instance, theArians knew and made much of the fact that Euse 62 The translations here and below are from P. R. The Church on Amidon, bius baptized Constantine his deathbed and so he has been substi History ofRufinusofAquileia (Oxford, 1997), 24-25. tuted for the Arian priest that we saw in Rufinus's account.

DOP 62 The Summer of Blood | 21

to one extent or even to the swayed by this story another, extreme it to a account. It of believing be wholly factual more Arian is nothing than propaganda.66

III.2. The Numismatic Evidence . Fig. Constantine II, 334. Silver miliarensis, 5.09 g (1:1) The evidence we have for date as RIC and same only exactly contemporary Siscia, unpublished (same 7.231-31 type one never as RIC 7 Trier the events of 337 has, with exception,67 been 581) This is Obv.: CONSTANTINVSIVN NOB C analyzed, and that is the coinage. unfortunate, Laureate and cuirassed bust because the coins struck between the and the right beginning Rev.:CONSTANTINVS CAESAR /SIS of are of in to cer end 337 great importance helping put Four standards sources a tain aspects of the literary into sharper focus and into a more accurate context. accession During the period between Dalmatiuss as caesar in in September 335 and Constantines death mints coins and seven May 337, six regularly struck gold silver in the names of the regularly struck emperors.68 not Of these mints, only three did strike precious metal in name the of the of Dalmatius: Trier, 2. coinage capital Fig. Constantine II, 337. Silver miliarensis, 3.95 g (1:1) Constantine the mint of with II; Rome, major Constans;69 Constantinople, unpublished (cf.RIC 7.124 CONS0) Siscia Obv.: and , the capital of Constantius. (a very anepigraphic rare Heraclea Diademed head right, looking up gold solidus type only), Thessalonica, (a Rev.:CONSTANTINVS CAESAR /C ? I rare silver and Nico siliqua type only), Constantinople, Four standards silver in his media (silver only) struck gold and name, at that including medallions Constantinople (at point directly controlled by Constantine). The absence of coins struck in the name ofDalmatius into a coins from is brought sharper focus by group of a Trier, Rome, and Antioch, part of series that appeared with reverses from many mints in both gold and silver, the depicting either four standards (representing emperor Fig. 3. Constans, 335-36. Solidus, 4.51 g (1:1) and three caesars, a revived from 324) or type Victory Constantinople, RIC 7.97 a often with advancing with wreath and palm frond, Obv. FL CONSTANS NOB CAES for various Laureate, and cuirassed bust anepigraphic obverses (see Figs. 1-4 examples). draped, right Rev.:CONSTANS-NOB CAESAR /CONS wreath and frond . Victory advancing leftwith palm 66 See, for example, Di Maio and Arnold, "Per Vim* ( above); "Con Bowersock,Julian, 13; Browning, Emperor, 34-35; Lucien-Brun, stance II," 600; Barnes, Constantine and Eusebius, 262 (all n. 2 above); 162 n. Hunt R. T. Ridley, Zosimus: New History (Sydney, 1982), 103; and 2 . ("Successors" [n. above], 4). More wisely, Olivetti ("Sulle strage" [ 2 it F. Paschoud Histoire above]: 75) calls "leggendaria," and (Zosime: 1 See Nouvelle, vol. [Paris, 1971], 244) "une fable (d'origine arienne?)." e la tradizione" . especially Tantillo, "Filostorgio ( above). 67 Di Maio and Arnold, "Per Vim," 194-95. a solidus forConstantine's tricen 68 Aquileia produced only single type Fig. 4. Dalmatius, 335-36. Solidus, 4.57 g (1:1) and silver but with obverses. nalia. Lyons Aquileia produced anepigraphic Constantinople, RIC 7.98 a silver with Constantine ITs name on Aries produced only single type Obv. FL DELMATIVS NOB CAES the reverse. These mints are therefore not considered here. bust Laureate, draped, and cuirassed right at it no mint. Rome was 69 Constans probably resided Milan, but had Rev.:DELMATI-VS CAESAR /CONS mint at time. wreath and frond the major Italian the Victory advancing leftwith palm

DOP 62 22 R.W.Burgess I

were a At Trier, four silvermiliarensis types struck with slightlylater issue).71 This indicatesthat specificincident and his three must have taken inwhich Constans wanted his obverse legends naming only Constantine place sons reverse Constantine audience to believe that the had been and legends again naming only empires security sons nos. or and his three around four standards (RIC 7 maintained protected. serieswas on and most there is the evidence of a 580-8$). In Rome this produced smaller sil Finally, important, ver obverses reverseswith twin issue of from to and siliquae with anepigraphic and coins, extending 337 340 clearly or three a vic linked to the aftermath of themassacre. Soon after the advancing Victory with palm branches (for a torious augustus and two caesars, revived type from 324) deathof Constantine (thechronology will be established his two eldest sons below in section I two new obverse and reverse and legends naming Constantine and V.6) types were on nummi from the mints of the (nos. 377-80). On theAntiochene siliquae the obverses issued bronze are reverses chief residences Trier also anepigraphic and the advancing Victory emperors' only: (Constantine II), name Constantine, Constantine II, and Constans (nos. Rome (Constans), and Constantinople (Constantius II). coins were Con On obverse of one is a of Theodora 105-7). No doubt originally produced for the type portrait stans at Rome and Constantius at Antioch, since these with the legendFL MAX THEODORAE AVG; on were Even an as their home mints, after all. though these silver the reverse, image of Theodora maternal Pietas, coins were no a on arm produced after his accession, similar type standingright and holding baby her left andher was struck in the name of Dalmatius at these mints. rightbreast with her righthand, with thelegend PIETAS Trier, Rome, and Antioch did, however, strike small ROMANA (seeFigs. 6-10).72The dativeof the legend s as commemorative bronze coins inDalmatius name; thesewill be discussed marks this type and posthumous. was on A commemorative was below (seeFigs. 24-25 for the type).It only the parallel type also produced for of the three sons of prestige issues, those intended particularly for imperial Helena, grandmother Constantine, payments to the army and the civil service, that the exis with theobverse legend FL IVLHELENAE AVG. The tence as a caesar was in reverse was PAX of Dalmatius implicitly denied legend, however, PVBLICA, with Hel ena as an a transverse theperiod beforeMay 337.Thus fromthe very begin Pax, holding olive branch and spear sons not seem not to two were ning, Constantine's only have fully (see Figs. 11-16).73 These types essentially the no matter accepted the legitimacy of Dalmatius and viewed him same, obverse and reverse, what themint, with as an to crown interloper, but also appear have communicated the exception of the type and minor differences with one another on this on a com in break.74 point and agreed the legend This indicates central design and mon response. Itwould therefore appear that they did dissemination. not accept the rehabilitation of Theodoras side of the

and the honors theirfather had his half names s family high paid 71 RICj:^4^ynos- 402-4, issued only in the of Constantine sons. reverse on brothers and their families. This continued the nummi into later 339 for only Con a or stantius and Constans the renewed GLORIA EXERCITVS The suppression of real perceived threat against alongside is a reverse on type (RIC8:250-51). the brothers implied by type small 72 RIC 8 Trier nos. Rome nos. 28, bronze nummi70 from themint at Rome 43, 48, 56, 65, 79, 91 (pp. 143-44); (controlled by nos. see 54 (pp. 250-51); Constantinople 36,50,51 (pp. 449-50); also pp. Whereas all issues of nummi from the other Constans). 6-7, 79-80,126,234, 442. Theodora is otherwise unattested as augusta not in in mints retain theGLORIA EXERCITVS one-standard (she does appear the list of augustae Barnes, New Empire [n. 3 Kent typesissued since the end of 335(described below, IV.6; above], 9). suggests that she received the title posthumously (RIC but there isno evidence that shewas not before see at Rome the reverses of all nummi issued 8:3), augusta mid-337 either, Figs. 20-25), so I leave thematter to one side. it be most However, would revealing if between the death of Constantine and the of to at promotion the brothers promoted her posthumously augusta this time. thethree caesars toSECVRITAS REI RIC STrier nos. Rome nos. change PVB(licae), 73 42,47,55,63,64,78,90 (pp. 143-44); 27, with a of Securitas on a column a nos. figure leaning (see Fig. 5, 53 (pp. 250-51); Constantinople 33-35, 38, 48, 49 (pp. 449-50). or 74 Trier: Helena: diadem with pearls, Theodora: laurel diadem; 70 The standard name of this has no ancient Rome: H: diadem with Th: coin, follis, authority. pearls, diadem with pearls; Constantinople: Here I use the later nummus, which means a "bronze H: Th: generic term, just diadem with and without pearls, diadem with pearls. In addi coin." this date these small nummi were worth 100 denarii each has minor variants in By probably tion, three regarding the break the obverse and were therefore almost the coins that are later referred to as variants at certainly legend (forHelena, the appear only Constantinople, and all R. S. in as centenionales (see Bagnali, Currency and Inflation Fourth Cen three variants appear for Theodora there well [note that variant T2 turyEgypt, 'ASP'Supplement 5 [Atlanta, 1985], 34, 44). was missed inRIC 8] ).

DOP 62 The Summer of Blood | 23

were At Trier these coins produced in great numbers duringthe interregnum (see Appendix 2). After that there were sixmajor changes ofmintmark between September were 337 and April 340.75The Helena and Theodora types were struckwith all six of thosemarks, meaning that they Fig. 5. Constantius II, 337. JE 3,1.66 g (1.5:1) struck routinely the two-and-one-half year throughout Rome, RIC 8.7 one period. Furthermore, they are,with exception (RIC 8 Obv.: D FL CONST-ANTIVS AVG no. as common as or more common than the rosette 82), (usually) Laureate and diademed bust right, draped and cuirassed PVB other types struck at the same time (theRoma and Con SECVRI-TAS REI /R$S names Securitas standing facing, head right, legs crossed, holding long stantinopolis commemoratives, and those in the in on a scepter right hand, left elbow leaning column of the three brothers and their father; nos. 37-92). On the other hand, mintmarks and hoards show that, unlike theTrier issues, the Theodora and Helena types were not from Rome and Constantinople struck until after the brothers had been proclaimed augusti. At Rome, Trier went six which like through changes ofmintmark (RIC 8 nos. 2-5$), theHelena and Theodora typeswere two struckwith only of those marks, the second and the Fig. 6. Theodora, 338/39-40. JE 3,1.68 g (2:1) last, thus in 337/38 and early 340 (nos. 27-28, $3-54). In Constantinople, RIC 8.50 cases were in both they struck small numbers, less than Obv.: FLMAX THEO-DORAE AVG or to the other obverse with the same mark. equal types Bust with plain mantle and necklace, hair elaborately dressed Rev.: PIETAS-ROMANA /CONS8 At Constantinople themintmarks changed much less fre an at 8 nos. two Pietas facing, head right, carrying infant her breast quently (RIC 23-54), but the issues forHelena and Theodora, and the one forHelena alone, were much smaller than those with other obverses, and the numbers issue were for the firstHelena much higher than those forTheodora (nos. 33-36, 38, 48-51). Thus, as at Rome, were these types produced much less frequently and in at at smaller numbers Constantinople than Trier. were ca. These coins produced until April of 340, when Constantine II was killed in his civil war against Fig. 7. Theodora, 338/39-40. JE 3,1.33 g (2:1) Constans. This, combined with other facts?that the Constantinople, RIC 8.50/51 (unpublished obverse legend break) were at FLMAX AVG majority of these coins produced Trier, the Obv.: THE-ODORAE Bust with mantle and necklace, hair dressed capital of Constantine II; that the earliest issues plain elaborately Rev.: PI ETAS-ROM ANA /CONS8 atTrier; that were struck atTrier; appeared they regularly Pietas head an infant at her breast was an to facing, right, carrying that there obvious reluctance produce them at were Rome and Constantinople; and that the types centrally designed and disseminated?indicates that Constantine II was one the responsible for designing as as and producing them, well convincing his brothers to do likewise. No coins had ever before been issued in the name

Fig. 8. Theodora, 338. JE 3,1.67 g (1:1) nos. same RIC 8.28 75 RIC 8Trier 37-43 all carry the mark, in spite of the dif Rome, + in not Obv.: FLMAX THE-ODORAE AVG fering placement of the the field. I do include Kent's last-listed on Bust mintmark in this analysis. Since it appears only issues forConstan with plain mantle and necklace, hair elaborately dressed tius seems to * and Constans, it have been employed after Constantine Rev.: PIETAS-ROM ANA /R E Us death 8 Trier nos. an at (RIC 93-95). Pietas facing, head right, carrying infant her breast

DOP 62 24 IR. W. Burgess

of Theodora; also, since her death ten years earlier there had been no other coins in the name of Helena. The

appearance of these types forConstantius Is two wives as a at same as linked, parallel issue, virtually the time is themassacre, clearly significant, and each type derives Fig. ?. Theodora, 338/39. JE 3,1.98 g (2:1) additional from its The of Trier, RIC 8.79 meaning partner. depiction Obv.: FLMAX THEO-DORAE AVG Theodora asmaternal Pietas cannot but refer to the extir Bust with mantle and hair dressed sons since plain necklace, elaborately pation of her and grandsons, the type clearly Rev.: PI ETAS-ROM AN A /TRP mirrors theSALVS REI PVBLICAE and SPES REI Pietas an at facing, head right, carrying infant her breast PVBLICAE types issued by Constantine between late in 324 and mid-326 honor of the reproductive abilities is on reverse of Fausta, who depicted the holding her infant sons Constantine and Constantius (who at the were seven ten see time actually between and years old; must seen as an act Figs. 17-19).76 The type therefore be to as of expiation Theodora themother of Constantius were Is children and grandchildren, nearly all ofwhom dead when the type was first issued (see below). Their jo. Theodora, JE 3,1.52 Fig. 337/38. g (2:1) deaths were not intended to reflect on her as a Trier, RIC 8.65 poorly a Obv.: FLMAX THEO-DORAE AVG mother. The choice of PAX as type forHelena also Bust seems to the circumstances when seen in the with plain mantle and necklace, hair elaborately dressed pertinent Rev.: PI ETAS-ROM ANA / ?TRP? context can as a of what only be described "civil war" an at Pietas head carrying infant her breast facing, right, within Constantine's family. If the Theodora type refers to must to her offspring, theHelena type then refer hers as can as a well. It only be intended promise of future "pax publica," "state/imperial peace," that is,peace among the state leaders of the (Helenas three grandsons) and their on relatives (those Theodoras side of the family). to see two It is easy then why the other brothers to are objected the production of these coins: they sub a. no versive and in a real undermine the official ver Fig. Helena, 337-38. JE 3, weight (2:1). very way Constantinople, RIC 8.33 sionsof theevents of 337by highlighting the closeness Obv.: FL IVLHE-LENAE AVG of the events to the three brothers. It is no wonder that Bust with ornamental mantle and necklace, hair dressed elaborately were not on themore and Rev.: PAX PV-BLICA ? /CONS8 they produced important gold transverse silver Pax standing left, holding branch and scepter coinage.

III.3. Synthesisand Summary s The analysis above clearly demonstrates thatConstantius was to cover original impulse up everything with the damnationes memoriae, but the court was soon forced

76 For a list of references to Faustas SALVS REI PVBLICAE Fig. 12. Helena, 337-38. JE 3,1.50 g (2:1) many and SPES REI PVBLICAE reverses, see RIC 7:749 and 753? Kent Constantinople, RIC 8.33 (fol . 2 n. Obv.: FL IVLHE-LENAE AVG lowed by Di Maio and Arnold, "Per Virn [ above], 195 211) believes that the nummi forTheodora and Helena were issued from Trier before Bust with ornamental mantle and necklace, hair elaborately dressed massacre raises Rev.: PAX PV-BLICA ? /CONSE the (RIC $17,126). This obvious problems, which Di Maio and Arnold note, but like Kent fail to realize that it is the coins that transverse they Pax standing left, holding branch and scepter massacres. provide the best evidence for the date of the

DOP 62 Summer The of Blood | 25

we are to (in stages, if judge from Libanius and Julian) to an were disseminate official position that the deaths a was the result of mutiny that beyond Constantiuss a to control. Eventually decidedly unofficial version began circulate widely thatConstantius himself had prompted Fig. 13. Helena, 337~38. JE 3,1.52 g (1:1) and the soldiers' actions. As this version promoted gained Constantinople, RIC 8.33 court was time general acceptance, the forced, by the of Obv.: FL IVLHE-LENAE AVG to counter at Bust with ornamental mantle and hair dressed the reign of Gallus (351-54), these charges, necklace, elaborately ? not to Rev.: PAX PV-BLICA /CONSE least internally, by continuing deny them, but by transverse Pax standing left, holding branch and scepter admitting their truthwith the claim that Constantius had genuinelyrepented of thedeed, which he blamed for his many later failures (as indeed he may have done). This version eventually became public. Later writers, for eventswere whom these history rather than contemporary news, took the official and themany and varied unofficial in versions circulation and, for religious and political reasons, altered them or combined them in defense of either Constantine or Constantius, producing hybrids Fig. 14. Helena, 338. JE 3,1.87 g (2:1) unknown. on the ofmodern previously Attempts part Rome, RIC 8.27 (unpublished mint mark) to Obv.: FL IVLHE-LENAE AVG scholars mix these distinct traditions, thereby them Bust with ornamental mantle and hair dressed selves new versions, or to admit from necklace, elaborately creating anything * Rev.: PAX PV-BLICA /R or to Q the late narratives of Philostorgius Rufinus are, my Pax left, branch and transverse scepter to standing holding mind, misguided, because they fail consider the prom various ulgation and chronological development of the versions, official and otherwise. In two general, there have been modern approaches to the above evidence: the first discounts as hostile pro accounts massacre paganda the that attribute the directly toConstantius. After all,most sources thatmention his are involvement manifestly hostile toward him. This leaves one to fashion a reconstruction from the official versions Fig. is. Helena, 337. JE 3,1.59 g (2:1) Trier, RIC 8.42 and found in Eusebius, Libanius, private explanations Obv.: FL IVLHE-LENAE AVG and theKG: in toConstantine s Julian, Gregory, response Bust with ornamental mantle and necklace, hair elaborately dressed deathbed decision that his sons should succeed him only Rev.: PAX PV-BLICA /i. and out of fear of a of transverse possible usurpation imperial power Pax standing left, holding branch and scepter on the part of at least one of Theodoras descendants (as a result of that decision?), the army declared that they would have no emperors but Constantine's sons and set about murdering Theodoras descendants and their was supporters; Constantius somehow tricked into acqui at was to escing in this first and anyway helpless stop it in the face of the army s violence, but he was able to rescue Gallus andJulian (or have themrescued) and thenfinally to Fig. 16. Helena, 337~38. JE 3,1.49 g (2:1) suppress the rebellion with great difficulty. Trier, RIC 8.47 mint mark) The alternative view is thatConstantius, almost cer (unpublished Obv.: FL IVLHE-LENAE AVG as a result of conflicts the succession tainly concerning Bust with ornamental mantle and necklace, hair elaborately dressed with Julius Constantius and his relatives and without Rev.: PAX PV-BLICA / ?TRS consent Pax left, branch and transverse the immediate of his brothers, instigated the standing holding scepter

DOP 62 26 R.W. Burgess

massacre in to two order accomplish purposes: first, to s return to a abandon his father plan?a tetrarchie system inwhich the descendants of Constantius I and Theodora and their wielded supporters great power through official appointments, marriage connections, and behind-the 17. Fausta, 32.4-25. & scenes to Fig. 3? 3?7 g (1.5:1) advisory positions?and allow the promotion Trier, ?/C 7.459 of all three brothers to the rank of augustus; second, to Obv.: FLAVMAX-FAVSTA AVG remove any for future power, influence, or Bust with bare head, waved hair, necklace, and ornamental mantle opportunity current or Rev.: SALVS REI-PVBLICAE /STR interference from future descendants of Theo Salus two dora. Gallus and survived either because were standing facing, looking left,head veiled, infants in Julian they her arms as no so were perceived being of immediate threat and allowed to or were live, because they rescued (byMark of Arethusa?). Itwas believed at the time thatConstantine s sons sons would eventually have of their own, and so Gallus and Julian would live out their lives in exile and The was obscurity. underlying problem that Theodoras meant that therewere too fecundity simply many legiti mate to claimants the throne for any real security for Constantine s sons and their descendants?the threat of future war was in civil quite real?especially view of the Fig. 18. Fausta, 324-25. Solidus, 4.54 g (1.5:1) attacks made some Constantine's Nicomedia, RIC 7.77 by against legitimacy, Obv.: FLAVM AX-FAVSTA AVG both as a son of Constantius I and as emperor. No son of Bust with bare head, waved hair, necklace, and ornamental mantle Constantine would have been unaware of the influence Rev.: SALVS REI-PVBLICAE SMN / that Galerius had had over Diocletian in 30$ in remov Salus head two infants in standing facing, looking left, veiled, Maxentius and Constantine from the succession and her arms ing own appointing his creatures, Severus and Maximinus, in as caesars. to see their place Itwould have been easy Julius Constantius, Flavius Dalmatius, or even Ablabius in was a matter not the Galerius role. It therefore just of removing Dalmatius and Hannibalianus from office, but of purging their fathers, all other possible contenders for the all those purple, and who supported them and s Constantine tetrarchieplans. In this context the "rescue" of the two youngest males, Julian and Gallus (rescued Fig. 19. Fausta, 326-28. JE 3, 2.39 g (1:1) in the sense that were not Constantius Thessalonica, RIC 7.161 they targeted by in sense. Obv.: FLAVM AX-FAVSTA AVG the firstplace), would make Bust with bare head, waved hair, necklace, and ornamental mantle two Of the views, only the latter is consistent with Rev.: SPES REI P-VBLICAE /SMTSA all the evidence. This is best demonstrated a two through Spes standing facing, looking left, head veiled, infants in her arms point-by-point summary analysis. .The weaknesses of the official explanation. The a first problem is that fundamental part of the official must as explanation be immediately rejected false: Con stantine did not have a deathbed conversion regarding his not succession, and he could "bequeath" the empire to his sons. That sons were not his proclaimed augusti on until 9 September disproves the official version itsmost once essential point. Second, thispatently false device has it seems extreme been removed implausible in the that

DOP62 The Summer of Blood | 27

the of its own accord?without a on army leader??would suddendesire thepart of rampagingsoldiers that only have rejected Constantines almost two-year-old settle Constantine's sons should succeed him, in violation of ment rebelled Constantius commander plan, against (a Constantine's express and already implemented plans. knew and under whose command were about are they they Pragmatically and politically, Constantius's actions to so campaign against the Persians), and assassinated completelyunderstandable and explicable.The official members of the and their version is As many imperial family supporters neither. Seneca said, "Cui prodest scelus, a caesar. was on one a including That maiestas, plain and simple, is fecit" ("The who benefits from crime is the one a in scale unparalleled Roman history. The resultwould who committed it").77 have been the execution of numbers 2. sons high-profile large The precious metal coinage. The of Constan of commanders and subordinates and the of tine did not name on or cashiering Dalmatius any gold silver struck entire units in an at was the aftermath of such outrage against their home mints while Dalmatius caesar and still the And there is no a we emperors' family. yet hint of such alive.From thebeginning findevidence for hostility ever an on response. No emperor could have allowed such toward Dalmatius the part of the other caesars and to severe a to as caesar on attack upon the imperial family pass without coordinated response his accession the even reprisals, ifhe had allowed theattack and did profit part of all three. from itsoutcome: the would have been far too The memoriae. is no precedent 3. damnationes There place for even one dangerous. And if credits the possibility of such any damnatio memoriae in the official version, since the a one to content were a revolt, might expect the troops have been descendants of Theodora the innocent victims of with the heads of Dalmatius and Hannibalianus, and mutinous army. Itmust be remembered that there is no Constantius and his in was a perhaps Julius brother, Dalmatius, claim any official explanation that there actually as since were ones well, they the only with imperial offices usurpation, only that the troops feared it.And the clear ones an and powers and thus the only posing immediate implicationof all theofficial versions that survive imply threat to Constantines sons and their succession. The that this fearwas erroneous and unfounded. to massacre was not an removal of all potential dynastic threats the brothers The damnationes prove that the no even and of theirmale supporters, but women, not the accident, as does the response of Constantine's sons to wives or not as daughters of those murdered, does sound the deaths of their relatives. It is just Pohlsander says a to like the result of spontaneous mutiny of ill-disciplined with respect the death of Fausta in 326 and whether troops. Furthermore, without the emended passage from it could have been accidental: theKG (viaAurelius Victor) thereis simplyno evidence on We must answer "no"... for then the bereaved forany previous hostility thepart of thetroops toward emperor or a Dalmatius, his family, their supporters. would have ordered splendid funeral,orations, and In to monuments. Her name not have been fact, according theKG, Dalmatius had been would erased seems to from Eusebius would have lauded her in popular with the soldiers and have possessed inscriptions. virtues in sons. As his Vita and her sonswould have and talents lacking Constantines Constantini, honored a some her result, he and his many relatives, of whom had memory.78 to been elevated by Constantine positions of power and a serious threat to the brothers and their prestige, posed . 1 77 Medea 500, quoted by Olivetti, "Sulle strage" ( above), 79? solehold on if held inmili we must Constanss as We must power,especially they appeal this light consider possible role well. also is since remember the of the of members of tary and political circles,which likely, Dalmatius example purges previously imperial a in families and their supporters undertaken in the East by Licinius after his p?re had been military commander himself Oriens defeat ofMaximinus in the summer of that were intended to leave and As andAmmianus the 313 (Cyprus Tyre). Julian hint, Licinius as the newly linked families of Constantine and the sole legiti matter seems to have come to a head when Constantine s mate see imperial dynasty. For this, Barnes, Constantine and Eusebius half and became involved in . more M. "From brothers nephews arguments ( above), 64 and, generally, Humphries, Usurper to over Emperor: The Politics of Legitimation in the Age of Constantine," with Constantius the details of succession. The spark 97-99: "Constantine was not an of a heated confrontation over the succession that was JLA (2008): esp. only accomplished master but also an His [at constructing legitimacy], excellent teacher. fanned fears on Constantiuss of sons it into action in by longstanding part learned the lesson well, and put repeatedly: first a ...." dynastic threat from Julius Constantius, his brother, 337 (p. 99). a more a 78 6 and their families is plausible hypothesis than Pohlsander, "Crispus" (n. above), 103.

DOP 62 28 R.W. Burgess I

The same would have been truefor Dalmatius and 7?The KG's account has often found themost favor not rest Hannibalianus, if for the of the family, had they with modern scholars, but its author clearly favored Dal a contra as was seen was been the innocent victims of mutiny. In clear matius and, above, it composed when diction to the various official versions thatwe know of, written accounts of Constantius s involvement were first it is were in as court was to Theodoras descendants who 337 treated appearing, when the imperial beginning of a state not the which was admit Constantius's involvement in the face of these being guilty crime, army, later forced to shoulder the blame. This crime may lie accusations, and while Constantius was still alive. As behind the tendentious account of Eusebius, but it is a result, it presents a transitional account between the not even at not an view hinted anywhere else. official and the unofficial versions, unbiased we not cannot What do and know iswhat this crime of reality. was. What could Julius Constantius and Theodora s rela 8. The Helena and Theodora nummi. The appear tives a ance coins at have been charged with thatwould justify dam of these this critical moment, issued only natioi is in unless themints of the no There nothing the surviving sources, from brothers' capitals and others, itwas in fact a influ a connection usurpation, charge that indirectly indicates very close and specific between encedGregory, Eunapius (VS),Rufinus, Philostorgius, the brothers and these coins, and therefore between the andTheodoret, but ifso, itwas quicklydropped officially. brothers and themassacre. It ishard not to conclude that coins were an act at This is,however, just speculation. these of expiation, aimed directly the The later treatment of Gallus and First of thewife of their 4. Julian. memory "pia" Theodora, grandfather no sense in on reverse as of all, the exile of Gallus and Julian makes Constantius, presented the the epitome of context was not at moment the of the official explanation, which Roman maternal pietas the very when her to or treatment designed explain the survival later of the childrenand grandchildren had justbeen slaughteredand two since itwas never rise memoriae. In boys, expected that theywould suffered damnatio this light, the facts? to or The exile is coins were in positions of power notoriety. nothing that these begun and issued the greatest more than an extension rise to came to an of the hostility that gave numbers by, and abrupt end with the death II one to themassacre, and that hostility clearly emanated from of,Constantine (the only who stood gain noth not one Constantius, from the army. ing from the removal of Dalmatius, the only who in assassina to 5.The involvement of Constantius the actually stood lose territorywith the promotion of n. tionof Ablabius. This isdescribed in detail by Eunapius Constans to augustus [see 27], and therefore the one in Lives least to have been in on a his ofthe Sophists, though the accuracy of the likely involved any plot)?take to as new in mes specificsis open doubt, such Ablabiuss propensity meaning, especially view of their subversive for usurpation. In his History Eunapius (via Zosimus) sage. If the assassinations had been the spontaneous acts makes it clear thatAblabiuss death was of the of a mutinous in part gen army Constantinople (or somewhere us no eralmassacre; Jerome tells that "multi nobilium" died in its vicinity), therewould have been need for such His to a alongwith Ablabius. deathhad nothing dowith personal statement, instigated by Constantine II, who itwas of themassacre undertaken was in no a issue any mutiny, yet part far away Trier, and need for parallel for by the army. Furthermore, Constantius seized Ablabiuss Helena, who had no connection with Theodora sdescen estates as he did the of the other victims. or property dants the supposed uprising. 6. The of There is no serious rea a testimony Julian. 9. As minor and subsidiary point, it is also pos son to doubt the evidence of the closest we was a use to Julian, have sible that there precedent for the of soldiers to an His toward Constantius is remove The eyewitness. hostility imperial threats. Origo Constantini imperato insufficient to vitiate what he in ris to says, especially the light appears say that in 325Constantine ordered Licinius of the it receives from the other His to in a support evidence. be assassinated military revolt: "Sed [Constanti account of the excuses to him and his brother in Herculii Maximiani soceri sui motus ne given nus] exemplo, seems and matches iterum in rei particular completely trustworthy depositam purpuram perniciem publicae the diachronic of the found in tumultu militari in shifting justifications sumer?t, exigentibus79 sources. other Moreover, when he touches upon events

forwhich we do have other evidence as the The reads "tumultum good (such 79 manuscript militari exigentibus," which death accounts are Mommsen emended to "tumultu militaribus His of his brother), his trustworthy. exigentibus."

DOP 62 The Summer of Blood | 29

sons events. is one Thessalonica [Licinium] iussit occidi" (5.29). Now there the later presented the There only are the natural source for the actual date of the the differentways of translating this, but way promotion, Descriptio seems to is to (it me) take the ablative phrase "tumultu consulum(s.a. 337.2),and itgives 9 September337, that ... not as with "motus three and a half months after the death of Constan exigentibus" coordinate exemplo" is, source on theother sideof thepurpose clause ("influencedby tine.81 Strange though this date may be, the itself also is been in ca. the example of his father-in-law... and because the very nearly contemporary, having compiled as it is source other soldiers mutinously demanded his death," Rolfe puts 342, and the sole formany important are or to it in the Loeb translation), but as part of the infinitive Constantinian dates that appear be accurate, "influenced the so its has phrase, which immediately follows: by testimony authority.82 Maximian he a example of his father-in-law Herculius80 This date ispartially confirmed by letterof Con a on a 2 ordered [Licinius]to be killed inThessalonica in mili stantine II written 17June 337 and by law of August so same was written taryrevolt with everyonedemanding his death, thathe of the year. The letterof Constantine II not to the detriment of the state take a second almost amonth afterConstantine's but Constan might up death, as timethe imperialpower thathe had laid aside." tine II still refers to himself as caesar, not augustus, he at In view of the above, the only internally consistent certainly would have done had he been augustus the reconstruction follows the narratives of time.83 In this letterConstantine II refers to and plausible explicitly events can take new Julian and Eunapius. Only this version of thedeath ofhis fatherand is clearlyexercising his account most role as senior Athanasius to his see plausibly of all the literary, epigraphical, emperor by restoring in it not a and numismatic evidence, the damnationes, the obvious Alexandria. Although does provide date for The the it confirms the basic fact of the political motivation, and the attempted cover-up. promotion, Descriptio on inac the brothers were not Constantine official version relies implausible and factually that promoted by curate or the the and the details and is contradicted by other facts. Finally, before his death, by armies, senate, even itmust be remembered that eventually Constantius people immediately afterward. on two summer gave up the official version. Of laws surviving from the of 337, massacre one was almost issued in Constantine's name Many different scenarios for the and the certainly events it are consistent with the above after his death?CTh CIC CI dated surrounding that 13.4.2 (= 10.66.1), can must 2 indicates that the broth evidence be hypothesized, but we, unlike others, ("data") August 337?which to create none ers not 2 No resist the urge order and clarity where still had promoted themselves by August. is to exists. location is given for the law,which addressed (Val a at erius) Maximus, praetorian prefect whose posting this date is unknown (see below, section IV. 2). Another Vaticana issued the law,Fragmenta (Frag Vat.) 3s,84 by caesars names not augustus and the (their actual do but was IV. The Chronology appear), bears the date of 29 August 337, origi in nally written inConstantinople the February before IV. . TheDate Promotion ConstantinesSons ofthe of Constantine's death (= CTh 3.1.2,dated 4 February 337).85 in the events of the sum Part of the problem evaluating a in 81 Even Jerome had version of theDescriptio front of him mer of 337 is that every source but one assumes thatCon though when he calculated the of Constantius's years, sons on length reign (twenty-four stantine's succeeded their father upon his death so five months, and thirteen days, p. 234), he did from Constantine's 22 is since was normal May. This logical, that the practice death. in his sons were caesars and thus 82 Chronicle with cor the empire, already See Burgess, ofHydatius (n. 43 above), 191-97, out as and thatwas how rections in Studies (n. 25 above), 270. The Consularia Constan already marked his successors, Burgess, isMommsen's name for the consulum. tinopolitana Descriptio

but 83 Athanasius, theArians contra Arianos emendation has been followed by Rolfe, Moreau/Velkov, and K?nig, Defence against [Apologia is so common n. that it c.Ar.]) 87.4-7 and//?/. Ar. 8.2. the expression "tumultu militari" (see, e.g., 45) [ApoL must remain. "cunctis" as a 84 Fontes iuris Romani 2nd ed. Boissevain suggested supplement. J. Baviera, antejustiniani, (Florence, to 469-71? 80 Maximian had twice resumed power after having been forced 1940), n. dates of Vat. and CTh are retire: Barnes, New Empire, 13 and Kienast, Kaisertabelle (both 3 85 The conflicting Frag. 35 3.1.2 easily is a of the law as itwas sent from above), 273. explained. Frag. Vat. 35 complete copy

DOP 62 R.W.Burgess 30 I

2 on ofCTh of Nevertheless, the delay makes clear the turmoil caused August and 9 September thebasis 13.4.2 Con 2 above in section which shows by the final preparations for the Persian war, by August 337 (noted IV.i), thatValerius Maximus?assumed to be Dalmatius s stantine's illness, and then by his death. prae ever so to torian on the basis oiAE where his name Eusebius subtly refers this strange period prefect 1925.72, to after his is for an erasure in an dated to the when Constantine appeared be still ruling supplied inscription e a summer of still Since the deathwhen he says ?a e a e a a 337?was praetorian prefect.87 ... a e e a a e a erasure means thatMaximus suffered damnatio memo ' a , a a a a e riae, he must have died with Dalmatius, and since an was addressed to him on 2 he e (VC 4-67.3;"Alone of mortals... he reigned imperial rescript August, were must even after death, and the customs maintained just have died after that date. as ifhe were alive,God havinggranted this to him and As it stands, Barnes's argument for the date of the no since time massacre is and was a advance over the other began"). cogent great jum of earlier and even later scholars. As will be describedbelow (sectionIV.6 andAppen bled thinking However, Trier a more Greek ofAE shows dix 2), the coinage ofAries, Rome, and indicates another, complete copy 1925.72 time between the death of Constan that both these must date to and that significant length of inscriptions 335/36 sons name isValerius not Valerius tine and the promotion of his three during which themissing prefect's Felix, to names Maximus was coins continued be struck in the of the three Maximus.88 Thus, although Valerius pre sons as caesars on 2 we not alone. fect August 337, do know whose prefect he name All the above evidence serves to confirm the accuracy was. Itmay be that the erasure ofValerius Felix's was to massacre in the summer of but of the date in theDescriptio consulum. related the 337, reasons alone have his removal strictly local may governed must from the inscription.89 The date of post-2 August IV.2. therefore be abandoned. Modern Scholarship massacre at some Most scholars accept that the took place date before9 337without statingwhy, though September Sources some a even a as IV.3. The suggest date shortly afterward, and date Literary as most late 338 is still espoused by some.86 The explicit TheKG noted thatthe death of Dalmatius had happened accounts massacre was and careful of all modern of the "haud multo post" ("a little after") the death of Constan it to tine or "confestim" that of T. D. Barnes, who dated the period between (Eutropius 10.9.1) ("immediately") a afterward (Victor 41.22). This indicates date soon after 22 337.Victor notes that Constantine II, who was to corrector of Picenum and includes May Aquileia by the praetorian prefect the at as as at killed in 340,fell three afterthe death of Dal the data date Aquileia well the accepta date Alba (18 Septem early years an matius which is thus in also ber). CTh 3.1.2 is excerpt of sections 3-5 of this law and preserves the (41.22), placed 337. Jerome, issuance of the law inConstan chronological details from the original seven in is of a tinople months earlier. The difference the dates the result . in to Vat. 87 See Barnes, Constantine and Eusebius ( above), 261-62; idem, delay the proclamation of the law the praetorian prefects. Frag. on New (n. 3 above), 8 and n. 34, 134-35; and idem, Athanasius has also mistakenly taken the year of the previous law (313). Empire and Constantius 35 34-35, 218, 219, 224, with New . 2 and Chantraine (n. above), Empire, 86 Olivetti ("Sulle strage" [ above], 70) (Nach 85-87. See also Klein, . 115-16 and PLRE 1:591, . 2 . extensive summaries "K?mpfe" ( above), [ above], 5-6 io) give useful and folgeordnung s.v. "Maximus . 49." of the various dates. Charles Pietri ("La politique" [ above], 122-23) 88 D. Feissel, "Une d?dicace en l'honneur de Constantin II C?sar et tries to use Jerome to date the assassinations to 338, a common mis in at n. . dates the les du pr?toire de 336," G. and D. Feissel, take (see below 90). Klein ("K?mpfe" [ above], 115?18) pr?fets Dagron "Inscriptions massacre to to in?dites duMus?e d'Antioche," TM 9 (1985): 421-34; Di Maio and Arnold, 338 and the meeting of the brothers June 338. Browning . 2 it Pot aPer Vim" ( above), 195-98; T. D. Barnes, "Praetorian Prefects, 337-361," (Emperor [n. above], 34-35) dates shortly after 9 September. 249-51 Eusebiusto ter(Roman Empire [n.2 above],688 n. 89) stillfollows Seeck inusing ZPapEpig<)\ (1992): (=From Augustine[Aldershot, to massacre to or 1994], paper XIII); P. Porena,Z,? della del Julian's age date the the very end of 337 early 338. See Hampshire, origini prefettura tardoantica, di storia antica 20 (Rome, 2003), 466-87, and O. Seeck, Geschichte des Untergangs der antiken Welt, vol. 4, Anhang pretorio Saggi Kaiser chart between pp. 582 and 583; and A. Co?kun, "Die (Berlin, 1911), 391-92, andRegesten der und P?pste f?r dieJahre 311 pull-out Praefecti im which themassacre between und die Praetorianer praefecturen bis 476n. Chr. (Stuttgart, 1919), 186, places praesent(al)es Regionalisierungder in to vierten Jahrhundert," Millennium (2004): 286-91, 302, 327. January and March 338 and the meeting May July. However, for 2 22. Julian's age, see Bower sock, Julian (n. above), 89 See Feissel, "D?dicace," 424-25.

DOP 62 The Summer of Blood | 31

on sources relying theKG, places the assassination ofDalmatius for these events, we cannot tellwhether he had after that of Ablabius and between two entries a reason relating for placing themurder of Ablabius beforethat to the siegeof Nisibis (234e); in spiteof the claims of of Dalmatius or not. Furthermore, Jerome notes that some this not since scholars, indicates 337 and 338, Shapur manyothers died alongwith Ablabius ("multinobilium"); we were began the siegeof Nisibis almost immediatelyafter the if to accept that Ablabius was killed later,we death ofConstantine to must not discount (June August 337).90 only Jeromes chronology but also The panegyricof Julian, and perhaps that of Libanius accept that therewere two separate massacres, which is in section indicate that Constantiuss unmentioned source (discussed III), by any other and implicitly denied problems with the army occurred immediately after byEunapius (viaZosimus). his fathers death and before the brothers' in does meeting Eunapius claim that Constantius acted against to out Pann?nia work the division of the empire. Julian Ablabius only afterhe had been allottedhis portionof same in to indicates the his Oration Heracleius theCynic the empire, and since Constantius sent the assassins when he that themassacre took to (Or. 7.228A-B) says place from Constantinople, the date would appear be after as a over result of quarrels the division of the empire fol Constantiuss return from Pann?nia, which was late Constantine's lowing death. September337 (see sectionV, below).But itseems clear states massacre Gregory that the happened that Eunapius has fallen into the natural, but erroneous, a ... a a a a a e a ancient sources assumption made by all but theDescriptio a and a ?a e a ("whenConstantius that Constantines sons succeeded their father immedi had just received power from his father" and "at the atelyupon his death.Thus Eunapius is infact dating the of his Or. and Libanius dismissal and assassination ofAblabius to the time beginning reign"; 4.21 22). says during itwas a e e e e e which Constantius in ("When resided Constantinople between Constantine had fallen ill and Or. Ammi his arrival from died"; 18.10). Antioch and his departure for Pann?nia anus likewise "inter exordia" at see section says imperandi ("right (onwhich, IV.7, below). This is consistent with the sources. beginning of [Constantiuss] reign," 21.16.8) and "post the other literary fratrisConstantini excessum" ("after the death of [Julius Constantiuss] brother, Constantine"; 25.3.23), all clear references to the summer of The 337. IV.4. Legal Epigraphic,and Ablabiuss assassination has often been separated PapyrologicalEvidence fromthat of theothers on thebasis of thenarrative of his The laws are of no for the massacre since, deathgiven by Eunapius (VS 6.3.9),which seemsto imply help dating Dalmatiuss name would have in the passage of some time between Constantiuss dismissal although appeared the to law issued until his death, ofAblabius immediatelyafter the death of his fatherand superscription every with Constantine, Constantine II, Constantius, Ablabiuss eventual assassination.91 However, Zosimus's along and Constans, all these names were removed the detailed account of themassacre (2.40) also derives from by later of the Theodosian Code of earlier which he to compilers except that Eunapius (his History, though, refers Constantine, the sole After 9 the here:VS 6.3.8-9)and thereAblabius isexplicitly included augustus.92 September, laws bear the name of Constantius, within the main massacre. Jerome (Chron. can. 234e) chiefly retroactively senior augustus for all laws between 337 and 340 after places his entry concerning Ablabiuss death right after the damnatio memoriae of Constantine II in 340. This the accession of Constantine II, Constantius, and Con shows that even if the laws had the names of the stans, and before themurder ofDalmatius, which he also preserved caesars, Dalmatiuss name would have been removed describes(from the KG). SinceJerome had twodifferent by

90 See Studies 25 233-38. et . ii Burgess, (n. above), 92 See Bagnali al., Consuls ( above), 73 and S. Corcoran, "Hid . n. Constantine den from The of in The Theodosian 91 E.g., Klein "K?mpfe" ( above), 114 17; Barnes, History: Legislation Licinius," Code, I. and Eusebius, 262; idem,Ammianus Marcellinus and theRepresentation ed. J.Harries and Wood (Ithaca, NY, 1993), 107. For the editorial pro n. "Successors" cess involved in the laws of the see F. ofHistorical Reality (Ithaca, NY, 1998), 133 17;Hunt, (n. producing Code, J. Matthews, Lay 2 on see Down theLaw: A the Theodosian Code above), 4 (citingJerome; whom, below); and Frakes, "Dynasty" ing Study of (New Haven, 2000), (n. 2 above), 99. 121-67, 200-253.

DOP 62 R.W. Burgess 32 I

memoriae s the compilers because of the damnatio against 13August 337Oxyrhynchus knew nothing of Dalmatius him.93 death or damnatio memoriae. However, the retention of are Constantine's here does not mean Nor inscriptions ofmuch assistance, apart from regnal year necessarily isno that were also unaware of Constantine's the negative fact that there known inscription that Oxyrhynchites to sons to as a sort refers the three of Constantine without their death. His regnal years continued be used of no father but with Dalmatius, and inscription with Dal "era" until 354/55 (POxy LX 4092.10).97 name is matiuss independently dated.94 The earliestfully dated papyrus afterPOxy XLV are more 2 The papyri, however, somewhat forthcom 3266 (13August 337) isPOxy XLVIII3386, of Phar even so ing than the laws and inscriptions, though the mouthi (28March) 338, dated year "32-22-14-5," thus are all that of Dalmatius. No results disappointing. Oxyrhynchus and many other retaining regnal years except cities as ameans of The of Egypt used regnal years dating. papyruslater than 13August 337includes the regnal year normal system treated each Egyptian civil year, begin of Dalmatius. New Years as a new now ning with Day, Thoth (29August), This raisesthe question of thespeed of thedis when came semination and of news from Constanti regnal year, regardless of the emperor actually employment to a In names new con power, with the fraction of year between the date of nople. the fourth century the of the accession on and the end of that calendar year always being suls normally appear Egyptian documents within as even itwas a a or two even seas counted year one, if only few months month of proclamation, though the or were to The in weeks long. The regnal years ofmultiple augusti and closed shipping.98 change of regnal years caesars were to form such as formulae accessions and deaths added together strings, "year dating following imperial is is a to 31-21-13-4-2"?which the Egyptian form of the year little harder gauge because of the relative paucity to 28 of the evidence and the fact that news of an from29 August 336 August 337?listingthe regnal emperor's or was at years of the emperors in order of seniority?in this case, of death accession acted upon different times in Constantine, Constantine II, Constantius, Constans, and differentplaces (evenwithin citiesand nomes),with the some were new or Dalmatius, respectively. Unfortunately, thismethod of result thatwhile scribes using dates had to out inmost cities not But dating began fall of fashion between discarded old ones, others had done so.99 where as is to 309 and 316 consular dates and indictions replaced the there sufficientevidence make reasonable judgments, some reason centuries aver cumbersome regnal year formulae, but for between the late third and late fourth the in use in time is news they continued Oxyrhynchus.95 age lag about four months for of imperial We have no documents from 335-36, the first year accessions or deaths to be disseminated and appear in ofDalmatiuss rule, but from 336-37 (= 31-21-13-4-2) And with the accession of Constantine's we havePOxy 192.4 (15Oct. 336),POxy XLVIII 3385.4 97 then, Julian, regnal year was = and the years of Constantius and Julian became fos Nov. SB VI 9191.9 9270 but with a dropped regnal (9 336), (undated, an in silized, creating "Oxyrhynchite era" that began 355-56 with year reference ahead to 28 the last of the and August, day year), "32-31" and continued at least to 668/69 with year "345- 4?" F?r this, see SB XIV PCollYoutie II 82.11 similar had earlier with the 12057.4 (undated).96 =POxy ibid., 56-62. Something already happened XLV written on 20Mesore retention of Galerius's years (alone, without his 3266, (13August) 337,makes regnal colleagues' years). He died in his nineteenth year, but there are instances of a reference to the started on 29 regnal preserved upcoming year (which to . years twenty twenty-six "Datation" [ 95 above], 235-36, as no (Chastagnol, "32-22-14-5-3," other with August 337) though papy Bagnali and Worp, Chronological Systems, 8-9, 246-49). rus of refers to this as such. Thus on et . ii 337/38 actually year 98 Bagnali al., Consuls ( above), 29-30. 99 For instance, Galerius died in of still A to early May 311, yet appears 93 similar problem, but with regard Licinius, who also issued twice in in nome full dating formulae the Arsinoite inAugust 311 and laws and suffered later damnatio, is discussed by Corcoran, "Hidden in on even in at Oxyrhynchus 3 Sept. 311, though Oxyrhynchus least from History," 97-119. hisdeath was knownby July 311 and with theremoval of 94 CIL Dalmatius iscalled AVG andNOBILISSIMO acknowledged 10:8015(where his name from consular it in dates (though still appears other consular is in error, as CIL 10:8021 shows. CAES) obviously in not 2 we dates there late August 311). It is until Dec. 311 that have evi see 95 In A. "La datation par ann?es dence for his removal general, Chastagnol, r?gnales Egyp from the regnal year formulae. See Bagnali and tiennes ? constantinienne," 'mAi?n: Le chez lesRomains, l'?poque temps Worp, Chronological Systems, ijj and 246-47. Even in the years that ed. R. Chevallier, Caesarodunum iobis 221-38, and R. S. concern us are (Paris, 1976), here there strange delays: POxy XXXI 2571 is dated to and K. A. 2nd Bagnali Worp, Chronological Systems ofByzantine Egypt, 27 July 338 by consuls (line 22), yet still refers to a a ed. (Leiden, 2004), 43-45. e a a e a a ( ) a ("all the unconquered and 96 Bagnali Worp, Chronological Systems, 251. masters, the augustus and caesars," 15-17).

DOP 62 The Summer of Blood | 33

use on with the as the Constantiuss in Egyptian papyri, delay increasing campaign thus belongs the period June as to fourth century progressed and regnal year formulae, August of 337.104 over us rather than consular dates, indictions, and eras, became Constantiuss victory the Sarmatians presents less used.100 an commonly with important piece of chronological evidence. The As a resultwe can see that XLV which POxy 3266, army thatConstantius would have used for this campaign in use on was to shows Dalmatius's regnal year still 13August theDanubian army,which ought have been under us at 337, tells nothing about the date ofDalmatius's death. the command ofDalmatius, who had been based Nais sus eastern On the other hand, the fact thatDalmatius's regnal year and had controlled the central and Balkan ismissing fromPOxy XLVIII 3386of 28March 338 dioceses (theMoesias and Thrace).105 IfConstantius was confirms that his death and damnatio cannot Dalmatius s and a strongly commanding army taking victory title in in was have occurred any later than the end of 337. his territory June-August of 337,Dalmatius already dead and Constantius had already claimed his territory. IV.5.A VictoryTitle at some Constantius took the title of Sarmaticus point W.6. TheNumismatic Evidence between 337and 340. The onlyopportunity he had to was summer There massacre can confront the Sarmatians the of 337.101 The date of the be pinned down very exactly was no time a once an for such campaign his brothers had through analysis of the immediately contemporary in was names arrived, and while Pann?nia he discovered that he bronze coinage, struck in the of Constantine and on to to caesars as as in names needed the frontier respond the Persian siege of his well the ofHelena and Theo so are a Nisibis and theArmenian revolt and would have been dora. The numismatic arguments of rather technical anxious to as nature so are conclude his necessary business and depart and only the general conclusions presented as as of the diocese are in in 2 quickly possible. Besides, 9 September here.The details presented fullbelow Appendix to was to ofMoesia reverted officially Constans, who then for those who wish follow them up. Most footnotes for to in responsible undertaking any necessary campaign relating thematerial includedboth here and the as in next ing himself, he fact did for the three years.102 appendixwill be foundonly in theappendix. no on There is record of Sarmatian activity the frontier Since late 33$ the mints in both East and West? between 334 and 337, though Constantine did campaign Lyons, Trier, Aries, Rome, Aquileia, Siscia, Thessalonica, was inDacia in 336, and Dalmatius's position inNaissus For this in in the third and see intended as a deterrent to further hostile activ 103 problem general fourth centuries, clearly M. "Constantine and theNorthern in seems Kulikowski, Barbarians," Lenski, It thatthe Sarmatians would . 2 ity. highlyunlikely Cambridge Companion ( above), 352-65. have started trouble while Dalmatius was stirring up 104 The calendar of Philocalus (written in 354) notes the celebration or or were a a over on there laterwhile any all the brothers present of single day of circuses for victory the Sarmatians 27 July on seems most CIL :268 and Inscr. Ital. 13.2:251with This notice almost the frontier. It plausible then that they (July, p. 488). to one of two victories: Constantine's in were in to news certainly refers 323 (forwhich, acting response of Constantine's death see Kulikowski, "Constantine" [n. 103 above], 359) or Constantius's in and to Dalmatius's that of departure (and important 337.1 believe that in 354 Constantius would have maintained a celebra the funeral in the own there is army commanders?) for capital.103 tion for his Sarmatian victory, rather than his father's, but sur ses no certainty. H. Stern {Le Calendrier de 354: Etude son texte et illustrations [Paris, 1953], 81, and cf. p. 82) believes that the form of the 100 The evidence is R. S. and K. A. com presented by Bagnali Worp, Regnal notice suggests that it is pre-Constantinian. But any tetrarchie Formulas in BASP 2 MT, 9-44 can in Byzantine Egypt, Suppl. (Missoula, 1979), memoration (it hardly be earlier) would have been recognition of and and 56-57, 224-51. was so on coin Bagnali Worp, Chronological Systems, the great victory of 294 that advertised heavily the silver in the 101 T. D. Barnes, "Two Victory Titles of Constantius," ZPapEpig 52 age of the period (which victory was, fact, only victory specifically to on C. H. V. The Roman (1983): 229-35 (= From Eusebius Augustine, paper XIV) and Barnes, named tetrarchie coinage: Sutherland, Impe From s to theDeath Athanasius and Constantius (n. 35 above), 219. rial Coinage, vol. 6, Diocletian Reform (a.D. 2?4) in it is Maximinus (a.D. 313) [London, 1967], 705: from the mints of Trier, 102 He, too, took the title Sarmaticus 338 (though just pos of a in sum Ticinum, Rome, Siscia, Thessalonica, Heraclea, Nicomedia, sible that he and Constantius conducted joint campaign the Cyzicus, in autumn mer Athanasius Antioch, and Alexandria) and that victory took the of 337). See Barnes, New Empire (n. 3 above), 199; idem, place 101 New 53 n. 32). and Constantius, 224; idem, "Victory Titles" (n. above), 230-31; and (Barnes, Empire, RIC 8:340. 105 Barnes, New Empire, 199.

DOP 62 34 R.W. Burgess

Heraclea, Constantinople, Nicomedia, Cyzicus, Antioch, one reverse on and Alexandria?had issued only type the small bronze nummi forConstantine and his caesars: two a or soldiers bearing single standard vexillum between them,with thelegend GLORIA EXERCITVS (=GE (i); Fig. 20. Constantine I, 337. JE 3,1.18 g (2:1) This reverse continued until A Siscia, RIC 7.161 Figs. 20-25).106 type 341. Obv.: CONSTANTI-NVS MAX AVG number of issues of theGE (1) type forConstantine and rosette Laureate and diademed bust right, draped and cuirassed his four caesars from themints inAries, Rome, and Trier Rev.: GLOR-IA EXERC-ITVS /ASIS* in addition to theHelena and Theodora series fromTrier Two soldiers, helmeted, at one another, inner hands standing looking described above the evidence on on in outer (III.2) provide necessary shield resting ground, reversed spear hands; between for themassacre. them, one standard dating we can That date very precisely the change of the obverse types of the nummi fromTrier, Aries, and Rome to at isdue the very frequent changes ofmintmark these us mints. No other mint provides with this type of evi into dence. Throughout 336 and 337 all fivemembers of were on the imperial college represented the obverses of reverse at the standard GE (1) type Trier and Aries, right 21. Constantine II, 337. JE 3,1.97 to Fig. g (2:1) down the very beginning of the final mintmark of RIC Siscia, 7.161 each series, the same mark that introduces the Obv.: CONSTANTINVS IVNNOB C * following names as series in the of the three brothers augusti. At the Laureate and cuirassed bust right of the to the final coins minted Rev.: GLOR-I A EXERC-ITVS / BSIS* point change mintmark, As no. 20 forConstantine and forDalmatius almost immediately at same disappear, and about the time.After the cessation of these two obverses, the same mintmark continues to on a issues GE reverse appear large number of with the (1) names in the of only Constantine II, Constantius, and Constans as caesar. At Rome we have a similar situation. Into 337 all five are members of the imperial college represented by the to standard GE (1) reverse, but with the change the last Fig. 22. Constantius II, 336-37. JE 3,1.7 g (2:1) mintmark of the series (the same mintmark that is used Nicomedia, RIC 7.201 issues as Obv.: FL IVL CONSTANTIVS NOB C for the first of the three brothers augusti), Dal matius from the obverses of theGE At Laureate bust right, draped and cuirassed disappears (1) type. Rev.: GLOR-I A EXERC-ITVS / SMNr about the same time, theGE (1) reverse forConstantine is As no. 20 a new to replaced with short-lived reverse, unique Rome and veryrare: VIRTVS AVGVSTI (Fig. 26). The GE reverse sons as caesar (1) for his three is then replaced by SECVRITAS REI PVB(licae),which was maintained for A them throughout the interregnum (Fig. 5). Dalmatius obverse never appears with a SECVRITAS reverse. at Thus all threemints all fivemembers of the impe are on nummi. rial college originally represented the Then Fig. 23. Constans, 337. JE 3,1.54 g (2:1) at Constantine and Dalmatius disappear virtually the Siscia, RIC 7.264 Obv.: FL CONSTANTIS BEA C Laureate bust and cuirassed In I common right, draped 106 the following discussion shall ignore the Vrbs Roma are to Rev.: GLOR-IA EXERC-ITVS /ASIS* and Constantinopolis types, which irrelevant my purposes here not name or As no. 20 since they do depict any member of the imperial family.

DOP 62 The Summer of Blood | 35

same time. coins was to Likewise, produced forConstantine II, production stepped up produce theHelena and Constantius, and Constans as caesar continue after the Theodora issues once the Constantine and Dalmatius But isno cessation of those produced for Constantine and Dal types had been abandoned. there question that in at Trier. the balance of shifted to the latter two matius, large numbers Aries and production types us to Hoards allow establish very exactly the chrono (see III.2, above).108 logicalstages of thebronze issuesfrom Trier (seeAppen Two conclusions must therefore be drawn from the to was dix 2), and their evidence leads four important conclu above analysis: first, thatDalmatius murdered very as reverse was soon in sions. First, noted above, the GE (1) type afterConstantines death, and probably earlier struckin thename of all fivemembers of the imperial June rather than later; second, that atTrier the Theodora and then in the name ofConstantine and and Helena followed the cessation of coins in the college minting types was at same name almost Dalmatius stopped virtually the time, just of Constantine and Dalmatius immedi to new mintmark. at same timeas after the change the Second, minting ately, the theVIRTVS AVGVSTI and to at of theHelena and Theodora types began while Constan SECVRITAS REI PVB typesbegan be produced tine II, Constantius, and Constans were still caesar. Third, Rome. were in these firstHelena and Theodora types produced that must very large numbers, which indicates striking been continuous the TheEvidence Itineraries have begun early and throughout IV.7. ofthe And fourth, these two became the interregnum. types At the time thatConstantine firstfell seriouslyill at dominant issues of the mint at Trier the inter during PythiaTherma (BAtlas52 E3, aroundthe bay and south 60 regnum, accounting for between 30 and percent of west was in ofNicomedia), Constantius the East, prob the total in the hoards studied. at a ably Antioch, awaiting the arrival of his father and On the basis of section III.2, above, and Appendix 2, in to a contingent of the army order launch campaign to As below, it is reasonable argue the following. will be II was in against the Persians.109 Constantine Trier, seen news below (IV.7 and Appendix 3), of Constan was in or Rome Constans probably Milan, perhaps (we tinesdeath arrived inTrier, quite quickly certainlyby do not know), andDalmatius was inNaissus (CIC CI theend of thesecond week ofJune. in thename two Minting 5.17.7,with Barnes, New Empire, 87). Of the eldest The numbers of Constantine stopped immediately. small was to sons, Constantius obviously the closest his father. of coins struck in the name of Dalmatius with the con Eusebius states that Constantius and his brothers were in name temporary mintmark shows that minting his not to summoned until Constantines body returned cannot cessation Con have continued long after the of is Constantine Constantinople. That unbelievable. and issues.The same can be seen atAries and stantine's pattern Constantius would have been in constant communica Rome, which would have received their orders striking tion with each other in the lead up to the Persian war, somewhat later than Trier. The the one longer period and Constantius would have learned immediately of his posits between the death of Constantine and the death No one fathers illness, ifonly from their couriers. would more to account ofDalmatius, the difficult it is,first, for to news an event have dared keep back of such important near the of the cessation of theminting of simultaneity until the body had been embalmed and then returned to nummi in their names, and, second, accommodate the to As first states that the capital.110 noted above, Julian emission ofHelena and Theodora types from Trier to large Constantius hastened his dying fathers sidewhile the before the appearance of the issues of the three brothers

as the end of Trier was issues both but shows no augusti (toward September 337). from 347, involving officin?? only reverses, of the bronze links coins. routinely producing large numbers regular among 148 is that of all other was issues at the time,107 so there isno reason to assume that 108 It possible, but less likely, minting types at a in the of the stopped later date the interregnum and that majority mint output was then dedicated to the Theodora and Helena types. 107 See R. A. G. Carson and A. M. Burnett et al., Recent Coin Hoards For to see n. For Constantius in British Museum Occasional 109 the this, 25. Antioch, from Roman Britain, Paper 5 (London, 1979), background see with the mint Zonaras 14.4.28 and Barnes, New Empire, 85. 91. Of 79 coins analyzed from the Chorleywood hoard was not a reverse die link. Studies of earlier 110 See M. Di and on the mark #TRP#, there single Maio, Jr., "Zonaras, Julian, Philostorgius on reverses the Constantine GOTR 26 121-22, and issues from both officin?? and both and obverses produced Death of Emperor I," (1981): a n. and n. one obverse link and one reverse link out of 373 coins, while study of 31, and Burgess, Studies (n. 25 above), 225-26 133.

DOP 62 36 R.W. Burgess

latter still lived, implying that he had been summoned (Or. 1.16D); laterhe states thatConstantine actually did summon reason him (Or. 2.94B). This claim isperfectly able. Itwas vital that at least one of the two eldest sons so receive promotion before their fathers death, and it Fig. 24. Dalmatius, 335-37. JE 3,1.4 g (2:1) seems almost certain that he summoned Constantius, Antioch, RIC 7.111 as two almost as Obv.: FL DELMA-TIVS NOB C physically the closest of his eldest sons, Laureate and cuirassed bust right soon as he had fallen ill. Rev.:G LOR-1A EXERC-ITVS /SMANI a The shortest trip for messenger from Nicomedia no. 20 As to route Antioch is via the standard through Nicaea, Ancyra, Parnassus, Archelais (Garsaura), Tarsus, and route 1 There Issus (see Appendix 4 and the map).111 were approximately 740 Roman miles and thirty-four mansiones between these two cities.112 In view of the

to numerous 111 Any such messenger, and those be mentioned times see below, would have depended upon the cursuspublicus. On which, in The A. Kolb, "Transport and Communication the Roman State: Cursus JE in in Roman ed. C. Adams Fig. 2$. Dalmatius, 335-36. 3,1.75 g (2:1) Publicus? Travel and Geography the Empire, Siscia,?/C 7.256 and R. Laurence (London, 2001), 95-105. Obv.: FL DELMATIVS NOB C 112 All the data for routes, distances, and mansiones in the follow Laureate bust and cuirassed are right, draped ing section derived from the Tabula Peutingeriana, the Imperato Rev.:GLOR-IA EXERC-ITVS BSIS itineraria et and the / risAntonini Augusti prouinciarum maritimum, As no. 20 K. Itineraria Romana: R?mische Itinerarium Burdigalense: Miller, an Reisewege derHand der Tabula eutingeriana (Stuttgart, 1916 [Rome, P. 1964]), O. Cuntz, Itineraria Romana (Leipzig, 1929), and Geyer and in Itineraria et O. Cuntz, Itinerarium Burdigalense, alia geographica, see . CCSL175 (Turnhout, 1965). In addition, in particular Broderson, Travel and "The Presentation of Geographical Knowledge for Transport in the Roman World: Itineraria non tantum adnotata sedetiam piet?" . in and Salway, "Travel,Itineraria and Tabellar?a" both Adams and m There is no doubt Laurence, Travel and Geography (n. above), 7-66. as mea that these lists they exist today contain inaccuracies, and their even accurate not accurate measurements surements, if textually, may be Roman such of the actual distances (although methods formeasuring Fig. 26. Constantine II, 337. JE 3,1.5 g (2:1) were are distances surprisingly sophisticated). However, these exactly Rome, RIC 8.4 (unpublished bust type) the sorts of documents that traveler in the Roman world had and Obv.: VIC CONSTA-NTINVS AVG every itwas to such documents that travelers would have Laureate and rosette diademed bust and cuirassed according planned right, draped their Thus itwould not matter whether the distance between Rev.:VIRTVS-AVGVSTI /R?> journeys. two was or is a traveler knew mansiones fifteen eighteen miles, the fact that bare-headed and in dress head Emperor, military standing facing, a a he could cover such distance in day with little effort.Two mansiones turned reversed in hand and left right, holding spear right resting a s It be twenty-fivemiles apart required solid day traveling. should also hand on shield noted that for the most part travelers had to travel in units of a mansio. a was 100 were If journey miles and there four evenly spaced mansiones, one cannot a and conclude simply postulate speed of fortymiles per day two to 100 at since itonly took and half days travel the miles that speed, not at there would have been any mansiones the forty- and eighty-mile was marks and thus therewas nowhere to stop unless the area well popu towns or Mutationes lated and the road had many villages along theway. were were for those available along theway between mansiones, but these cursus to in using the publicus mainly change horses. For this reason, the to terms discussion below I try deal in of mansiones, rather than simple are itineraries averages. There also discrepancies among the three regard in certain in some cases ing the actual number of mansiones places and routes they also offer different (see the comparison table for the Egyptian a routes of the three itineraries and papyrus itinerary from Theophanes'

DOP 62 The Summer of Blood I 37

an This map is based upon the relevant maps in Atlas; Konrad Miller, Itineraria Romana: R?mische Reisewege der Hand der Tabula Otto Itineraria 1 as well as the of the Ancient World Peutingeriana (1916; reprinted Rome, 1966); and Cuntz, Romana, voi. (Leipzig, 1929), maps Drawn Rasmussen Mapping Center (http://www.unc.edu/awmc/mapsforstudents.html). by Karen (archeographics.com).

in a return over 800 Roman miles evidence presented Appendix 3 for the speed of travel had longer journey, just on even terrain since to on to foot and horseback, with themountainous and thirty-eightmansiones, he had travel route an from of the between Ancyra and Tarsus, emergency Constantinople, though the last leg of the journey news serious have been traversed more messenger bearing of the emperors illness Nicomedia could quickly by to route1 and perhaps imminent death Constantius, and thus ship (seeAppendix 4 and themap). It isunlikely at to it as fast as the mes traveling top speed, should have been able make that he could have traveled original to or not on Antioch within six days, covering sixmansiones senger, since he would have been traveling his man two or a too conserva between 125 and 145miles per day,with only four own, but threemansiones day is on an in siones and sixty-fourmiles the last day. Constantius tive estimate, given the various examples noted was Appendix 3 and how important it forConstantius in dossier [seeAppendix 3] C. Adams, "'There and Back Again: Getting to at as soon as arrive his fathers side possible.113 If he Around in Roman inAdams and Laurence, Travel and Egypt," Geog traveled four mansiones on the first to m In some were per day (five day raphy [n. above], 161). addition, roads wide, straight, or in various and well paved, others could be narrow, winding, unpaved, states a met who was of disrepair, which would slow travelers down. As result, all the 113 No doubt along theway he themessenger traveling I use mansio times are to tell him of Constantine's death and the for the of figures for numbers, distances, and by necessity plans preparation inexact to a certain for the funeral. degree. the body

DOP 62 R.W. Burgess 38 I

on in Mopsuestia for ninety-five miles and five the second other days with quicker travel. There are, fact, about to mansiones so lastday, putting him inNicomedia, for 107miles) he 154 along theway, Athanasius did indeed an one were could average between eighty-two and ninety-five miles travel average of mansio per day. Since there on mansiones per day, with sixty-five the last day. Unencumbered about fifty between Trier and Viminacium, and , cover no one (e Libanius, Or. 59.74), he could this distance probably forhim tovisit along the way, itthus took in nine a exactly days. Athanasius maximum offifty days to travelfrom Trier Ifwe assume Constantius was when Con to a summoned Viminacium, total of about 1,161Roman miles (by the stantine first fell he would have arrived inConstan at an ill,114 longest route), average speed of twenty-three miles at or in at on 6 tinople thevery end ofMay thevery beginning of per day, arriving Viminacium the latest August, June. Even ifwe allow that he was not summoned until dependingon when he leftTrier (Ihave assumed18 June). Constantine had infact died, he stillcould easilyhave Ifhe had no one tovisit (and itseems most likelythat he been in 6 The Constantinople by June. funeral would did not), he probably traveled somewhat faster than his have followed since then Constantine s was to two immediately, by average. Although Theophanes able travel to on body would have been embalmed, returned Constan mansionesper day abouthalf thedays of his journeys and lain in state for seven VC themansiones s route tinople, days (see Eusebius, (seeAppendix 3), along Athanasius and were too to as 4.66-67 70). spaced far apart for this have been practical Athanasius states that he met Constantius atVimi often. Constantius must therefore have been inVimina nacium cium no or (Defensebefore Constantius [Apologiaad Con later than the firstweek ofAugust 337 rather, stantium(Apol. ad Const.)] 5.2).The only timethat both more probably,the lastweek ofJuly. Constantius and Athanasius could have been inVimina met at Although Constantius Athanasius Vimi cium at the same timewas Constantius was was eitherwhen nacium, it hardly the type of city that would have the Sarmatians on theDanube or when he was suited a was battling meeting of the imperial brothers. Sirmium his brothers in Pann?nia and Athanasius was a more a awaiting fitting location, being centrally and strategi fromTrier toAlexandria.115 The date therefore located center returning cally military (especially for campaigns must be the summer of the an 337.116 against Sarmatians), with imperial residence,mint We know that Athanasius was inTrier on and various times 17 June (atmany during the fourth century, when he received the letter from Constantine II read in not in arms starting 324, though 337), and factory. It to see mittinghim his (seeAppendix 3).He arrivedin had been the imperialcapital of Diocletian (285-296) Alexandria on same 23November of the year, 159 days and Licinius (308-316), and Constantine had spentmuch later.117 to According the itineraries, the total length timethere himself while on campaignbetween 317and of Athanasiuss from Trier to Alexandria was to to journey 329, especially June August 317,October 318 April Roman miles route to to to approximately 3,352 (seeAppendix 4 319,May August 320, June September 321,May and the A of this distance over to 4 map). journey averaged July 322, and March April 329.118Most important, it a of 159days gives speed twenty-one miles per day,which is in Pann?nia, where Julian says that the brothers met is what one would for someone exactly expect traveling (Or. 1.19A and 20C). on normally foot (seeAppendix 3),assuming stopsof Barnes believes that Constans may have been in more than a in some at and on see day places (as Viminacium) Aquileia 29 August (the date o? Frag. Vat. 35; above,Vi)I with his praetorianprefect, just before head to Pann?nia tomeet with his brothers and 114 statement was ing accept Malalas's that he ill for six days (Chron. 13.14; Thurn, seems to on but this 249) reliable but cannot be verified. promotion augustus 9 September,119 cannot be the case. is much closer to 115 See Barnes, Athanasius and Constantius (n. 35 above), 34-35, 41. Aquileia certainly Viminacium is over the border from Pann?nia inMoesia, but the Sirmium than either Trier or just Constantinople, between will be below. discrepancy explained 399 and 416 miles, and sixteen or seventeen mansiones. 116 Barnes (ibid., 219) suggests July 337, A. Martin (Athanase et au dAlexandrie V?glise d'Egypte IVe si?cle (328-373) [Rome, 1996], See also Di Maio and "Per Virn 2 . 394-95) September. Arnold, (n. above), 118 See Barnes, Constantine and Eusebius ( above), 8,9,32,68,72, 198-207. and idem,New Empire (n. 3 above), 49 and n. 16,51,52-53, 69, 73-78, 80. 117 E estai Index 10 with Barnes, Athanasius and Constantius,

34-36. 119 Barnes, New Empire, 86-87.

DOP 62 The Summer of Blood | 39

on theroute taken route depending (seeAppendix 4 5and would havebeen aboutnineteen days (seeAppendix 3).If themap). Itwould have been necessary for Constans to we assume that the summons was sent sometime between have traveled at to arrive in il forty miles per day Sirmium and 17June (seeAppendix 3), themessenger would on 8 he lefton have arrived in at September(assuming 30August), but that Constantinople right the end of June seems in the extreme. There is no reason or unlikely why during thefirst week ofJuly. Sirmium is about 700 Constanswould stillhave been in at theend of miles and mansiones Aquileia thirty-four from Constantinople. and then make a mad dash for Sirmium at the Constantius a August could have completed leisurely journey lastminute. such a timetable would leave in seventeen Furthermore, days120 and thus easily have been inVimi no to him time conduct the contentious and difficult nacium beforethe end ofJuly (see Appendix 4 route3 thatwe know took negotiations place (see below). How and themap). isno reason ever, there why Constans and his praetorian It is about 1,063 miles from Trier to Sirmium and should have been in the same at therewere or mansiones prefect necessarily city forty-five forty-six (seeAppendix the same time.When Constans traveled tomeet with route two 4 4 and themap). An easymarch of mansiones his his would have remained as and brothers, prefect simply about forty-fivemiles per day would have put Con close to him as while still within stantine II in Sirmium in a over practicable remaining little twenty-three days. to out Italy carry his duties. Therefore, the presence of He could thusvery easily have been inSirmium by the Constans s in in fact indicates middle of on praetorian prefect Aquileia July,having overtaken Athanasius. If, the that Constans was in Sirmium on other he already 29 August. hand, had been summoned by Constantius, These dates then?17 June inTrier forConstantine he stillcould have arrivedthere before the end ofJuly in II, thelast week ofJuly (or first week ofAugust) Vimi (approximatelyforty-two days after the dispatch of Con nacium for and in close stantiuss summons in Constantius, 29 August (or to) early June). Sirmium for us our We can about since we Constans?give parameters. say nothing Constans, do There can be no doubt that themost and not know where he was at or important residing this point, Milan issue s pressing facing the brothers after their father death Rome, but themessenger sent to him would have arrived was to to one meeting together be promoted augusti by the before the sent to Constantius (orConstantine II) to a army, then having this proclamation forwarded Rome and he would have had shorter distance to travel to senate as for acceptance by the and people, Eusebius shows Sirmiumthan his brothers(see Appendix 4 routes3 and did in fact occur. It is clear from Constantine Ils letter 5and themap). of that he was in as senior n. 17June acting his capacity the As notedabove (at 103),I thinkit likely that Sarma It is that itwas he who informed his tian on theDanube was emperor. possible activity prompted by knowledge brothers of the need tomeet, and when and where. As of Constantines death and Dalmatiuss absence from we saw was above, it certainly he who devised theHel the frontier.News of Constantines death would have ena andTheodora coins forall three.Although we do traveledquickly in thewake of thefirst messengers sent not know when he have summoned his broth on 22 might May, and the Sarmatians would have responded itmust have been before as news ers, the letter forAthanasius just quickly. Messengers with of hostilities could since the was of (17June), promotion greater importance, have reached Constantinople from the frontierwithin was no reason to seven to ten a and there (now obvious us) why he would days. In such case, it could have been news It is also it delay. possible that was, rather,Constantius of Sarmatian hostilities that prompted Constantius's who initiative soon as as took the and summoned his brothers departure from the capital, early the end of June?a after his in in more arrival Constantinople early June. If so, departure made all the imperative if the brothers the in is tomeet in a to difference the following timeframes neverthe had already decided Sirmium, city close lessminimal. Sarmatian territory.Conversely, it could have been the A II in to to message fromConstantine Trier Constan need quell the Sarmatian uprising that determined two tinople(the farthest of the capitals)would probably the location of the brothers' meeting. Constantius could have taken about as to arrive as the notification have covered threemansiones just long easily per day throughout a of Constantine's death, less day (the distance between Nicomedia and the of two a more Constantinople), given urgency 120 At mansiones per day he would have faced journey of route than miles a three times. the situation (seeAppendix 4 3 and themap). This fifty day only

DOP 62 40 R.W. Burgess I

and he have been in Sirmium in less his journey, could V. A Reconstruction in orNaissus in even Hypothetical than twelve days, and Viminacium was at less time. When Constantine struck by illness Pythia a it is he realized that itwas serious and that As result of the above analyses, clear, first, that Therma, quickly was in for his two eldest sons would have no means of Constantius probably present Constantinople easy pro a or a a atmost: to ifhe He month month and half from early June moting themselves augustus died. quickly as closest of to lateJune or mid-July. This is consistentwith all the had Constantius summoned the physically it is clear that his two eldest in case he should take a turn for the other evidence presented above. Second, sons, to worse. No doubt he or his advisors believed that even in the brothers should easily have been able accept pro not. extremis Constantius to in motion togetherat theend of July. They did What he could promote augustus We cannot Like events the of the and Constantius could then happened? know. the surrounding presence army, scenarios can Constantine II.When Constantine died themassacre, any number of plausible be promote finally as that for at noon on 22 before Constantius could swift suggested, such problems delayed departure May, arrive, one ormore in were sent to four caesars inform of the brothers, slow travel, complications messengers officially the Constantiuss preparations for the Sarmatian campaign, ingthem of theirfathers (and uncles) death. in the or dur Constantius arrived in fromAnti difficulties subduing Sarmatians, problems Constantinople once but we have ing negotiations the brothers did arrive, och duringthe first days of June and immediatelybegan we s must no facts. All can say for certain is that the itinerar preparations for his father funeral, which have soon two who ies indicate that the gap in time between Constantine's taken place after, absent his brothers, sons was not a in a death and the promotion of his result of remained theircapitals. By theend ofJune meeting to the distances the brothers and theirmessengers had of the brothers had been arranged, either by Constantine II or to at a location travel. But beyond that and the parameters established by Constantius, take place Sirmium, we cannot in for them to find a out above, proceed. central for all three, order way were to ter Once all three in Sirmium, from the end of of their constitutional difficulty and arrange their at not ritorial divisions. This in neutral not in August least if before,they began theirdifficult meeting territory, been the does some and protracted negotiations.121 When these had Constantinople following funeral, suggest were tension completedthey proclaimedaugusti together by the among the brothers, but the location may simply armies on news was sent to in to Sarmatian Danubian 9 September and have been response the uprising. senate itmust was nine Rome for acceptance by the and people. Constantius, be remembered, only teen at none the time and, although he possessed of the at was genius that characterized Octavian that age, he certainly cunning and ruthless, and he had before him his fathersbloody examplesof solvingfamily and dynastic 121 Julians excessive and his claims of the broth protestations repeated Indeed, Constantines sons were all a a s problems. young: ers' ("unity") and a ("perfect unity"), ofConstantius e a a e or nineteen, and thirteen or fourteen. giving his brothers ("no occasion for complaint"), twenty twenty-one, of his his brothers a a a a the same time Constantius was to Constantius s treating ("at married Julius daughter, a e a a justly and moderately"), and of his ("mod his sisterwas married toHannibalianus, son of Flavius eration and all for the a and e and magnanimity"), ("unity" Dalmatius, and Constans was betrothed toAblabius's "peace") of all Romans, indicate just the opposite (Or 1.18D, 19A-20B, The for interferencewith 2.94B-C). In Or. 2.94B-C he admits that Constantine II and Constans daughter, Olympias. potential and with each other not with of at or atworst at quarreled fought (though Constantius, the brothers best their overthrowal the until the former's death. Constantine II was not satisfied course) certainly hands of the elderDalmatius, Julius Constantius, orAbla with the final as he demonstrated less than three later arrangements, years bius must have been obvious to Constantius. After his n. 2 (see 27, above). See RIC 8:7-8; Potter, Roman Empire (n. above), 462; 2 arrivalin he dealtfirst with Ablabius and Frakes, "Dynasty" (n. above), 99-100. See also RIC 8:32,125,170; Constantinople, by Barnes, Constantine and Eusebius, 200 n. 14; and idem,Athanasius and firinghim and no doubt dissolvingthe betrothal of his n. to exer Constantius, 311 5 for the possibility that Constantine II tried brother to There are in Olympias. enough hints the liter cise a over his brothers that had not to. This primacy they clearly agreed sources, Ammianus, and seem to a over see ary especially Julian, Eunapius would also include quasi-regency the young Constans: 2 eine to show that some of soon Bleckmann, "B?rgerkrieg" (n. above), 236-41, esp. "Constantin Vor (Zosimus), type disagreement mundschaft ?ber seinen Bruder Constans aus?bte" arose between Constantius and his relatives j?ngeren (p. 239). (especially

DOP 62 The Summer of Blood 41 |

succession. into his uncle Julius Constantius) concerning the immediately put thedamnationes effect and stopped on Almost certainly this dispute centered the exclusion minting coins forDalmatius. Constans produced nummi of Dalmatius and Hannibalianus from power, in spite with theSEC VRITAS REI PVB reverse,as ifthe state to some of Constantines obvious wishes, and attempts free had justbeen savedfrom danger (Dalmatius and sons or the three from any interference from their elder Julius Constantius, themutinous army?). Constantine to relatives. This decision exclude Dalmatius may well II immediately initiated theHelena and Theodora issues, sons at a since in have been made by the much earlier date, and after his arrival Sirmium he forced his brothers as as not to to suit. early 335 they had clearly all agreed strike follow coins precious metal for him. After reaching theDanube, Constantius quelled the massacres in Almost certainly the took place Con threats from the Sarmatians with Dalmatius's army and or its environs. in stantinople Itwould be only natural for took the title Sarmaticus recognition of the victory to on met all of Constantines physically closest relatives gather (perhaps 27July). At theend ofJuly he Athana so were sius at on there for his funeral, and that many killed all Viminacium, perhaps while still campaign. at once most were some with their supporters suggests that At unknown date before 29 August his brothers in one contentious place. Constantinople, rather than, for instance, arrived and they all began their delibera most a tions new now Naissus, is the obvious location for such gather concerning the division of the empire, was was ing. It only twenty-three mansiones and 486 miles that Dalmatius dead. fromNaissus, where Dalmatius and his fatherwere based, Libanius (Or. 59.73-74) indicates that Constantius so to news itwould have been easy for them have arrived in received of the Persian siege ofNisibis before the was in to time for the funeral. Constantius certainly Con meetingwith his brothers(though he dates it thetime stantinoplewhen he hadAblabius assassinated(though ofConstantine's funeral) and Julian says he learnedof the was Ablabiushimself was not).The army,too, would no doubt Armenian revoltwhile he meeting with them (Or. in in numbers for Since the started somewhere around have been present the capital large 1.18D-19A). siege excuses the funeral, and the mentioned by Julian make themiddle ofJune123 (we know nothing about therevolt), a it almost certain that Constantius and the "mutinous" news travelingat the standardfifty miles daywould were in at time weeks to reach army close physical proximity the of the have taken almost four Constantinople massacre,122 as does the rescue of Julian and his brother, and another twoweeks to reach Sirmium fromConstan more The news traveled much but either byConstantius himself or, plausibly, byMark tinople.124 probably faster, must not of Arethusa (who have been in Constantinople it could have overtaken Constantius before he had was to Sirmium a late from rather than anywhere else). The army employed reached (even assuming departure But no matter situation on assassinate Julius Constantius, Dalmatius Caesar, Opta the capital). what the the no was or news tus, and Hannibalianus, and doubt others, and it frontier how early arrived, Constantius could not not return in was then made the scapegoat for their deaths. Those in until his business Pann?nia done and were he was the city, likeAblabius, hunted down and killed by augustus. were as no set out assassins. Excuses made for their deaths well (such Constantius doubt for Constantinople as we s on read in Eunapius VS). Constantius condemned immediately after his promotion 9 September, eager to return to Antioch. abandoned the of Dalmatius, probably Julius Constantius, and perhaps Shapur siege to so news others, then sentmessengers his brothers informing Nisibis around mid-August, and that probably At some between the in at the latest. them ofwhat had happened. date reached Constantinople mid-September set Thus time Constantius was able to reach Con veryend of June and themiddle of Julyhe offfor by the were now in ruins. before the lastweek of the he had Sirmium. Constantines plans stantinople, month, II Constans received the news of retreat Libanius In Gaul and Italy, Constantine and already Shapur's (as

massacre the obvious excuse 123 See Burgess, Studies (n. 25 above), 232-38. 122 If the had happened anywhere else, was not 124 News would have traveled from Nisibis via Antioch. As would have been that Constantius physically present and therefore probably was to actions: how he have? The we have itwas about 800 miles fromAntioch to and unable have prevented the army's could seen, Constantinople mentions also close between more than miles to Sirmium from Itwas almost "deception" that Julian indicates proximity 700 Constantinople. 450 to Constantius and the army. miles from Nisibis Antioch via Edessa (following the Tab. Peut).

DOP 62 42 R.W.Burgess I

turn he have had notes, Or. 59.75) and could therefore his attention have known about whatever plans may toother matters, chiefly the deposition of Paul and the beforehand. Assassination need not have been apart as there have been to remove Dalmatius ordination of Eusebius ofNicomedia bishop of Con of any plan may He to to and his from the succession and the stantinople.125 then returned Antioch prepare family posi to invasion. tions of influence that had been to them his response Shapur's granted over the brothers. The intention was at the time to remove Theodoras male relatives from the dynastic ensure nor succession and to that neither they their or sons supporters could influence threaten the three of Constantine at a later date. Constantius did, no VI. Conclusions doubt, feel remorse later on for his actions. The assassinations took in or can be sum 4. place Constantinople The major conclusions of this paper, then, in are its environs early June of 337. The details lost marized briefly: to scenarios can be history, though many plausible . s sons a Constantine three showed marked hostility imagined. was one set in towardDalmatius fromthe very beginning of his 5. There only of assassinations, which all to or silver coins in his of Constantius s were removed within a reign, refusing strike gold opponents was to name at their home mints. short space of time.The army employed murder 1. and silver to Dalmatius Caesar and Constantine used the gold coinage pro julius Constantius, Optatus, mote two sons two but assassins were from his eldest equally above the other Hannibalianus, dispatched caesars over to execute like who in spite of Constantine Us seniority Constantinople those, Ablabius, Constantius. This confirms other evidence that indi were not within the soldiers' reach. Damnationes cates was a return to memoriae'were Dalmatius Caesar thatConstantine planning for proclaimed against a II ifnot others.We tetrarchie system headed by Constantine and and probably Julius Constantius, one on not Constantius, but based blood ties of family do know the basis for these condemnations. ties as was case 6. Constantius left between the end rather than simply of marriage, the Constantinople sons were to on with the first tetrarchy. His young and ofJune and themiddle ofJuly campaign the He he intended that Flavius Dalmatius, Julius Con Danube against the Sarmatians. may have cel act as over on stantius, Ablabius, and perhaps others would ebratedvictory them 27July (thus implying to even as sons an advisors and perhaps regents for his early departure). were to assume Constantius met Athanasius atViminacium at the and nephew until they old enough 7. on own. or power their veryend ofJuly thebeginning of August 337. can no was met in 3. There be serious doubt thatConstantius 8. The three brothers Sirmium, and although all was we not behind the assassinations and that it he who could have arrived during July, do know why s succession not it took until for them to be rejected his father and dynastic plans, 9 September proclaimed not mean the soldiers. This does necessarily that the augusti by theDanubian army. were in II was for and assassinations planned advance; they could 9. Constantine responsible designing as nummi. have been, but they could well have been the result issuing the Helena and Theodora They a reaction on s to have been an act of atonement to their of spur-of-the-moment Constantius appear to some or for the assassinations of her sons part real perceived impasse, difficulty, step-grandmother or on or as Theodora her threat the part of Dalmatius, his family, and grandsons. Just represented cannot can we sons and so Helena their supporters.We know. Nor know dead grandsons, represented howmuch his brothers(especially Constans) may her three living grandsons. Furthermore, Theodora represents maternal piety, while Helena represents a promised imperial peace (i.e.,within the imperial 125 Socrates, HE 2.7,with T. D. Barnes, and A.D. "Emperor Bishops, These coins first from Trier in the in family). appeared 324-344: Some Problems," A]AH 3 (1978): 53-75 (reprinted Early 66 immediate aftermath of the news of the assassinations and theRoman Empire [London, 1984], paper XVIII), and Barnes, Athanasius and Constantius and were in (n. 35 above), 213 219. and struck there great numbers throughout

DOP 62 The Summer of Blood | 43

the and to in on interregnum continued be struck large great failures of his personal and professional life rest numbers for the of Constantine Ils reign. At his actions in 337. in theirmeeting Sirmium, Constantine II compelled to coins at his brothers strike similar Constantino More could perhaps be teased from the sources, and many so scenarios ple andRome, thoughthey did only reluctantly, alternative and plausible theories could be pre in at am intermittently,and smaller numbers than Trier. sented, but many readers, I sure,may feel that I have was too on too as With Constantine Us death their production already gone far little it is. So be it.But it at immediately shut down all mints, ismy hope thatthe evidence presented here will setour io. The version events over time. events on a new more official of the evolved understanding of these and solid At were or first there the damnationes and nothing footing, whether the individual overall conclusions was was are or not. said about the murders. Then the army of this paper accepted was as a blamed and Constantius presented hero Universityof Ottawa Later for resisting their uprising and restoring order. Department ofClassics and ReligiousStudies was as in Constantius presented helpless the face of 70 Laurier Ave. East militarydeception and overwhelmingodds against Ottawa, Ontario KiN 6Ns was themutinous soldiers. Finally his involvement Canada no was to longerdenied and he said have repented [email protected] ofhis role in themurders and tohave blamed all the

Constantines. Plans for Succession as Seen in the Appendix Coinage

The jointand equal superiorityof thetwo eldest caesars, various twinned obverse and reverse typeswith an overall over two not exact Constantine II and Constantius II, their col though parity (nos. 65-66,69-71,93-96,105, Constans and in the to Constantius more solidus leagues, Dalmatius, period 333 109-12,115-16,119-20). had is in 337 made clear by the contemporary coinage. types 333, the year of his quindecennalia (nos. 70-72), mint in II more vicen At the Constantinople, Constantine's home and Constantine has types celebrating his on two in II mint, silver produced between 333 and 337 the nalia 336-37 (nos. 116,119-20). Constantine and caesars are on one issue in are a junior named only tiny 337 Dalmatius missing from the obverses of donative {RIC7 nos. 136and 136A[see p. 719]),while Constantine Festaureus series inwhich Constans appears (nos. 103-6), in most Constantine are II and Constantius have rough parity (for the though all four types (two for I) repre reverse the sented one coin so one was part) twinned obverse and types throughout by only surviving each, type RIC nos. struck forConstantine II as and period: 333-34 (see pp. 718-19), 55-57A (see probably originally well, nos. is on p. 718); 335-37 (see pp. 718-19), 123-25,127,127A-30, perhaps forDalmatius. Constantius alone named the one two at 131B, 133-34, 135A (see p. 719). Apart from nine obverses of unique medallions minted the end of no. in name or see solidus medallion from 333 (RIC 67, the 335 very beginning of 336 (nos. 88-89; for the date caesar is nn. of Constans; described below) neither junior 88-89, p. 583), though others naming Constantine a named on thegold between 333and thevery end of 335 IImay well havebeen minted. The first(no. 88) depicts are on a I two cae (nos. 64-72, 87-89). They named special issue, nimbate Constantine enthroned between equal in sars both obverse and reverse, early 336 (nos. 97-98; cf. (SALVSET SPES REIPVBLICAE, "TheSafety and are nos. 90-96) but fromthen until September 337 they Hope of theState"), obviously his eldestsons, while the on one 102: issue him enthroned named only sesquisolidus (no. Dalmatius; other contemporary (no. 89) depicts a 1.5 solidi), an aureus (no. 106: Constans), solidus (no. amongall fourcaesars (SEC VRITAS PERPET VA, "Eter a 121: In outer caesars inner 113:Dalmatius), and fraction (no. Constans). nal Security"), the of equal height, the two senior caesars continue to share their caesars of amuch shorter The medallion the gold the yet equal height.

DOP 62 44 R.W. Burgess I

of struck in the name of Constans noted above once: no. can 333 (no. (Dalmatius only 147). Little specific there offers the same reverse a caesars on 67) legend and shows standing fore be said about the representations of the Constantine I a holding parazonium (a short, sheathed these coins.128 sword cradled in the left a of and vexil At Constantius arm, sign valor) Thessalonica, slightly surpasses his two caesars a reverse lum, flanked by equal each with long scepter elder brother in the number of obverse and types and shield and a much shorter caesar on his with at of right the very end 335and during 336 (nos. 209-11,215-16, to a in what appears be short staff his right hand. On 219-20A [seep. 718])but is strangelyabsent from three reverse no. the of 89 (above) it is the outer, taller caesars earlier silver types of 335 that focus on Constantine II who each hold a In reverses one no. parazonium. all these the (nos. 194-96, plus for Constantine I, 197). A of the two elder caesars is name parity obvious, yet according nine-solidus medallion in the of Constantine I to one strict seniority would expect all caesars to be of mirrors the nine-solidus medallion described above from a size. two a decreasing Only the elder brothers share Constantinople(SALVS ET SPES REIPVBLICAE), but reverse series of tremisses as special solidi and with their depictsConstans slightlytaller and Dalmatius (who is one an as father(nos. 107-8,114-20:VICTORIA CONSTAN theonly without invertedspear) slightlysmaller on TINIAVG, VICTORIA CONSTANTINI CAES(AR), than they appear theConstantinopolitan medallion, VICTORIA CONSTANTI [sic]CAESAR). Though yetboth remainsmaller than the two equal figuresof caesars Constantius ismissing from the tremissis series inRIC, the elder (no. 204). a at one a specimen has recently appeared auction.126 Only the At Heraclea, Constantius has obverse type in reverse sesquisolidus forDalmatius and the fraction forConstans single silver serieswhile Constantine II has two notedabove celebrate the "VIRTVS CAESARVM NN," (nos. 146,148-49). that is, the "Valor of theTwo Caesars," who must be the At Nicomedia there are two similar nine-solidus two caesars. junior medallions in the names of Constantine and his eldest At Trier Constantine II and Constantius likewise son reverse a was (the shows that type certainly minted for a dominate thegold and silvertypes with roughparity Constantius), which depicts nimbate Constantine enthroned with one caesar to either side (nos. 565-68,570 [Constantius only], 572-74,127 581-84 and the legend [Constantius has three obverse variants]), whereas Con FELICITAS PERPETVA AVG ET CAESS NN ("Eter stans does not on the until late or our our appear gold 335 336 nal good luck for augustus and two caesars"; nos. or on (nos. 575-76; cf. 564-70), the silver (no. 585) until 173-74). This medallion was minted in 335,a year to two as 336-37. years after the proclamation ofConstans caesar (for the At Rome is issues see n. Constans missing from all gold date, RIC 7: 627, 160); itmirrors the similar con between and the 333 337 (nos. 340-41, 374-75), namely temporary medallion from Constantinople noted above typewith thereverse VICTORIA NOB CAESS, "Vic (no. 88,SALVS ET SPES REIPVBLICAE). Therewas of theTwo Noble in name a tory Caesars," struck only the also specialissue of solidi in 335only for the two eldest ofConstantine II (nos. 341,374-75) and Constantius (no. brotherswith thereverse legend VIRT VS CONSTAN two caesars are on 341). Likewise, only the eldest named TINI CAES andVIRT VS CONSTANTI CAES (nos. the silverminted Since Rome There are two 336/37 (nos. 376,378-80). 181-82). single specimens of argentei of was under the jurisdiction of Constans himself, this lack different reverse types in the names of Constantine I, of is representation puzzling. 128 In RIC Constantine II is from the silver series of Apart froma gold seriesof 334 (nos. 225-28), the missing 334 (nos. 129-32) and Constantius is from the series of 335 (nos. and silver of Siscia 229-34, 242-51, missing gold gold (nos. 257-60) and the silver series of 242-51) 336-37 (nos. 259-60), though the latter are and there are incompletely preserved many types marks Constantine IFs vicennalia, so an obverse forConstantius would that for one or other of the not be an solidus and in the appear only caesars, though expected. However, unpublished argenteus name of Constantius date to the but for all four do appear on the obverses between 334 and 337 clearly period 335-36, parallels at no. the types appear only other mints (Antioch 97 and Thessalonica no. no. and no. 216, respectively [Heraclea 146 Constantinople 127A 126 & Mosch Giessener auction Gorny M?nzhandlung 133 lot 529, are similar to the latter have a reverse (p. 719) but different legend break]). Ii October 2004. at This implies parallel types Siscia in the names of the other caesars that 127 RIC lists obverses for Constantine II and but a have not only Constans, survived. For these unpublished coins, see Numismatica Ars for come to specimen Constantius has recently auction: Numismatik Classica auction 25 lot 597, 25 June 2003, and H. D. Rauch auction 71 Lanz auction 128 lot 22 2006. 28 895, May lot 1063, April 2003.

DOP 62 The Summer of Blood | 45

Dalmatius, and Constantius (nos. 186A-187) but issues 97 [gold]and 105-7 [silver]),but we would expecthim same to in since reverse of the types for the other members of the college have appeared both, the of the former no doubt existed. Constantine II and Constans share a refers to him (see above) and Constantine I,Constantine reverse a a in we type for semissis and tremissis (nos. 183-84). II, and Constans appear the latter. Similarly, would II to in On thegold from Antioch between335 and 337there expect Constantine have appeared theVICTO are solidi in the name of Constantine II and Constantius RIA solidus series (nos. 102-4, above), which has obverses with thereverse legends PRINCIPI IVVENTVTIS, "To forConstantius and Constans. theleader of theyouth," and VICTORIA CAESAR NN, In conclusion, while Constantius may not overall our two an on as "Victory of caesars" (nos. 94-95,97,102-3), the make appearance quite many surviving obverse must to two caesars even or reverse or variants as latterof which refer those though types of types Constantine II, an in name can no mints two there is obverse the ofConstans (no. 104). The there be question that the presented the a as is a an as to latter isprobably mule, solidus with obverse of brothers of equal rank, second only Constantine a reverse I on reverses. The Constantine II and with the legend VICTO both obverses and purpose of such at two RIA CONSTANTINIAVG (no. 101;cf. 98-100, all with promotion, the expense of the youngest caesars, at can a intention to obverses of Constantine I). As other mints, Constans only have been result of Constantines not on or sons to as does appear thegold silveruntil 336/37(nos. have both succeed his position augustus. two series 104,107). Constantius ismissing from (nos.

Appendix 2.Hoard Evidencefor the Date of theMassacre

At Arles, thefirst GLORIA EXERCITVS one-standard The most plausible explanationfor thisdegree of (GE (i)) reversetype carried the mintmark PCQNSTand rarity is that the issues for Constantine and Dalma was a issue tius or and almost followed by smaller with the mintmark, ceased, simultaneously nearly so, issues were in names as soon as mintmark There were more pcoNST-129These struck the of all the changed. Con coins struck for Constantine because of the fivemembers of the imperial college: Constantine, originally see stantine II, Constantius, Constans, and Dalmatius (for hierarchyof striking(on which, below).After the not the see cessation of the Constantine and Dalmatius the examples of the types, though mint, Figs. types a new GE reverses for 20-2,5). In 337 mintmark appeared, PConst> ^e pcoNST niintmark continued with (i) firstthat would be employedon thenew issuesof the Constantine II, Constantius, and Constans as caesar. threebrothers as augusti (RIC 8:205) and thereforethe There can be littledoubt, then,that the PC(j^STnummi as caesar. were were last of their issues These coins also issued produced almost entirely between the deaths names of Dalmatius and the first issues of in the of all fivemembers of the college. However, Constantine and are as in the the coins minted for Constantine and Dalmatius the brothers augusti.131 Interestingly, spite of rare?so in known time frame available for the serieswas extremely rare, fact, that the only short it, PC(^ST were since it the issue in specimens found in theWoodeaton hoard, four large, outnumbers PC(^ST hoards, one itwas not so as the issue.132 for Constantine and forDalmatius,130 indicating though nearly large PC(^ST were we a situation. that coins for these two emperors struck inmuch In Rome have similar Through 336 smaller numbers than those for the three caesars. are so 131 They dated inRIC 7:278-79 and RIC 8:197. 129 In mintmarks the "P" indicates the officina and is used as are (prima) 132 For instance, in theWoodeaton hoard there 6 ? specimens a standard form for citation. The other officin?? had two in total, 18 two 2 in (Arles and 6 O specimens; in the Freston hoard, ^, X, and O; the Rome had used theirown letters T, and 21 1 2 in the & and five) (S, Q, ?). Appleford hoard X, and O; Ihnasyah hoard 5 nos. in ? 2 in the & 130 Woodeaton hoard (see n. 137),Constantine 1314-17, and Dal O; the Bicester hoard 4 and O; Chorleywood 31 as in Hamble hoard and in theMetternich hoard ? is matius no. 1318 (misnumbered 1319 in the text, but correctly labeled and 5O; the 4 rep on are so rare inRIC resented all five O two, and X one. For these Plate 12). They that neither appears 7 (pp. 278-79) by emperors, by by hoards, see n. and I have found no other reference to further specimens. below, 137.

DOP 62 R.W.Burgess 46 I

into are A western us to and 337all fivemembers of the imperialcollege number of European hoards allow the standard GE reverse. This con establish the of the bronze issues represented by (1) very exactly chronology tinues R*P to fromTrier.137 The was in with thechange of mintmark from R^P, Ollmuth hoard closed the early same autumn one the mintmarkthat is used forthe first issues of the of 337: it contained only coin struck after three as a brothers augusti (RIC8:249-50). Consequently, thebrothers became augusti, PIETAS ROMANA for this mintmark indicates their last issues as caesar. The Theodorawith ^ ,the first mintmark used by thebroth standard GE (1) reverse ismaintained for Constantine ersas augusti(RIC 8:143).Almost half thehoard, 74 out and are of was in issue I, Constantius, Constans,133 though they very 155specimens, minted Trier, and the largest rare.Almost immediately after themintmark change, the bears themintmark #TRP# (28 of 74), thelast mark used GE reverse forConstantine Iwas a new (1) replaced with before the brothers became augusti. The distribution of reverse,unique toRome: VIRTVS AVGVSTI (Fig.26, the obverses of this issue is set out in column O ofTable a similar struck at the in name . one specimen end of 337 the Although havingonly certainpost-9 September of Constantine II as reverse it contains augustus).134 But theGE (1) coin, three #TRP# pieces forHelena and five for the caesars then to forTheodora. And in of nineteen quickly changes SECVRITAS spite having specimens REI a at in the names of the as PVB(licae) (Fig. 5, specimenstruck theend of three brothers caesar, ithas only in the name of Constantius as one none 337 augustus) for only the forDalmatius and forConstantine. as caesar three brothers (RICj, nos. 402-4). At the same More thanhalf of theWeeze hoard ismade up of timethe VIRTVS AVGVSTI typefor Constantine I is issues from Trier (668 out of 1,198). Itwas closed a little The SECVRITAS PVB is in earlier than the dropped. REI legend unique Ollmuth hoard, during the interregnum, this form on theConstantinian and this since it no issues as coinage particular has of the three brothers augusti, and of typeis unique on thebronze, though a similarone had the 668 coins fromTrier, 217 have the #TRP? mintmark, at to been used earlier Trierbetween 317 and 336chiefly for the last appear under the three caesars. The distribution are no name solidi.135 There coins in the ofDalmatius for among the emperors is listed inTable 1,column W. As in this hewas a the Ollmuth are mintmark,though partof theearlier GE (1) hoard, although there many coins for down toR*P. As was the case with the issue sons as are issues, PC(^ST the caesar, there very few forConstantine and ofAries, therecan be no doubt thatthe SECVRITAS Dalmatius. On the other hand, the number of Helena REI PVB was issued the after and Theodora is type during interregnum specimens extremely large. Constantine and Dalmatius's deaths.136 The publicationof theCranfield hoard isnot very The pattern atAries and Rome is clear: all fivemem 137 The hoards to be discussed are as follows: Ollmuth, Westdeutsche bersof the imperialcollege were originallyrepresented Geschichte und Kunst j (1888): 123-24; Weeze, Westdeutsche on the nummi. Then Constantine and Dalmatius dis Zeitschriftf?r Zeitschriftf?r Geschichteund Kunst j (1888):124-29; Chorleywood, in at about the same time. At both coins . appear mints, Carson and Burnett, Recent Coin Hoards ( 107 above), 4 _98; Ham in vol. 6 produced forConstantine II, Constantius, and Constans ble, ibid.; Cranfield,iVCser. 6, (1946): 159-62; Woodeaton, NC ser. vol. 12 as caesar continue after cessation 138 (1978): 38-65; Freston, NC 7, (1972): 145-57; Appleford, the of those produced RBN123 (1977): 41-100; Bruckneudorf,NZ89 (1974): 5-17; Ihnasyah, forConstantine and Dalmatius, and occur in they large JIAN16 (1914): 1-27; Metternich, i?/145 (1940): 80-125; Halifax, York numbers atAries. shire ArchaeologicalJournal 23 (1914-15): 444-51; and Bicester, Coin Hoards from Roman Britain, vol. 2, ed. A. M. Burnett, British Museum are Occasional Paper 31 (London, 1981). These all the contemporary and 133 RIC 7 lists the first two struck only (nos. 400-401). Specimens are in near-contemporary hoards that well reported the literature. The in the name of Constans in the hoard. There appear Ihnasyah certainly ser. Bishop's Wood hoard (NC 3, vol. 16 [1896]: 209-37) nas been left must have been an issue for Constantine II as well. out because the mintmarks are recorded and and^types poorly include 134 no. known from a RIC'7, 405, unique specimen. AlthoughLRBC incorrect mintmarks as xrs for 19 of the 30 issues for recordsVIRTVS AVGVSTI with theR*P mintmark V. Hill and P. (such Dalmatius); (P. J. see Carson and Burnett (above), 45. The breakdown of the hoard is very C. Kent, Late Roman Bronze A.D. $2 4-4g S. Parti: The Bronze Coinage, as peculiar well. it contains 2,455 for Constantine, the A.D. 6 no. Although specimens Coinageof Houseoj"Constantine, 324 -34 [London, 1978], 15, 3,679 forConstantine II as caesar, and 2,197 forConstantius as caesar, it the editor was to 566), 0?RIC7, Patrick Bruun, unable verify its existence contains none at all forConstans as caesar. For the brothers as it n. augusti (pp. 295 and 344 391). As a result I do not take account of it here. has 4 forConstantine II, 4 for Constantius, and 450 for Constans. And 221. same 135 RIC 7:178,185-86, 211, and The on in legend appeared spite of itsmany eTRP# specimens forConstantine I,Constantine II, bronze reverses was forHelena between 324 and 329, but the type differ no and Constantius, it apparently has eTRPe specimens forDalmatius, ent a (see RIC 7:750 for list of themany issues and mints). or even Helena, Theodora, though it contains 312 specimens forHelena are so 136 They dated inRIC 7:294-95 and RIC 8:234. and 271 for Theodora.

DOP 62 The Summer of Blood |47

? ? detailed, but there are no coins of the three brothers as Table Distribution: of issues in so itmust same time augusti, have been closed about the threehoards of late 337 as were in theWeeze hoard. Of 1,700 coins, 517 minted owe Trier. There are no coins in the names of Constantine or no. % no. % no. % there are six for the sons as caesar. The Dalmatius, though GE (1)and Helena and Theodora types(no mintmarks 0 CI 0.0 17 7.8 0 0.0 are aregiven, though for the latterall must be #TRP#) 11CU 39.3 74 34.1 2 12.5 listed inTable 1,column below. C, 7 Cs 25.0 36 16.6 3 18.8 These threehoards show a doubta numberof beyond 1 Cn 3.6 15 6.9 1 6.2 in name important facts. First,minting the of Constan 0 0.0 soon D 1 3.6 3 1.4 tine and Dalmatius stopped very after the change ofmintmark and at almost the same time; second, mint 3 10.7 40 18.4 6 37.5 ingof theHelena and Theodora typesbegan with the 5 17.9 32 14.8 4 25.0 TRP# mintmark while Constantine II, Constantius, and Constans were still caesars; and theHelena third, O: Ollmuth; W: Weeze; C: Cranfield were in numbers: in CI: CII: Constantine Cs: Cn: and Theodora types produced large Constantine; II; Constantius; Constans; D: Dalmatius; H: Helena; T: Theodora theOllmuth hoard theyaccount for almost 30 percent of in the surviving specimens, theWeeze exactly 33percent, over 60 and in the Cranfield percent.138 ratios continued the summer of from to these coins? These large beyond produced 335 337.Unfortunately can notations themintmarks 337 and be verified from the frequency with TRP, TRP*, and^/TRP?rarely occur recorded inRIC 8 (pp. 143-44). Both Helena and in the hoards under examination here, respectively were next two are so Theodora types struck with each of the six only 65, 2, and 5 specimens (the latter marks rare not even in mintmarks from Trier that include types for Constan that they do appear RIC 7). Comparable autumn to six ratios on the from other mints are of value tine II from 337 early 340. In five of these based output in a in the #TRP# emissions, each Helena and Theodora type outnumbers general way evaluating issues, although sometimes it should be out that exact cannot every other obverse type, quite significantly, pointed comparisons exceptfor TRPj#, where theyare equaled by one type bemade. InTable 4 below I have includedGE (1) issues for all five from mintmarks and outnumbered by another. emperors well-represented us to 1. in Let return Table A second interesting feature inGaul (Lugdunumand Aries) found contemporary eastern western issues of thistable is thefact that the numbers for Constantine hoards and from and from large are so low. This can be confirmed from other contempo contemporary hoards in bothWest and East. 2 it rary hoards as well, as can be seen inTables and 3. Examination of the data makes clear that there coins was a of with Constantine Ils coins The extremely low number of forConstantine hierarchy striking, 1 issues fromwestern Aries. I and Dalmatius observed in Table is confirmed by dominating the mints, except was in as these additional seven hoards. This pattern also noted Constantine clearly dominates the East, would be context and Dalmatius all above in the of their representation among the expected. Constantius, Constans, same of coins in all col last mintmarks from Aries and Rome in 337. The best have roughly the proportion to numbers is to for Constantius and Constans's issues in way evaluate the significance of these umns, except to in compare the output of these emissions those of earlier Lugdunum (lower)and Dalmatius's Aries (higher).139 Trier issuesfor all fivemembers of the imperialcollege Constantine II dominates in theWest because of his as caesar in West. position senior and senior emperor the In theEast he is ahead ofConstantius, who 138 It shouldbe noted thatin spiteof theearly publication dates for only slightly no RIC LRBC these three hoards catalogue, including 7 and (n. 134 to be but there are above), lists these Helena and Theodora types before 9 September 337. 139 Lyons appears anomalous, many possibilities was to account for not least data. The o? RIC Since the eTRP# mintmark reused in 338/39, all publications of later it, incomplete frequency figures mirror cf. also but are of little value hoards and RIC include the earlier eTRP* typeswith the later coinage? 7 my totals (nos. 285-88; 271-72), at no to them?thus the since it is clear that few hoards were consulted I can cite 26 GE since there is present way distinguish skewing (e.g., (1) issues nos. and %(PLG issues forConstantine from five hoards inTable 4, RIC 7 cites survival figures for those later (RIC 8, 63 65). yet

DOP 62 48 R.W. Burgess

2: ? ? Table GE (i) pre-9 September 337 issuesfrom seven contemporary and later hoards.

4 12 3 5 6 7 total

CI 0 17 9 4 12 5 49 6.6

CU 11 74 84 51 72 55 354 48.0

Cs 7 36 61 12 52 52 224 30.4

Cn 1 15 27 7 29 15 95 12.9

D 1 3 3 2 4 3 16 2.2

: : 6: Ollmuth; Weeze; 3:Chorleywood; 4: Hamble; 5:Woodeaton; Freston; 7: Appleford

Table 3:TRP and *TRP* GE (i)1

Hal Men

CI 1 8 7.1

cu 20 31 40.5

Cs 12 27 31.0

Cn 6 18 19.0

D 0 3 2.4

Hal: Halifax; Mett: Metternich; %: percent of total

isonly slightlyahead ofConstans (2.6percent and 1.9 stans s in number, but that isnot the case. Aries seems to percent respectively). Constantius and Constans have be anomalous in this context, because the other groups fewer coins each because of their lower show that the coins of were progressively Dalmatius normally present and in amounts between 6 and 8 seniority, averaging 19 15percent respectively overall, percent (average 6.7 percent), ifwe discount the low survival rate of coins is unusually which about half of the percentages forConstans. This from in column 1.Since Dalmatius was at Lyons proclaimed higherfigure for Dalmatius Aries in the PC(^ST issue two fewer than years afterConstans (seven years separate highlightsjust how remarkablethe drop too percentin Constantine II and and eleven the issue is Constantius, separate succeeding PC(^ST (see above). Constantius and Constans), his representation in these The differencesbetween Table 4 (mostlypre-22 May issues should be about the same as 2 theoretically Con 337)and Table (mostlypost-22 May 337)are quite strik ing.The hierarchyof strikingnoted above is stillvisible a in the BM: no. the only single specimen 2.85; of hoards cited here only inTable 2 but for percentages Constantine have dropped appears in theRIC The same can be said Appleford bibliography, p. xxv). to over one of his usual in for theAries in are in just quarter average output figures, but addition they anomalous comparison with the earlier and Dalmatius to under one third. thefigures given inRIC y (nos.394-99), but againRIC is lacking period just thehoard data This confirms the (againonly Appleford). conclusion, stated above, thatminting 140 These hoards are from Table 2 because are not at same segregated they stopped almost the time for both Constantine mintmark in the Thus the totals distinguished by original publications. and Dalmatius, soon after Trier switched to the contain at some very least non-#TRP#, issues, except for CII and Cn in the TRP? while coins in the name Metternich which do not. As mintmark, of the three hoard, will be explained below, however, are caesars and Helena and Theodora were struck the pre- #TRP# GE (1) issues not voluminous and, as can be seen, do throughout not distort the the the in percentages excessively. interregnum, latter large numbers.

DOP 62 The Summer of Blood

Table 4: Distribution of obverse typeswith pre-9 Sept. 337GE (1) reverses

% % no. %

CI 26 21.9 25 26.9 61 25.0 74 25.2 308 42.1

CU 72 60.5 20 21.5 83 31.0 102 34.7 143 19.5

Cs 7 5.9 17 18.3 50 20.5 60 20.4 124 16.9

Cn 4 3.4 15 16.1 35 14.3 41 14.0 110 15.0

D 10 8.4 16 17.2 15 6.2 17 5.8 47 6.4

Totals 119 93 244 294 732

. #PLG (Lugdunum) (Woodeaton, Freston, Chorleywood, Appleford, Weeze) 2? PCONST (Arles) (Woodeaton, Freston, Chorleywood, Hamble, Appleford, Bruckneudorf, Ihnasyah, Bicester) 3.Metternich hoard, western issues mints 4. Metternich hoard, all issues, including uncertain eastern 5. Ihnasyah hoard, issues, including Siscia

on Appendix 3.The Speed ofTravel Foot andHorseback were to cover in were Ordinary Romans expected the distance per day Italy,where the roads much better than two in one on In in rest between mansiones day foot. general this the of the empire.143 out to works between sixteen and twenty-fivemiles per A numberof close groupings of lawsissued by Diocle an a tian in and Constantine in day,which suggests average of around twenty, figure 290,293, 294, and by 318,320, in a it that is explicitly stated number of sources,141 though and 326make clear that Diocletian normally traveled one or two the actual distance between mansiones varied consider either mansiones per day, sometimes three, on terrain cover to in a ably depending the and local conditions.142 and could easily twenty thirtymiles day, not Couriers forthe imperialpost carryingdispatches along if more, and that Constantine could likewise easily cursus on are cover over miles comitatus.1*4 the publicus horseback generally thought thirty per day with his to to or accounts have been able travel about fiftymiles, between Papyrus records and of the trip of Theo two to an from Pelusium and threemansiones per day, and up eightymiles phanes, imperial officialwho traveled to cursus in Egypt Antioch and back using the publicus in 2.11.1 as well as between 317 and 323, show that could 141 See, for example, Gaius's prescription Digesta ordinary people . . statement an cover same as , 38.15.2.3, and 50.16.3, and Vegetius's that army should easily the distance the emperors with their to cover in at a be able twenty miles five hours regular pace and twenty since comitatus, Theophanes averaged between thirty at a is an four miles quick step (1.9). This during the summer, when hour was over to just seventy-six minutes. However, thirty thirty-five miles seems not to in the A. . "The of the Roman per day have been unusual for the army Republic. 143 Ramsay, Speed Imperial Vosu'JRS 15 see in in next note. The Roman C. W. "New Evidence for the of the For this, particular the works cited the (1925): 63, 68-69; J. Eliot, Speed was meters or Roman L. in mile 1480 1618.5 yards. Imperial Post," Phoenix 9 (1955): 76-80; Casson, Travel see theAncient World 1994 [1974]), 188; O. Perler, Les 142 For the mansiones of Italy, Britain, Gaul, Spain, and Asia, (Baltimore, Voyages R. de saint (Paris, 1969), 31;R. Chevallier, Roman Roads (Berke Laurence, The Roads ofRoman Italy: Mobility and Cultural Change Augustin An Inter 1976), 191-95 and Laurence, Roads, 81-82. (London, 1999), 88-92, and idem, "The Creation of Geography: ley, in and locations of can be seen most in pretation of Roman Britain," Adams and Laurence, Travel and Geog 144 The dates these laws easily in New raphy (n. above), 81-87. Barnes, Empire (n. 3 above), 51-54, 74, 77.

DOP 62 R.W. Burgess 50 I

one two mansiones on and fortymiles, and mansiones, per day dur miles and 84 fromNicomedia, depending route ingtwo tripsof eighteendays out and sixteendays back the routetaken (the throughthe Alps is shorter mansiones. outward but involves more see route covering twenty-six On his longer mansiones-, Appendix 4 4 one over in and the and we know that Constantine II knew journey he covered mansio per day four days map), ten over six nine over was the desert, then mansiones days, about his fathers death and already acting upon next on it on later the eight,and thenthree the lastday, when he 17 June, twenty-six days (Athanasius, Apol. into no c.Ar. and Hist. Ar. Ifwe assume that news was covered sixty-fourmiles Antioch, doubt leaving 87 8). return he sent toConstantine II fromNicomedia on early and arriving late.On his shorter journey immediately two over ten one he covered mansiones per day days and per the afternoon of the twenty-second ofMay and that over the news the before his letter forAthanasius day six days.145 received day in circumstances In addition, special and emergency (whichwould hardlybe thefirst item of businessafter are to the but messengers and other individuals known have hearing about death of his father, let that pass to a Theo for it been covered up 150miles day.146 On 29 June 431, the sake of argument), would have necessary sent a to news to to at dosius II letter from Constantinople Cyril of for the have traveled him about sixty-nine was to to three mansiones twice Alexandria inEphesus, andCyril able reply it miles and per day, Theophanes' on were 28 rate. This to July.147There approximately 478 miles and messenger would have have traveled faster news was not sent mansiones between Constantinople and Ephesus, and if immediately after Constantines so must was a a Theodosius's messenger, Palladius, have covered death and ifthere gap of numberof days before 160miles and nine mansiones Constantine II could write the letter for as approximately per day,with, Athanasius, no doubt, little time for rest. This seems excessive, but would in fact be reasonable. Under these circumstances, was so sowell which allow a time and an Palladius's speed remarkable and known traveling of, say, twenty days an entire to him in on 11 a that Socrates devoted paragraph his arrival June, messenger between Nicomedia and us history(HE 7.19).These figures give an absoluteupper Trier would have covered between 86 and 113miles and limit for emergency travel. aboutfive mansiones per day (withfive days of fourand can time for one three: varies These data be compared with the taken of the distance considerably between news to a the of Constantines death reach Trier, when mansiones along this stretch of road). As result, five was not or s mansiones seems a rate speed important but Tiberius Palladius break per day reasonable of progress itineraries we can in summer neck speed. From the calculate that for emergency travel the of 337. Trier was between 1,785miles and 89 mansiones and 1,831

145 Catalogueofthe Greek and Latin Papyri in theJohn Ry lands Library Manchester 4, ed. C. H. Roberts and E. G. Turner (Manchester, 1952), nos. 627-28 and 630-38. See also Roberts and Turner (above), 106-7 The and Casson, Travel(n. 143 above), 190-93. J.Matthews, Journey of in Roman East Theophanes: Travel, Business, and Daily Life the (New on Haven, 2006) provides complete translations of and commentaries and these fascinating valuable documents.

146 Ramsay, "Speed," 62-65, 67; Casson, Travel, 188; and Laurence, 81. most is in c E was Roads, The famous example Tiberius, who 9 able to 182 to travel miles from Ticinum the bedside of his dying brother, at on Drusus, his camp the Elbe within twenty-four hours (Pliny,Nat ural History 7.84). = 147 Frag. Vat. 83-84 E. Schwartz, Acta conciliorum oecumenicorum 1.1.3 (Berlin, 1927), 10.

DOP62 The Summer of Blood | 51

Appendix 4. RoutesAcross theEmpire

Figure(s) inbrackets indicates map number in Atlas. Vindonissa (i8), Brigantium (19),Cambodunum in are main Cities cited italics the departure and arrival (19), Bratananium (19). Pons Aeni (19), Iuvavum cities discussed in the text above, (19) Virunum, (20), Poetovio (20),Mursa (20/21), . Nicomedia (52),Nicaea (52),Ancyra (86),Archelais (or) Iuvavum (19),Ovilava (12),Vindobona (13), (63),Tyana (66),Tarsus (66), Issus (67),Antiochia Scarbantia (20), Savaria (20),Mestrianae (20), Sopia (67). nae (20),Mursa (20/21),Cibalae (20/21),Sirmium 2. Antiochia (67), Issus (67),Tarsus (66),Tyana (66), (21),Viminacium (21),Naissus (21), Serdica (21), Archelais (63),Ancyra (86),Nicaea (52),Nicomedia Philippopolis(22), Hadrianopolis (51),Heraclea ($2), (52),Constantinopolis (52). Constantinopolis(52), Nicomedia (52),Nicaea (52), 3. Nicomedia (52),Constantinopolis (52), Heraclea ( $2), Ancyra (86),Archelais (63),Tyana (66),Tarsus (66), Hadrianopolis (51),Philippopolis (22),Serdica (21), Issus (67),Antiochia (67), Laodicea (68),Tripolis Naissus (21),Viminacium (21), Sirmium (21), Cibalae (68), Berytus (69), Caesarea (69), Pelusium (70), (20/21), Mursa (20/21), Poetovio (20), Virunum Andronpolis (74).Alexandria (74). (20), Iuvavum (19) (or)Mursa (20/21),Sopianae 5. Mediolanum (19/39),Cremona (39),Bedriacum (39), (20), Mestrianae (20), Savaria (20), Scarbantia (20), Verona (19/39),Iulia. Concordia (19/40), (or)Roma Vindobona (13),Ovilava (12), Iuvavum (19),Pons (44), Fanum Fortun?? (42),Ariminum (40/42), Aeni (19),Bratananium (19),Cambodunum (19), Ravenna (40), Spina (40),Hatria (40), IuliaCon Brigantium (19),Vindonissa (18),C?mbete (18), cordia (19/40),Aquileia (19),Emona (20), Siscia Argentovaria (18/11),Argentorate (11),Divodurum (20) ,Cibalae (20/21),Sirmium (21). Trever orum (11),Augusta (11). 4. Augusta Treverorum ( ),Divodurum (11),Argen torate (11),Argentovaria (18/11),C?mbete (18),

DOP 62