Ancient History & Civilisation


Sole Augustus

The Overthrow of Magnentius

Having restored liberty to the city of Rome for a second time during the summer of 350, Magnentius set about consolidating his authority in the western empire. The chronology of his conquests is uncertain although it is frequently said that Magnentius extended his range quickly and efficiently.1 Beyond the coinage produced by mints under Magnentius’ control, the seemingly unglamorous evidence of milestones indicates clearly the extent of Magnentius’ penetration into the former provinces of Constans with inscriptions honouring both the usurper and his Caesar, Decentius, having been found in the Iberian Peninsula, Gaul (including the north-eastern provinces of Germania and Belgica), Sardinia, Africa, Italy and the city of Rome.2

The one exception to this is the island of Britain for which no epigraphic evidence referring to Magnentius and/or Decentius is extant.3 However, it is apparent that Magnentius took the diocese of Britain out of Constantinian hands for a considerable period of time and did so in such an effective way that a special commission was launched following Magnentius’ defeat in order to recover the island and to punish the conspiring officer class who had enabled the usurper’s rule.4 An account of Constantius’ political expansion into Britain, the island that had ensured the fame of his brother as a dashing, young warrior following his voyage to the island in the winter months of 343 where he shocked the inhabitants with an unexpected visit,5 is given by Ammianus. It is here that the historian’s audience is introduced to the villainous figure of Paul nicknamed “the Chain” on the account of his ability to ensnare his targets in complicated plots.6 A notarius regarded as Constantius’ “special investigator”,7 Paul was despatched to Britain to mop up resistance to Constantius II and to bring to trial the supporters of Magnentius in the winter of 353/354.8 Ammianus’ portrayal focuses on Paul’s absolute ruthlessness in carrying out his task, which in the case of his presence in Britain led him to go beyond his original brief and torture and imprison many who were innocent of complicity with the Magnentian regime. For Ammianus, the voice of conscience in this affair was Martinus, the Vicar of Britain who was (to quote the historian) “governing those provinces as substitute for the prefects”9 in the interim period prior to the appointment of a new Praetorian Prefect of Gaul.10 Martinus sought to defend the innocent against Paul’s charges. Tensions must have been running very high since Martinus himself, after being charged by Paul with treason, turned on Paul and attempted to assassinate him: His attempt failed and Martinus, “a most just ruler who had dared to lighten the unhappy lot of many”, committed suicide.11

The impact of Magnentius’ reign across the empire was considerable and there should be little doubt that it took a massive military effort on Constantius’ part to reclaim Constans’ territory. Details of the higher-profile campaigns in Italy and Illyricum are known (higher-profile since they precipitated the defeat of the usurper), having been conveyed by a selection of sources (see below), but we know precious little about the expeditions that led to the collapse of Magnentius’ hold over other areas in the west. Julian’s praise of Constantius’ naval endeavours in the Mediterranean following his defeat of Magnentius at the battle of Mursa in September 351, specifically Carthage in Byzacium, and the Iberian Peninsula, reveals the prominence of such episodes in memories of the civil war but which have curiously left near-zero trace in the historiography.12 Such episodes were key not simply in hastening the end of Magnentius’ hold over Africa, Iberia and southern Gaul, but were instrumental in shrinking his hold over the Mediterranean. Civil conflict, however, is never simply a military affair: Foremost, it is about winning and maintaining allegiances and the politics of partisanship. The milestones of the period reflect this very neatly. Ignazio Didu’s survey of the epigraphy demonstrates the significant regional variations in the presentation of Magnentius’ regime in the west. As Didu indicates, the epigraphic evidence from Iberia (predominantly Galicia) pays little concession to the regime of Constantius II,13 in contrast to a number of inscriptions from Africa (the Mauretanian provinces and Numidia) which are dedicated to both Constantius and Magnentius.14 Didu views these differences as a sign of the firm control of Magnentius’ consolidation of Spain, in contrast to his hold over Africa which required the exercising of political opportunism by way of his alignment with the eastern emperor. Questions relating to dating are clearly relevant here (the African inscriptions are likely from the first year of Magnentius’ reign) but also the identity of the dedicator(s) and the question of whether or not to interpret milestones as messages from the emperor or dedications to him from an individual (a governor), a community or a military unit, a distinction which commentators believed they were able to discern by the use of either the nominative or dative case of address in the inscription, a judgement that Eberhard Sauer has lately problematised in light of regional epigraphic trends in the later empire.15 (“The wide variation in style demonstrates that the texts were not normally prescribed or even sanctioned by central authorities, let alone the emperor himself”.16) Sauer argues valuably concerning the important role played by milestones in advertising the political allegiances of individuals and communities, particularly in light of their principal audiences, namely civil and military officials and the army in transit. Thus, in the case of the inscription (CIL 8.22552) from close to Icosium (modern-day Algiers), the dedicator(s) had understood that the political climate required a show of support for both emperors, an interpretation which may not have been ultimately correct but at the time was a reflection of the prevailing political mood. This milestone (and another for Magnentius from Galicia17) bears the distinctive formula, “Born for the good of the State” (BN RP NATIS) which is found on coins and inscriptions for Constantine and a number of other emperors from the fourth century18 (intriguingly, it is found on a number of milestones and honorific inscriptions from Britain19).

Matters of allegiance become greatly heightened during times of civil conflict. Some awareness of this fact offers a way of reading the treatment of the guilty and innocent alike by Paul during his investigation of Magnentius’ supporters in Britain during a period marked by danger and anxiety following the death of their emperor some four months previously. Ammianus’ tendency to exaggerate the suspicion, intrigue and bitterness of Constantius and his functionaries should not prevent us from seeing his portrayal of Paul in Britain as a genuine representation of the time and its events. In this regard, Ammianus’ evaluation that Paul’s recourse to torture exceeded his original instructions should be questioned since in the case of conspiracies against the emperor it would have been a feature of investigations to apply judicial violence even against those in the higher social classes.20 In this regard, the fate of those convicted – freeborn citizens (ingenui) clapped in irons – is convincingly narrated:

After perpetrating these atrocious crimes, Paul, stained with blood, returned to the emperor’s camp, bringing with him many men almost covered with chains and in a state of pitiful filth and wretchedness. On their arrival, the racks were made ready and the executioner prepared his hooks and other instruments of torture. Many of the prisoners proscribed, others driven into exile; to some the sword dealt the penalty of death. For no one easily recalls the acquittal of anyone in the time of Constantius when an accusation against him had even been whispered.21

Such commissions, “ruthless purges” or “treason trials” (call them what you will) were as essential in securing an opponent’s defeat in civil war as an actual military victory. They had the potential to transform entirely the post-conflict landscape by ensuring the complete rewiring of a region or community’s political allegiances. In the case of Magnentius and Britain, it has long been suspected that the impact of Paul’s investigation and trial of the usurper’s supporters was profound. In the case of the landed Roman estates on the island, changes of use or even the demolition of villas have been identified in the period stretching from the mid-350s to 360, with the known political proscribing of Constantius’ opponents (a sentence that would have included the confiscation of the convicted individual’s property to the imperial fisc) regarded as the likely cause. As Graham Webster has noted in relation to the changed configuration of villas in the fourth century:

[i]t would be reasonable to assume that the considerable changes that took place in Britain were a result of these large-scale confiscations. Some estates may have received new owners … or they could have been divided and redistributed; others may have been enlarged with the addition of adjacent lands or groups of estates under a single owner.22

The villa at what is now Barnsley Park near Cirencester underwent a radical transformation c. 360 from small farm to a stone corridor villa with a large barn (coins of Constantius II have been discovered in the stone building’s foundations). By contrast, Gadebridge villa in Hertfordshire which housed a large pool that may have served a public (religious?) function was completely demolished c. 360.23 Such radical changes in so narrow a corridor of time may very well be explainable in light of the social upheaval of regime change brought by Paul to Britain.24

Constantius’ path to victory over Magnentius began some twenty-six or so months previously when the armies of the two emperors met at Mursa on 28 September 351.25 In the months prior to this fateful engagement, Magnentius had pushed forward into southern Pannonia following the failed attempt by Constantius to break through to Italy and in so doing had scored some notable successes, including the short-lived capture of Siscia. The capture of this important city is evident by an output of bronzes from the city’s mint proclaiming VICTORIA AVG ET CAES in the names of both Decentius and Magnentius, with the accompanying, provocative reverse image of the emperor standing on a prisoner in a scene of total subjugation. The victorious emperor holds aloft a standard emblazoned with the Chi-Rho.26 While the coin likely commemorates an early success by Decentius against the barbarians in Gaul, it is tempting to speculate that this was an audacious commemoration of Magnentius’ success against Constantius half-way through the year. However, Siscia was the limit of the usurper’s gains over Constantius. When the attempt to push eastwards to take Sirmium failed, Magnentius’ troops moved north-west to Mursa. According to Zosimus (2.49.3–50.1), the inhabitants of Mursa closed the gates to the usurper, prompting Magnentius’ forces to launch a full-blown assault on the city’s fortifications. Constantius’ defence of the city precipitated the engagement on the plain outside Mursa. Zosimus’ account (2.51.3) of the battle portrays a scene where the will to slaughter on both sides was so absolute that,

when it was late at night, they were still wounding each other with spears, swords, or anything else within reach. Neither darkness nor anything else which usually brings a respite from battle could make the armies cease their mutual slaughter; rather they considered it the greatest good fortune to be wiped out together.27

In a similarly detailed account of the battle,28 Zonaras (13.8) notes that Constantius stood over the battlefield at dawn and shed tears for those killed, undoubtedly an example of poetic licence, although Julian’s account of the battle (Or. 1. 35d–38a) places Constantius in the thick of the action.29 Indeed, the whereabouts of the emperor were a matter of discussion in the ancient sources: In contrast to Julian, Sulpicius Severus in his Chronicle (2.38) places Constantius in a martyrium outside Mursa where he took refuge as a result of his cowardice (see below). Regarding the final outcome of the battle, Zonaras narrates the following numbers: “For it is said that from [Constantius’] men, who numbered altogether 80,000, about 30,000 fell, and from those of Magnentius, who were 36,000, 24,000 were destroyed”. The numbers may have been intended to convey the scale rather than actual number of soldiers killed, but if anywhere near correct, the battle of Mursa was one of the most bloody civil war battles in the history of the Roman period. The immediate impact on Magnentius’ fortunes was a retreat behind the fortress on the Julian Alps (claustra Alpium Iuliarum) which he strengthened and a further retreat to Aquileia.30

As discussed in Chapter 3, the battle of Mursa was central in shaping hostile judgements about Constantius. This encounter together with the broader “culture” of civil war in the 350s was regarded by a number of ancient witnesses as providing a plausible explanation for Rome’s severely weakened military response to the threat from Sasanian Iran. However, in the context of that decade, the victory of Constantius was a highly celebrated event that was instrumental in re-establishing the emperor’s legitimacy following years of major social and political upheaval. As Mark Humphries has argued, it became a central feature of Constantius’ propaganda in the period, at a time when he was consolidating his hold over the entire empire as sole Augustus.31 The episode shaped the monumental landscape of the empire in the guise of the triumphal arches in Gaul and Pannonia – commemorating Constantius’ victories at Mons Seleucus (see below) and Mursa, respectively – which so greatly pained Ammianus Marcellinus, and was also marked by the erection of the great obelisk of Thutmose III, pharaoh of the Eighteenth Dynasty, on the spina of the Circus Maximus following (or during?) Constantius’ visit to Rome between April and May 357 (see Chapter 8). It was also, as Humphries has demonstrated, a central episode for the writers of panegyrics delivered for Constantius, including Julian’s two important orations and the lost epos of Proba, the wife of the Urban Prefect Clodius Celsinus Adelphius (also Chapter 3). In line with Bleckmann’s proposal from his article in 1999, it is also likely that a panegyrical template – perhaps more than one lost speech – underlay the intermittingly positive portrayals of Constantius and his troops which can be glimpsed in the Greek historiographical accounts of Zosimus and Zonaras.32 With reference to the former’s incredible but muddled version of events,33 this template is not difficult to spot. It is apparent in the claim that Constantius sought to avenge the death of his brother, Constans, in the public speech delivered in response to the criticisms of Magnentius’ apparatchik Titianus (2.49), which appears to have been central to the panegyrical and monumental commemoration of the war as may be seen in the plausible reconstruction of the attic inscription of Rome’s Arcus Divi Constantini by Mateos, Pizzo and Ventura (see Chapter 5).34 It can also be seen in the distinction made by Zosimus between the “courage and reputation of the Romans” on Constantius’ side, and the blind fury of Magnentius’ supporters (2.51.2–3) during the battle itself, a judgement which carries the subtle distinction of separating “the Romans” of Constantius from the troops of Magnentius who by implication are not (or no longer) deemed to be Romans. This distinction is observable in the period immediately following Constantius’ reign, and is preserved by Jerome in his Chronicle where the usurper’s supporters in Rome are identified as the “Magnentians”.35

Evidence of a Christian panegyric may also be glimpsed in the brief notice about the battle of Mursa in the Chronicle of Sulpicius Severus, who was born in the year that the emperor Julian died. The passage in question (2.38) is replete with the familiar criticisms of Constantius, a gullible, cowardly emperor in league with the Arians. However, certain aspects of Sulpicius’ account betray a different orientation not least the claim that an angelic apparition heralded the emperor’s victory. The centrality of Valens, the bishop of Mursa, in the account is an intriguing feature in light of the bishop’s public break at the Council of Milan in 345 with the position of the eastern court regarding Athanasius’ status as persona non grata.36 It may be that Constantius’ defence of the city marked a reconciliation between the bishop and the emperor which was commemorated in an oration.

For at that time, when a battle was fought at Mursa against Magnentius, Constantius had not the courage to go down to witness for himself the conflict, but took up his abode in a church of the martyrs which stood outside the town, Valens who was then the bishop of the place being with him to keep up his courage. But Valens had cunningly arranged, through means of his agents, that he should be the first to be made acquainted with the result of the battle. He did this either to gain the favour of the king, if he should be the first to convey to him good news, or with a view to saving his own life, since he would obtain time for flight, should the issue prove unfortunate. Accordingly, the few persons who were with the king being in a state of alarm, and the emperor himself being a prey to anxiety, Valens was the first to announce to them the flight of the enemy. When Constantius requested that the person who had brought the news should be introduced to his presence, Valens, to increase the reverence felt for himself, said that an angel was the messenger who had come to him. The emperor, who was easy of belief, was accustomed afterwards openly to declare that he had won the victory through the merits of Valens, and not by the valour of his army.37

Socrates – also no supporter of Constantius – preserves another overtly Christian instance of praise for the Constantinian emperor which is clearly part of the same tradition. In the following passage, Constantius’ divinely ordained destiny as a victor in the civil war is the predominant theme arising from the battle of Mursa. Notably, it is divine favour which transforms the loyalty of the enemy troops:

… but at last, Magnentius having been defeated near Mursa, a fortress of Gaul, was there closely besieged. In this place the following remarkable incident occurred. Magnentius desiring to reassure the courage of his soldiers who were disheartened by their late overthrow, ascended a lofty tribunal for this purpose. They, wishing to give utterance to the usual acclamation with which they greet emperors, contrary to their intention, simultaneously all shouted the name, not of Magnentius, but of Constantius Augustus. Regarding this as an omen unfavourable to himself, Magnentius immediately retreated to the remotest parts of Gaul.38

Leaving authorial agendas to one side for the moment, it is clear, as Humphries outlines, that the battle of Mursa was the major engagement of the civil war against Magnentius. For some ancient authors, notably Zosimus, the remaining two years of Magnentius’ reign were incidental to the battle and were hardly worth mentioning (2.53). However, a series of further, difficult engagements lay ahead for both rulers, and while the chronology for the post-Mursa stage of the conflict is uncertain, numerous significant events took place which were instrumental in reshaping the configuration of power in the empire. In the interim period, Constantius set about consolidating his fleet and issued an amnesty (Gk. adeia, Julian, Or. 1.38d) for Magnentius’ supporters with the exception of those who “shared the guilt of those infamous murders”,39 a reference to those responsible for planning and carrying out the assassination of Constans and other members of his entourage.

In the months following the battle, Constantius’ fleet sailed up the Adriatic and blockaded the mouth of the river Po to prevent Magnentius’ escaping by sea. His army also captured the fortress of Ad Pirum along the line of the claustra during mid-summer 352.40 In the meantime, Magnentius had pressed inland to Aquileia, the city that had since February 350 been his effective headquarters. Julian alleges that since Magnentius felt so secure in the city as a result of the protection afforded by the Alps, he spent his time attending public shows (Or. 1.38d). With the loss of Magnentius’ defensive line, Constantius broke into Italy in early autumn 352. Aquileia fell following Magnentius’ desertion having been given notice of Constantius’ advance while he was attending a horse-race in the city.41 It was at this point that Magnentius decided to move his forces into Gaul and surrender his hold over Italy. On 26 September 352, Neratius Cerealis – the maternal uncle of the Caesar Gallus – was installed as the Urban Prefect of Rome, thereby marking the return of the eternal city to a Constantinian loyalist.42 The new Prefect duly dedicated an equestrian statue in the Roman Forum to the emergent victor in the civil war inscribed with the following pregnant terminology: “To the restorer of the city of Rome and of the world, and the exterminator of the pestilential tyranny”.43 It is perhaps at this point in time with the loss of Italy and his retreat to Gaul that we should situate the claim made by Zonaras (13.8) that Magnentius made overtures of peace and reconciliation to Constantius by deploying a wide range of representatives, including bishops and members of the Senate who had remained loyal to the western emperor. Constantius dismissed all such attempts, secure in the knowledge that he had effectively led his rival into a corner. By this time, Constantius was in Milan where, on 3 November 352, he issued together with Gallus a constitution “to the provincials and the people” which invalidated all the laws of Magnentius (tyrannus) and his judges – on the basis that they were contrary to law – and ordered the return of property confiscated to those evicted during his reign, although all emancipations, manumissions, pacts and transactions (presumably because they were not deemed contrary to the law) were to remain in place.44

In the summer of 353 – possibly late July45 – the two field armies met once again, this time at Mons Seleucus (La Bâtie-Montsaléon) in south-eastern Gaul. This battle is not commemorated on the same scale as Mursa even though it is credited with forcing the final capitulation of Magnentius.46 From there, Magnentius retreated further inland to Lyons, where he had spent the majority of his time since the loss of Aquileia.47 Zonaras reports that some of Magnentius’ troops devised a plot to betray him to Constantius.48 When he became aware of their subterfuge, he committed suicide. Zonaras’ account at this point mentions that before he killed himself, Magnentius murdered those around him, his kinsmen and friends, along with an attempt on the life of his brother, Desiderius, who nevertheless survived and defected to Constantius. It is suspected by DiMaio and the editors of the PLRE49 that the sibling named Desiderius was an invention by Zonaras, introduced in order to account for the discrepancy in his source, namely Socrates (2.32), where two brothers are mentioned although both are evidently references to Decentius. The Consularia Constantinopolitana records Magnentius’ death as occurring in Lyons on 10 August 353.

In fact, Magnentius’ Caesar was elsewhere in Gaul during the final stages of his Augustus’ downfall. Decentius is reported in Sens (Senonae)50 where on 18 August he also committed suicide.51 Decentius’ final weeks were marked by serious civil unrest in the city of Trier where the residents barred entry to Decentius and his forces around late June/early July 353.52 The event itself is known only from a few lines in Ammianus Marcellinus (15.6.4) who records that a certain Poemenius was chosen by the citizens of Trier to defend the city against Decentius: The reason for Ammianus’ interest in the episode arose from the investigation by the notarius Paul into the affair of Silvanus, which we discuss later, and the execution of Poemenius as a supporter of Silvanus. It is certainly appealing to see the events in Trier as “the usurping of a usurper” – to quote the contribution of Walter C. Holt53 – but the evidence would seem to indicate that this was yet another loyalist rebellion spearheaded by Constantinian supporters rather than an independent attempt to seize imperium. Decentius and his army had returned to Trier following in all likelihood an expedition prompted by Rhine Germani raids, which had reached a peak during 353. Indeed, Drinkwater sees a link between Alamanni or Frankish incursions and Trier’s rebellion – the inference being that Decentius had failed to protect the city which out of necessity had turned to its own strategy for protection.54 Whatever the cause, the incident left its mark also on the numismatic record. Two coin types, a very rare gold issue and a bronze issue, were produced with the same obverse legend (D N CONSTANTIVS P F AVG) during the period of Trier’s rebellion. The solidus issue with reverse type depicting winged Victory holding a palm and spear leading the emperor who is holding a globe and spear and bearing the legend VICTORIA AVG NOSTRI has no precedent for Constantius at any other mint.55 The debate must remain open about whether this Victory issue marked a victory won – for example, Constantius’ success at Mons Seleucus – or a victory anticipated over the usurper, although in line with Holt’s argument, the timing of the emission with the final victory over Magnentius is very tight indeed and given the distance needed to convey the news of Magnentius’ loss, it is unlikely the solidus marked a historic moment.56 The bronze issue also for Constantius imitated – albeit on a reduced scale from the large size of the original Magnentian issue57 – a recent emission for Magnentius and Decentius, although the reverse legend was altered to reflect the acknowledgement of Constantius as sole Augustus (SALVS AVG NOSTRI, “The health of our Augustus”, as opposed to the Magnentius-Decentius issue, SALVS D NN AVG ET CAES) while retaining the same reverse type unique to Magnentius of a Chi-Rho flanked by the Alpha and the Omega.58 This bronze emission was discontinued following Constantius’ consolidation of his position after the defeat of Magnentius in line with the discontinuation of all types associated with the usurper in the period following Constantius’ victory. Indeed, Constantius demonetisation of Magnentius’ coinage appears to underlie the constitution (Cod. Theod. 9.23.1) addressed to Vulcacius Rufinus, Praetorian Prefect of Gaul, on 8 March 354, which ordered the seizure of coinage which had been prohibited from public circulation.59

Ammianus’ brief note about Poemenius’ fate in the affair of Silvanus during 355 is a sure indication that Decentius failed to retake Trier, since Poemenius would likely have been executed as a leader of the rebellion should Trier had fallen. Furthermore, the fact that Decentius’ final moments were played out many miles away in Sens where he committed suicide on 18 August is also another sign that Trier remained in the hands of Poemenius and its citizens until Constantius’ troops finally reclaimed it. The emperor himself was in Lyons by September where on the 6th of that month, he issued a ruling addressed to Neratius Cerealis in Rome (Cod. Theod. 9.38.2). As Cuneo points out, the constitution was heavily excerpted by the Theodosian-era editors who gave it the title, “On the forgiveness of crimes”, and the impression is of an amnesty for all except those guilty of the five crimes deemed capital offences (murder, poisoning, magic, adultery and rape).60 However, despite its partial preservation, the ruling was intended squarely for the attention of the Urban Prefect of Rome alone and should therefore be understood as part of Constantius’ attempt to resolve the severely partisan climate in the city which we have glimpsed from time to time, both within the senatorial ranks and among the populace at large. Indeed, Julian supplies a clear indication of the background for the amnesty when he writes in Or. 2 (96a) that Constantius, following his final victory, had disavowed the use of the sword even against those who had been Magnentius’ very close friends (Gk. oikeiotera), including those who “had stooped to win a tale-bearer’s fee by slandering the emperor”.61 Maraval is surely correct to identify the villain in question as Magnentius’ former Urban Prefect, Fabius Titianus, the Constantinian loyalist turned ally of the usurper. Zosimus (2.49.1–2) records Titianus’ visit to Constantius, while the latter’s forces were encamped at Cibalae, the site of Constantine I’s decisive battle against Licinius in 324. The former Praetorian Prefect in Gaul under Constans read out a collection of charges against Constantine and his sons, including the allegation that Constantius had ruined the cities as a result of his carelessness. As Mark Humphries notes, Titianus had been instrumental “in promoting the usurper’s legitimacy in Rome”,62 and had dedicated during his time as Urban Prefect a number of statues of Magnentius in the city which proclaimed the usurper to be the “expander of the world and the Roman state”.63 Following Cibalae, Titianus was allowed to return to Magnentius. In the period following Magnentius’ defeat, it appears that he was not so fortunate. Despite Julian’s claim that no punishment followed Titianus’ act of betrayal, the partial erasure of his name on one (CIL 6 1166a/LSA 1281) of the bases from Rome suggests otherwise.64

As a sign of imperial magnanimity and clemency, the amnesty issued by Constantius in Lyons likely marked the beginning of his tricennalia which got underway in Arles in November 353.65 Appropriately for a critic of the emperor, Ammianus emphasises the opposite situation to have been the case. The commemoration of the emperor’s longevity heralded an intensive persecution of Magnentius’ supporters, including those simply rumoured to have favoured the usurper.66 This is the period which included Paul’s attendance in Britain to investigate and punish Magnentius’ supporters there. Hartmut Leppin has rightly drawn attention to the disparity of the sources in the period following Magnentius’ defeat, but he diminishes the value of his observations with the judgement that Ammianus is not a partisan commentator, in contrast to the panegyrists and church historians, because he “writes at temporal and personal distance”.67 One solution, as Leppin proposes, is to evaluate Ammianus’ account as focused on those who were not permitted to take advantage of the amnesty, as in the case of Gerontius,68 a count in Magnentius’ retinue. The historian tells us that Gerontius was tortured and exiled (Leppin undermines his idea about Ammianus’ personal distance with the claim about Gerontius that “the historian was particularly sensitive in [Gerontius’] case”69). However and in spite of its excerpted form, the constitution issued in Lyons indicates that the amnesty was intended to be selective in two ways: First, it was to be applied locally, in the context of Rome and its environs in order to remedy any outstanding issues of loyalty; and second, it absolved all except those guilty of the five capital crimes. Other known victims of the trials which followed Constantius’ victory include those condemned by Paul in Britain, including Martinus who should be counted among the victims of the political violence. Very little is known about the prosecution of Magnentius’ inner circle of advisers and supporters. As discussed in the previous chapter, his wife, Justina, survived the turmoil and possibly also her father, Justus,70 whose appointment as the governor of Picenum could in fact have occurred under Constantius II rather than under Magnentius as is commonly supposed. Other victims are known by name tangentially. For instance, Ammianus reports that Gratianus, the father of Valentinian I and Valens, had been the subject of a confiscation order for having given shelter to Magnentius during his passage through Cibalae during the summer campaign season of 351 that culminated in the battle of Mursa. However, the confiscation order can only belong to the period of investigations following the usurper’s defeat. Ammianus’ notice indicates that Gratianus had served as one of the protectores domestici, as Tribune, and then as a military count (comes rei castrensis per Africam) in Africa,71 all during the reign of Constantine I in the 320s and 330s (since he was born in Cibalae, his initial appointment may have been under Licinius).72 During his time in Africa, he was accused of theft, although the allegation does not seem to have impeded his career greatly since under Constans, Gratianus was also a military count in Britain during the winter months of 342–343. It was during his retirement that he gave support to Magnentius, an act which led to the confiscation of his countryside residence. However, his reputation was rehabilitated during the reigns of his sons when he was honoured with a statue in Cirta in Numidia,73 and by a bronze statue in Constantinople.74

The Glittering Career of Gallus Caesar

According to Ammianus’ narrative of events, the period following the defeat of Magnentius was a time of unparalleled barbarism in the empire. In the conclusion (14.7.21) to his famous description of Gallus Caesar’s “atrocities and savagery”, the author described civil life in the eastern empire as a place where the innocent were condemned alongside the guilty in the absence of a fair trial, and where “all justice vanished from the courts as though driven out”. Ammianus is our main source for the post-Magnentian period and as such it is common to follow the historian’s line in reducing the tumult in the eastern provinces to the question of Gallus’ moral temperament. However, looking beyond the historiographical interpretation of a partisan author, the period beginning in autumn 353 is more appropriately understood in contextual terms as the aftermath of a brutal civil war where the conflict’s end witnessed a heightening of the underlying social tensions that had brought about the war in the first place, and which subsequently shaped the political culture of the empire for many more years to come. This was predictable in light of the fact that civil war tends to be one of the most intense forms of human conflict – a truism applicable both historically and contemporaneously75 – as a result of the emotive ties that bind combatants. The months following the cessation of hostilities in September 353 were characterised by a persistence of emotive markers such as suspicion, intrigue and ambition, the very things that had driven events to spill out into open conflict in the first instance. A justifiable objection to this interpretation views the “atmosphere” of intrigue and suspicion in book fourteen as the invention of Ammianus whose complex narrative portrayal of Gallus’ reign – recently unpicked by Alan Ross – serves as a foil not only for the author’s classicising judgement about Constantius II but is also preparatory to the introduction of Julian in book fifteen (15.8), the putative hero of this portion of his work.76 Underlying this is the rejection of the belief that Ammianus’ work captures the political conditions of the early 350s in something equivalent to a historical petri dish, itself part of an established trend which has evaluated Ammianus as if he had been “cool and unbiased in his judgments”,77 which is now regarded as a misplaced judgement that fails to recognise his work as a narrativised series of literary reminiscences composed during the reign of Theodosius I. However, in light of the fact that the events conveyed in Ammianus’ narrative are attested in sources that are both earlier (notably in Julian, Libanius et al which Ammianus drew on) and later (e.g. Philostorgius), his history suggests that some understanding of the political culture of the time as exceedingly tindery can be justifiably retained in the mind of the commentator. It is possible, therefore, to look beyond Ammianus’ portrayal of Gallus as a cipher for the suspicion and scheming of the time and see his reporting of the political turmoil afflicting the empire as a sign of an elite, and indeed of a society at large, deeply affected by a protracted civil war.

In the first instance, the problem of usurpation did not go away with the death of Magnentius and the suppression of his troops and political supporters. Embedded in the affair of Domitianus,78 the Praetorian Prefect of the East appointed following the death of Thalassius, and Montius,79 the Imperial Quaestor (quaestor sacri palatii), who were both lynched by Gallus’ soldiers in Antioch c. April 354,80 is the curious episode involving the production at Tyre of an imperial robe which had been commissioned in secret. While Ammianus states that “it was uncertain who had ordered it or for whose use it was made” (14.7.20), the implication of such a step was clear: Someone was planning to “don the purple” and assert a claim to rule. Tyre lay (ostensibly) under Gallus’ jurisdiction and as such the order came, presumably, from Antioch to arrest Apollinaris, the governor of Phoenicia. While the legal evidence restricting the production and possession of purple-dyed cloaks or tunics, those “dedicated only the Emperor and to his Household”, derives from a later period,81 the arrest of the governor of the province where the imperial dyeworks was based indicates that the activities of the works at Tyre were closely policed by the imperial administration and any breach in protocol by workers or those involved in the oversight of imperial textile production was deemed a treasonable matter.

At the same time as events in Phoenicia were unfolding, Domitianus and Montius were being punished and tortured by Gallus’ soldiers ostensibly because the new Prefect had undermined court protocol by demanding in an insolent manner Gallus’ return to Italy under instruction from Constantius himself. Gallus, angered by Domitianus’ abruptness, ordered the palace guards (scholae palatinae) to arrest the Praetorian Prefect. Montius became implicated following his attempt to calm the atmosphere at court, although the argument he put to the palace troops that arresting the Praetorian Prefect was tantamount to overturning statues of Constantius – i.e. rejecting the authority of the Augustus – was the catalyst for his physical assault (14.7.9–18). Both were subjected to brutal treatment overseen by a certain Luscus (the Curator Urbis of Antioch) and their bodies were thrown into the Orontes (Ammianus, 14.7.16). A significant detail according to Philostorgius’ narrative, preserved in the Passion of Saint Artemius (12–13 = 28a), indicates that the brutalised corpses were recovered by “the bishop of the city”, namely Leontius, and accorded a burial, which, in turn, reveals two concerns: First, both men were Christians, and second, Leontius, although allied with Gallus as the leading lights in the homoian Christian scene (see below), was prepared to enforce Christian tradition against imperial wishes.82 Montius, the Imperial Quaestor in Gallus’ retinue – the principal legal adviser and draughtsman of imperial constitutions83 – was, on the basis of Ammianus’ description (14.7.18), evidently tortured since he disclosed the names of two individuals, Epigonus and Eusebius, who were later implicated as co-conspirators and tried soon afterwards. Ammianus’ narrative is characterised by a number of temporal connectives (eisdem diebus; dum) and while the events themselves are not linked explicitly to one another, it is apparent that the brutality in Antioch was associated in some way with the events in Tyre. This becomes apparent in the episode (14.7.9) that occurred “in the same days” as the murders of Domitianus and Montius, Ammianus reveals that the son-in-law of Domitianus, who was also the son of the governor of Phoenicia – also named Apollinaris – was arrested for enquiring about Gallus’ ambitions among the soldiers of Mesopotamia, where he had recently been posted following a period as “the one in charge of the palace” (cura palatii). The younger Apollinaris had tried to evade capture after hearing about the fate of his father-in-law in Antioch and headed for Constantinople via Armenia minor – circumventing the settlements around the Black Sea – but was captured and later tried along with his father.

The various incidents recounted by Ammianus in chapter seven of book fourteen lead to a very complex plot whereby someone was viewed as seeking to usurp Gallus’ role as Caesar in Antioch. The most likely candidate appears to be Domitianus. His direct nay discourteous handling of Gallus, while the basis for his seizure by Gallus’ soldiers in the accounts of both Ammianus and Philostorgius (3.28), is clearly not the whole story; rather, his arrival in Antioch with orders to recall Gallus to the west was interpreted seemingly by Gallus himself as an attempt by Constantius to demote him from his caesarean rank and install Domitianus as his replacement.84 Gallus detected a high-level coup engineered from within Antioch itself involving the Praetorian Prefect of the east, the Imperial Quaestor, the Governor of neighbouring Phoenicia and his most recent “caretaker” of the Antiochene imperial palace. This is informed speculation, but the main point to bring to mind is the susceptibility of the imperial court in the east to the fear of usurpation in the immediate aftermath of Magnentius’ defeat.

In this regard, the response of Gallus and his troops was in keeping with the prevailing political culture of suspicion and intrigue of the time. His response was also in line with the precedent set by Constantius with regard to the mopping up of pockets of factionalism in support of Magnentius following the capitulation of the usurper’s regime. Therefore, at some point soon after the murders of Domitianus and Montius, likely during mid-summer 354, Gallus recalled Ursicinus, the magister equitum, from Nisibis (where he was spearheading the defence of the city) to Antioch in order to preside over investigations into the treasonable activities of the Antiochene administration. Acting, in the description of Ammianus, “seemingly as a judge” (14.9.3), the author’s former commanding officer ruled on the fates of those accused of treason by Gallus’ government. The names supplied by the dying Montius, Epigonus and Eusebius, led to the arrest of two men whose only crime was sharing the same names as the men suspected of offering weapons in the expectation of a rebellion. Ammianus, presumably as a result of his personal connection to Ursicinus, knew that the Epigonus and Eusebius of Montius’ mind were in fact tribunes of an imperial arms factory – most likely the one situated in Antioch itself85 – and that the two men arrested were nominal doppelgängers (this begs the question, when the information that the wrong men had been arrested actually came to light?). Both were tortured: Epigonus confessed, while Eusebius remained resolute in his ignorance of any involvement in the events (14.9.5–6).

Torture was also used to extract a confession from the dye workers of Tyre to the effect that they had produced a purple sleeveless tunic. A Christian deacon named Maras was summoned during proceedings, and the evidence of a letter written by him to the chief (Praepositus) of the Tyre factory requesting completion of a garment was produced. Ammianus makes no connection between Maras’ commission and the classified royal robe, which was possibly intentional since it is apparent that Ammianus wished to convey the sense that the trial’s investigations were suspect. Ammianus notes that many were put to death, including the Apollinares who were both initially sentenced to exile but were summarily executed once they had reached their family villa named Craterae, some twenty-four miles from Antioch. Ammianus is highly critical about the proceedings in general and it is apparent that his main concern lay in protecting the reputation of Ursicinus, his commander, from claims that he had been complicit in Gallus’ activities. Thus, Ammianus states that Ursicinus had initially written to Constantius in secret prior to the beginning of proceedings in order to alert the emperor to the corrupt practices that were evident to him. However, it seems that there was little that Ursicinus could do to prevent the injustices which emerged during the trials since Gallus and Constantina were guiding events from close range.

Accordingly, on the day set for the fatal examinations the Master of the Horse took his seat, ostensibly as a judge, attended by others who had been told in advance what was to be done; and here and there shorthand writers were stationed who reported every question and every answer post haste to Caesar; and by his cruel orders, instigated by the queen, who from time to time poked her face through a curtain, many were done to death without being allowed to clear themselves of the charges or to make any defence.86

The events which led to the lynching of Domitianus and Montius were but one example of Gallus’ broader mishandling of political affairs in the eastern empire as portrayed by Ammianus. Another major flashpoint was Gallus’ alleged mismanagement of food shortages in Antioch – dated by Barnes to autumn 353 – which Ammianus presents as an instance of wilful negligence on the Caesar’s part (“he did not, after the manner of princes whose widely extended power sometimes cures local troubles, make any arrangements or command the bringing of supplies from neighbouring provinces”(14.7.5)). However, the same supply crisis is mentioned by Ammianus at the beginning of the chapter (14.7.2) where he reveals that Gallus had sought to lower prices in order to counter the inflationary price-fixing which characterised the activities of local landowners and suppliers.87 As Thompson noted, many of these individuals would have been members of the Antiochene council, and their intransigence to Gallus’ intervention was so great that the Caesar – as Ammianus relates – threatened to execute the leaders of the senate. Ammianus’ interpretation is thereby intended to portray Gallus as a disinterested spectator of events: By refusing to take responsibility for the supply crisis and accept that food shortages lay in the purview of the central administration, the blame was shifted onto government at the provincial level. As such, the consular governor of Syria, Theophilus,88 was lynched by a mob in Antioch, and the property of the prominent sophist, Eubulus,89 was attacked and burnt to the ground.

The intricate circumstances surrounding the murder of Domitianus, the Praetorian Prefect of the East, were reflected in the treatment of his counterpart in Gaul, Vulcacius Rufinus, the maternal uncle of Gallus (see Chapter 6). In the spring months of 354, Constantius had set out from Arles for Castrum Rauracense (Kaiseraugst) via Valentia (modern-day Valence) in order to combat raids on Gaul by the Alamanni led by the fraternal kings Gundomadus and Vadomarius (Ammianus, 14.10.1).90 News of Gallus’ conduct in Antioch reached Constantius while he waited for supplies in Valentia from a Protector Domestici Herculanus, the son of Hermogenes, the magister equitum killed in the riots of 342 in Constantinople following the contested election of Paul as bishop of the city (see Chapter 5). The delay in supplies caused by bad weather prompted a protest by troops assembled at nearby Cabyllona (Chalon-sur-Saône) and as Praetorian Prefect of the region, Vulcacius Rufinus was sent to pacify the soldiers. These events presented Constantius’ court with an opportunity to engineer the downfall of Rufinus, a figure who had been of central importance in the suppression of both Vetranio and Magnentius, and who had also engineered the elevation of his nephew Gallus in March 351 and his marriage to Constantina (see Chapter 5). In this regard, Ammianus’ description of Rufinus as “a very powerful” (praepotens) individual is on point (14.10.5). The immediate context for this astute political wrangling was unquestionably the communiqué regarding the situation in Antioch and the east brought by Herculanus. Following the murders of the east’s Praetorian Prefect and an imperial Quaestor, Gallus was a major concern for Constantius’ government and Rufinus was judged, rightly or wrongly, to be an extension of the Caesar’s influence in the west. Indeed, we would be warranted in reading this move against Rufinus as the first by the court in the west to eradicate Gallus by dismantling his network of support there. As we noted previously, Vulcacius Rufinus had been one of Constantius’ intimates during the decade of the 340s having served in the prestigious position of comes ordinis primi intra consistorium91 and as comes orientis under the emperor. As conveyed in the report by Ammianus, this was an opportunistic affair and while details of the plan itself are meagre, one suspects that the enemies of the Praetorian Prefect perhaps hoped that he would be set upon and killed by the soldiers. However, this was a dangerous business. Rufinus’ high standing on the political scene did not make him the sort of target whereby those seeking to engineer his downfall could escape with impunity. Indeed, Ammianus notes that the uprising was handled in a more circumspect manner by way of bribes in gold donatives distributed to the leaders of the rebellion by Eusebius, the praepositus sacrii cubiculi, the high-profile chamberlain and adviser at the court of Constantius, whose finger prints in the case of Gallus’ downfall were detected by Julian.92

Constantius returned to Milan from the Rhine having secured a peace treaty with the Alamanni, the culmination of “a rare and remarkable campaign” in the assessment of John Drinkwater.93 Despite his success against the Germani, Constantius had little respite from the affairs of state. Ammianus describes in a familiar conspiratorial way the details of a plan which proposed that Ursicinus and his sons were planning to seize control – aiming for a “higher (altiora) position” in the historian’s language describing the act of usurpation (cf. 14.11.6 and Gallus) – of the eastern half of the empire should the emperor remove Gallus from Antioch in light of the unfolding drama in the city (14.11.2–3). The rumour regarding the ambitions of Ursicinus and his sons – grown men serving in the army – derived from the Grand Chamberlain Eusebius and Flavius Arbitio (also a magister equitum) who seem to have feared the magister’s political influence: The Grand Chamberlain’s insinuations about Ursicinus present the impression of a political adviser who sought to clear all problems (real and potential) out of Constantius’ way, a feature caricatured by Ammianus who claims that the emperor was dependent on Eusebius (rather than the other way round!).94 As Ammianus portrays events, this led to a staggered recall of eastern personnel to Milan. First Ursicinus with Ammianus in attendance were recalled to court, ostensibly to discuss strategies against Persia with Prosper95 sent as his temporary replacement to the east, followed by a request addressed to Gallus that he present himself before the emperor.

It is highly significant that a request was also made for Constantina to attend court in Milan at this point in time. Ammianus indicates that this was done in order to put Gallus’ mind at ease in light of Constantius’ decision finally to move against his cousin. However, in light of Constantina’s role as a likely back channel (to borrow Drinkwater’s characterisation96) between the camps of Vetranio and Constantius during 350, it is surmised that Constantius intended his sister to play a more direct role in managing Gallus in light of the emperor’s plan to remove him from Antioch. A much less charitable interpretation of Constantius’ overtures to Constantina involves linking the emperor’s request for his sister to attend court with the aborted plan to hasten the demise of Vulcacius Rufinus: If, as Bleckmann and others have proposed (see Chapter 6), these two individuals had been instrumental in securing Gallus’ position in March 351, it would be meet to try and extinguish the Caesar’s leading supporters from the political scene prior to the extermination of Gallus. However, the indignant response of Constantius to Gallus’ attempt to pin the blame for the murders of Domitianus and Montius et al on Constantina reported by Ammianus (14.11.23) suggests that Constantius had continued to count his sister as an ally. Indeed, Gallus’ fate could have been very different if Constantina had been alive as he made his journey towards Italy. As it was, however, while Constantina was travelling to Italy, she died from fever at Caeni Gallicani in Bithynia. Her near-half-a-decade-long marriage to Gallus had seen the birth of a girl but no male heir.97 François Chausson98 endorses and develops the thesis proposed by Angelo Silvagni in 1929 that the child of Gallus and Constantina was named Anastasia, an observation based on a number of important inscriptions from St. Peter’s basilica in Rome during the papacy of Damasus (366–384),99 one of which attests to Anastasia’s development of the basilica’s baptistery. In the second of these inscriptions, Anastasia – a name with a pedigree in the Constantinian line (see Chapter 2, the sister of Constantine I and wife of Bassianus) – is noted as the mother of Gallus, and her son is referred to as the one who has “enhanced the glory of the [ecclesiastical] palace” (Gallus, Anastasiae natus, decus addidit aulae). As Chausson notes sagaciously, the involvement of Anastasia and her son Gallus in St. Peter’s reflects an atavistic engagement with the principal churches of Rome, following in the footsteps of Constantina and her initiative to develop an imperially derived ecclesiastical presence in the city. Constantina’s development of the area outside the Aurelian Wall on the Via Nomentana honouring the Diocletianic-era martyr Agnes included the foundation of a basilica dedicated to the virgin martyr alongside which was added a mausoleum where Constantina was interred on her death in 354 (see Chapter 6). Hillner has made the case that the mausoleum was subordinated to the basilica, thereby providing the impression that “the entire complex was less focussed on imperial commemoration than on the saint’s cult”.100 It appears that the same devotion to the sacred topography of Rome continued down the generations to Constantina’s daughter and grandchild.

Constantina’s untimely death from illness – Chausson surmises she was in her mid-thirties in 354 – thereby deprived Gallus of an absolutely crucial figure with regard to his relationship with Constantius and the emperor’s coterie of advisers. As reported by Ammianus, the letters sent from Constantius to Gallus requesting his presence in Milan contained the pretence that the Caesar was required to assist in the reconstruction of Gaul, a request reflecting Bleckmann’s suggestion (see Chapter 6) that Gallus was originally appointed to serve the western empire and Gaul specifically. An identical message was conveyed to Gallus in person by Scudilo, a tribune of the Scutarii, that Constantius intended, “to make him a sharer in his rank [i.e. an Augustus], to be a partner also in the labours which the northern provinces, for a long time wearied, demanded” (14.11.11). It is unlikely that these were genuine offers on the part of Constantius, and indeed Ammianus’ account emphasises the subterfuge which lay at the heart of the emperor’s plan to entice his Caesar to return to Milan in a way that did not arouse suspicion. Ammianus offers a hint of the possible backlash facing Constantius over his intention to dispose of Gallus: The Caesar’s Theban legions based close to Adrianople where Gallus was residing on his journey from east to west attempted to make contact with him in order to let him know the fate which lay ahead. Clearly, Gallus remained a very popular leader among his troops. However, Constantius’ intelligence network prevented contact from being made between them and Gallus. Ammianus discloses the effective leaders of this network. At its apex were Eusebius, together with Pentadius,101 a notarius who would later serve as Julian’s magister officiorum in Gaul, and Mallobaudes,102 a tribune of the Scholae Armaturarum, together led the enquiry into Gallus’ conduct while he resided at Pola in Istria, the final stage of his journey. Flavius Leontius,103 the quaestor sacri palatii, Lucillianus,104 Gallus’ comes domesticorum, and Bainobaudes,105 a tribune of the Scutarii, kept Gallus under surveillance during his time in Adrianople. As Gallus was moved to Poetovio (modern-day Ptuj) in Noricum, Barbatio,106 another comes domesticorum, and Apodemius,107 an agens in rebus, took over and were the ones charged with disrobing the Caesar and clothing him “a tunic and an ordinary soldier’s cloak” (Ammianus, 14.11.20). At Pola, where twenty-eight years previously Constantine I’s eldest son Crispus had been executed, Gallus was tried for the murders of the senior officials in Antioch. In his defence, Gallus put the blame on Constantina, a response which apparently enraged Constantius and effectively sealed the Caesar’s fate (14.11.23). And so, Gallus was bound and decapitated. The fact that Ammianus includes the detail that Gallus’ head and face were disfigured following his execution is important since it reveals a political reason – as opposed to a dynastic explanation – underlying the decision to eradicate the Caesar. It puts us partially in mind of the treatment of Nepotianus’ head which was impaled on a javelin and paraded through Rome following his defeat in summer 350 (see Chapter 6), with the main difference being that Nepotianus was killed during a civil conflict for the control of the city of Rome, while Gallus’ decapitation was the outcome of an event that bears the sign of an extra-judicial killing. Despite these differences, the mutilation of Gallus’ head invites comparison with Nepotianus in the sense that it too was a ritualised response by those who viewed the Caesar as having violated the principal “rules” underpinning the security and well-being of the res publica. As such, and as Troels Myrup Kristensen has pointed out, Gallus’ decapitation was very much part of a “political discourse” that viewed disfigurement as entirely apposite for enemies who had imperilled the safety of the state.108

A subtly different version of the last days and hours of Gallus to Ammianus’ account is presented in the Photian epitome of Philostorgius’ historical work (4.1: Amidon 2007: 63–65). As noted in Chapter 3, Gallus was Philostorgius’ imperial hero on account of his association with Aetius and Eunomius, the doyens of the heterousian cause, and Philostorgius’ account of the condemnation of Gallus reflects the historian’s regard for both him and Constantius II. In Philostorgius’ account, Gallus is exiled for the murders of Montius and Domitianus to an island off the Adriatic coast (“one of the islands of Dalmatia”: 4.1); according to Philostorgius, the deaths of the officials were warranted in light of their efforts to undermine Gallus (“those evil men [put] to an evil death”: 3.28). Theophilus, the Christian missionary who had earlier led an embassy to the Himyarites in order to convert them to Christianity (see below),109 accompanies Gallus as his adviser and serves as an intermediary between Gallus and Constantius. He attempts to intervene on the Caesar’s behalf but is also, in turn, exiled. The figure of Eusebius the Grand Chamberlain is prominent here, as he is in Ammianus’ narrative, but with the difference that he actively deceives Constantius in order to ensure that a sentence of death is passed on Gallus. Constantius finds out about this too late and despite an attempt to rescind the order as an act of clemency, Gallus is executed in his place of exile. Constantius is portrayed as repenting for Gallus’ demise. Constantius and Gallus are, therefore, portrayed as pawns in the political games of Eusebius in a manner befitting Philostorgius’ efforts to exonerate either imperial figure of wrongdoing. Indeed, the actions of Gallus in Antioch prior to his exile are attributed to the nature of the imperial office which made him “overbearing, ungovernable and implacable in anger” (3.28a from the Passio Artemii); the portrayal of the corresponding arrogance and conceit of Domitianus and Montius also, however, suggests that their grisly ends were entirely justified.

Philostorgius’ narrative of Gallus’ reign thereby allows a different image of him to emerge. The heterousian Gallus is a war hero, leading successful campaigns against the Sasanians (3.28) and, according to the material from the history extant in the Passion of Saint Artemius (3.28a), forging so enormous a reputation as a young and vigorous military leader that the Sasanians refused to engage the Roman army in combat and so, “it was at that time especially that the Roman Empire was completely at peace”. The truth about Gallus’ reputation and abilities lies somewhere in the middle. It is certainly correct to argue that Gallus’ record against the Sasanians was subject to, in the words of Peter Crawford, “some Ammian manipulation”,110 and it is perfectly reasonable to argue on the basis of the evidence that during his tenure as Caesar, Gallus scored some notable military achievements, regardless as to whether he was involved in the engagements himself. During the time of Gallus’ oversight of the eastern provinces, Sasanian forces were preoccupied with the demands of migratory parties on the eastern-most fringes of the Persian Empire (see Chapter 6) with the result that there was an almost instantaneous halt to the siege-style of warfare which had characterised Sasanian policy towards Rome during the 340s and the first year of the 350s. However, the Sasanians continued to try and find gaps in Rome’s defensive lines during the first half of the 350s, an approach which utilised intelligence and surveillance activities to a greater degree than had previously been the case. In September 353, a Sasanian official with the title (rather than a proper noun!) Nohodares111 bypassed heavily fortified Mesopotamia in favour of the less well-defended Osrhoene. At that point in the year, the Osrhoene town of Batna (modern-day Sarug112) held a yearly festival which drew large crowds attracted by the market which took place as part of the celebrations, where “traffic in the wares sent from India and China” were sold (Ammianus, 14.3.3). In the estimation of Ammianus, the resulting assault would have been devastating to Roman fortunes in the east, presumably because it would have struck a highly popular albeit illicit trading zone (recall the mandating of Nisibis as the sole area of cross-border exchange of goods in the treaty of 298; see Chapter 6), with the resultant loss of revenue for key settlements in Osrhoene.113 As it was, however, the Sasanian assault was betrayed by defectors to the Roman army (14.3.4). Ammianus notes that some months later Gallus (c. March 354114) was on the point (profecturus) of leaving for Hierapolis to take part in a campaign, but it is unclear whether this took place and whether it was related to a renewal of hostilities against the Persians in light of the plan to assault Batna some months earlier.115

More concrete evidence of military activity during Gallus’ reign emerges from Ammianus’ account of an encounter with the Isaurians, a “barbarian” people from southern Asia Minor, which Barnes dates to 353 as opposed to the widely held date of 354.116 The late Keith Hopwood, one of the most authoritative commentators on Isauria in recent times, provides the following description of the geography and history of Isaurian territory:

Isauria was one of the Diocletianic provinces created in the last decades of the third century AD. It corresponded to one of the eparchies of the earlier province of Cilicia, existing at least from the time of Antoninus Pius. The origins of this division are disputed. The Diocletianic provinces corresponded to the geographical area of Cilicia Tracheia or Aspera (‘Rough’ Cilicia). It represents the Taurus Mountains’ eastern thrust to the sea, between Coracesium (Alanya) and Seleucia-on-the-Calycadnus (Silifke). It consequently forms the southern-eastern mountains barrier between the Mediterranean coast of Asia Minor and the central Anatolian plateau.117

Ammianus narrates the details of a conflict which he claims arose as a result of an incident in the northern city of Iconium (Konya) where captured Isaurians had been thrown to the wild beasts in the amphitheatre.118 This was “beyond the established custom”, and as a result led to the outbreak of “a serious war” (14.2.1). As Hopwood’s analysis highlights, these very specific details about the events of 353 are meshed together with a number of historiographic commonplaces about the Isaurians as “bandits” (latrones): “Ammianus’ description fits both our derived concepts of the portrayal of outsiders in Roman historiography and our ethnographically derived views of rural revolt”.119 On the basis of Ammianus’ narrative of events, John Matthews characterises the conflict “as a case of rebellion or civil war”,120 and views the episode (and later such Isaurian rebellions in 359 and 368) as determined by the frequently experienced dire economic circumstances of mountain people. Indeed, as Matthews notes, their initial target was the coastline adjacent to Cyprus, in the opposite direction from Iconium, where they raided harboured boats and confiscated cargo. Next, they moved to neighbouring Lycaonia where they blockaded roads and commandeered the belongings of travellers. For the most part, they managed to evade Roman soldiers who had been deployed against them and headed for Pamphylia. From there, the Isaurians tried to cross the river Melas but were repulsed by troops quartered at Side (14.2.10). Ammianus concedes that “severe hunger” pushed the Isaurians to attack the town of Laranda in Lycaonia and attempt an assault, which failed, on a fortification called Palea which served as the food store for troops serving in Isauria. Seleucia on the banks of the river Calycadnus, the region’s metropolis, was then besieged following a stand-off with the troops stationed there. The Isaurians captured the grain supply headed for Seleucia which left the soldiers and citizens of the city in dire straits. It was at this point that Gallus ordered Nebridius, the comes orientis, to relieve Seleucia – Ammianus notes disparagingly that the Caesar had to be contacted incessantly to act in defence of the city (14.2.20) – at which point, with the threat of a large-scale relief army appearing over the horizon, the Isaurians disappeared to their mountain territories.

In the same period (353), Ammianus describes the raiding activities of the Arabs (Saraceni121) whom, he notes, “we never found desirable either as friends or as enemies” (14.4.1). Irfan Shahîd in his monumental study, Byzantium and the Arabs in the Fourth Century, develops the argument that the Arabs in Ammianus were former allies (foederati) of Rome who had been bound by a treaty of alliance (symmachia) that was renewed by Constantius II in 338 when he was engaged in his early campaigns against the Sasanians and preoccupied with the question of Rome’s sovereignty over Armenia (see Chapter 6).122 Julian alludes to the renewal of this alliance with in all likelihood the Lakhmids of southern Arabia and the Tanūkhids of the north when he refers to Constantius’ achievement in sending embassies to Arabia with the result that “you (sc. Constantius) turned the marauding Arabs against our enemies” (Or. 1.21b): Here, Julian is indicating that the Arabs were part of Constantius’ early military campaigns against the Sasanians. According to Shahîd, Constantius’ renewal of the alliance with the Arabs in 338 followed a period of discontentment on their part following the likely deposition of Pamphilus during the 330s, the bishop of the Arab foederati whose name is included in the subscription list of the Council of Nicaea as Pamphilus of Tayenoi, which Shahîd associates with the Tanūkhids.123 Shahîd therefore includes Pamphilus among the prominent Nicenes exiled c. 334–335,124 one outcome of which was that the alliance between Rome and the Arabs fell into abeyance. It was revived by Constantius during 338, but was dissolved once again in the early 350s, which is offered as the explanation for the renewal of Arab raiding described by Ammianus at 14.4.1. Aspects of Shahîd’s thesis are built on circumstantial evidence and triumphalist church historical sources, but as more recent studies of the period between Nicaea and the death of Constantine have shown, the Eusebian “Arian” network “ended the decade [of the 330s] in virtual control of the churches of the East”.125 It is, therefore, plausible that the rupture in Roman-Arab relations following the death of Constantine I and the renewal of their alliance in 338 was driven by the politics of ecclesiastical history.126 Indeed, in light of the fact that Shahîd’s study dates from 1984, his efforts to expose the way in which “the contours of political and military history follow those of ecclesiastical history”127 were remarkably farsighted and as such his observations about the reign of Constantius remain very valid.128

Ammianus’ near pairing of the Isaurian uprising and the raiding activities of the Arabs – the two accounts are separated by 14.3.1–4 detailing the aborted Sasanian raid on Batna – was intentional: Both accounts were intended to suggest that Gallus’ “tyranny” encouraged episodes of instability to arise within the empire among long-standing associates of Rome and to convey the judgement that his government was ineffective in this regard since it was riven by intrigue and infighting. Despite the characterisations offered by Ammianus which presents both parties as “outsiders” in a classically conceived sense, the extent of the integration of both the Arabs and Isaurians in various aspects of the empire was considerable. As Matthews valuably highlights about Isauria: “For all its unruliness, Isauria was part of the Roman empire, and it is an error to think of it as a permanent enclave of dissension”.129 A positivist interpretation of Ammianus’ narrative at this point is certainly encouraged in light of the general depiction of Gallus as a tyrant who failed in all aspects of government. Thus, Shahîd wonders whether Arab unrest was caused by the “unenlightened policy” of Gallus which led to their “alienation” (he fails to elaborate on what an “unenlightened policy” would comprise!).130 The idea that the truncated reign of Gallus presented the impression of a weakened administration in the eastern half of the empire which, in turn, encouraged a number of internal revolts by allies and associates of the empire is a plausible but ultimately unsatisfying explanation for the events of 353, as is the idea that Gallus’ style of rule encouraged insurrections. The multitudinous layers of the civil and military administration comprising the workings of the empire together with attendant projections of imperial continuity implicit in the promotion of the imperial persona made it unlikely that citizens, not to say non-citizen allies, were aware of a failing emperor. The curtailing of the Isaurian conflict by the deployment of the comes orientis and troops to Seleucia-on-the-Calycadnus demonstrates that the measures needed to maintain the internal security of the empire worked effectively even under Gallus.

One of the more famous internal threats to the stability of the eastern empire during Gallus’ reign is not mentioned in the extant portions of Ammianus’ work.131 The conflict appears to have arisen in Diocaesarea (Sepphoris) in the central Galilee region and is referred to by the earliest extant source for the event as “a revolt of the Jews” (seditio Iudaeorum).132 The ensuing Roman suppression of the revolt led to the destruction of Diocaesarea along with Tiberias (on the Galilean Sea’s western shore), and the southern city of Diospolis (Lydda, modern-day Lod).133 Archaeological investigations of destruction and abandonment layers from sites in the Palestinian region point to major disruption during the mid-fourth century,134 although the revolt during the reign of Gallus is but one likely cause for the evident social upheaval which also included the devastating earthquake that hit Palestine in May 363.135 Aurelius Victor adds the enigmatic detail that the Jews “criminally elevated Patricius to a sort of kingship (regnum)” (42.11–12), and the twelfth-century author George Cedrenus136 (I. p. 524, Bonn Corpus) notes uniquely that the Jews killed Hellenes and Samaritans. The entry covering the episode in Jerome’s Chronicle is placed under the year 352, although there are grounds for also regarding 351 as a valid date.137 While the Greek and Latin sources are near-unanimous in declaring this a Jewish uprising, according to Joseph Geiger, the evidence of the Jewish sources – notably the Jerusalem Talmud – does not support the idea that this was a Messianic liberation movement in the mould of the rebellion led by Simon ben Kosevah (Bar Kokhba) between 132 and 135 CE138: Indeed, the silence about Patricius there is described as “roaring”.139 Geiger’s assessment of the episode proposes seeing the conflict as a usurpation by the figure of Patricius (a Roman officer?), to whom the Jews of Palestine lent their support, on the basis that the term regnum in Victor is used elsewhere to denote an imperial pretender as, for example, in his description of Calocaerus (41.11), the Master of the Camel Herd, who attempted to usurp in Cyprus in 334 (see Chapter 4). Geiger argues: “… there is nothing inherently impossible in a pretender whose main, or even exclusive, support comes from one ethnic element”,140 and cites the example of the Bagaudae as an “ethnic support base” for Aelianus and Amandus at the end of the third century.141 While the silence about Patricius is “roaring”142 in the Jewish sources, we do find references to Ursicinus in the Jerusalem Talmud143 (e.g. Sanhedrin 3:3, 21b (= Shevi’it 4:2, 35a)) where two amoraim, Rabbi Yonah and Rabbi Yosi, instructed the baking of bread on Shabbat for Ursicinus; it is therefore assumed that the context for Ursicinus’ appearance in the Talmud is the insurrection during the reign of Gallus, although there is nothing in the text to suggest that Ursicinus compelled the rabbis to transgress the law in this example.144 The deployment of the magister equitum to Palestine is more than a hint that a serious incident or series of incidents had occurred which threated the security of the region; as Barnes points out, however, Ursicinus’ presence was a sign that Gallus “did not visit the theatre of war himself”.145

In contrast to Geiger, Oded Irshai has lately cautioned against dismissing entirely a messianic-religious explanation for the violence in Palestine during 351/352. Irshai presents the episode not as a full-scale insurrection but a “chain of violent regional outbursts” which attest to “a possible prevailing religious, political, and even messianic ferment” in Palestine.146 Irshai notes the prevalence of apocalyptic concerns in Christian literature in this period, most notably the appearance of the cross over Jerusalem in May 351 which Cyril of Jerusalem interprets eschatologically in light of Mt. 24.30 in his letter to Constantius II (see Chapter 3), and also in pagan circles via the appearance of traditions associated with the Sothic cycle, the notion of the “Great Year” and the iconography of the phoenix: In this regard, Israhi cites the article by Olbrich from 2004 discussed in Chapter 2 without seeming to realise that Olbrich’s contribution situates the revival of these motifs in the fourth century squarely within the imperial iconography of two Christian monarchs, namely Constans and Constantius II. Nevertheless, Israhi manages to connect what appears to be a rising interest in messianism across a range of traditions with the social and political circumstances of the time. Two factors are identified in this regard. The upheavals caused by Magnentius’ seizure of power in the western half of the empire, and the ongoing conflict against Sasanian Persia, encapsulated in the siege of Nisibis during 350. In relation to the latter, Irshai offers valuable commentary that builds on an earlier idea in Menahem Stern’s Greek and Latin Authors on Jews and Judaism (vol. 2:501) that the war between the Sasanians and the Romans, in particular the third Persian assault on Nisibis, spurred the Jewish uprising in Palestine. The “cultural significance” of Nisibis to the eastern empire was such that,

[i]n the eyes of that generation, it was a genuine symbol and a focus of interest among travelers and merchants who reported at length on the goings on-there, and a city on whose behalf it was fitting to invoke divine powers.147

The evidence from the ninth- or tenth-century Jewish chronicle, the Seder ‘Olam Zuṭa (The Minor (Smaller) Order of the World), cited by Irshai, may very well capture Jewish responses to the siege of Nisibis in 350 and the sense in which the city was understood to have been “conquered” by Shapur II, although as Irshai notes, it is a moot point as to whether the entry in the chronicle in fact records the conceding of Nisibis to the Sasanians by Jovian in the treaty of 363.148 As we noted in Chapter 6, the siege of 350 itself was rapidly mythologised by numerous authors, but it was the ultimate failure of the Sasanian assault rather than the ineptitude of Constantius and the Romans which was recalled in these sources and was portrayed so that the defence of Nisibis became central to the image of Constantius II as a dynamic emperor who, according to the fourth-century source underlying the Chronicon Paschale, led its defence in person (see Chapter 6). It is difficult to agree with Irshai’s assessment, therefore, that at the time of the unrest in Palestine, the siege of Nisibis was judged to be a sign of the Roman Empire’s impending implosion: Indeed, the opposite impression seems to have been the case. Furthermore, a thesis proposing rising messianism or apocalypticism does not constitute an explanation for the so-called “Gallus Revolt”, and in this regard there is zero consensus about the reasons for the actions of the Jews. An economic explanation for Jewish “discontentment” that once held sway for understanding the events of 351/352 seems to have fallen out of favour: The idea that enforcement by the Roman authorities of punitive economic measures such as exorbitant rates of taxation or laws prohibiting Jews from owning slaves (e.g. Cod. Theod. 16.9.2149) now runs counter to the assessment that the year 350 marked a time of prosperity in Palestine and Syria as seen, for example, in the great period of synagogue construction.150

Israhi’s wider political explanation for the Jewish revolt is very welcome since it stands in contrast to the idée fixe, which derives ultimately from Ammianus Marcellinus and is prevalent in much modern commentary about Gallus, that the Caesar’s style of governing the east was chaotic and cruel with significant internal unrest arising as a result. The events in Palestine may very well have been handled in a similar manner to Ammianus’ presentation of the threats emerging from the Arab tribes in the Arabian Peninsula and the Isaurians in southern Asia Minor in the portion of his work that is now lost (book thirteen), although two considerations militate against this idea. First, book fourteen is from its outset a jaundiced primer about Gallus’ regime, thereby seeming to exclude a prior presentation of the events in Palestine during 351–352 in the manner of his account of the activities of the Isaurians and the Arabs; and second, the role of Ursicinus in the suppression of the revolt which is so well-attested in the Rabbinic sources suggests that Ammianus’ portrayal of the episode may have reflected his admiration for the magister equitum.151 It could have been the case therefore that Ammianus narrated details of the Jewish rebellion without mentioning Gallus at all! Israhi’s examination of the “Gallus Revolt” elevates discussion of the east in this period beyond a fixation with the nature of Gallus’ government and locates it in a broader context whereby the evidence for all localised episodes of unrest and violence in the period during and after the usurpation of Magnentius, including the oft-repeated deplorable condition of Gaul (cf. Ammianus, 14.11.10), indicates that ways of thinking (messianism; apocalypticism) and acting (armed rebellion; piracy; raiding) were emerging – or at least re-emerging – across the empire as responses to what was perceived as the susceptibility of Roman power in the period during and after a civil war.152

Saints and Missions, Trade and War

A further aspect of Gallus’ reign which has been all but lost to posterity was the extent of his connectedness to the Christianity of the eastern empire. From the beginning of his time in Antioch, Gallus appears to have enjoyed close relations with a number of key figures in the history of the Eastern Church during the mid-fourth century beginning with Basil of Ancyra and Eustathius of Sebaste. Both individuals had become “imperial favourites”153 and were to be instrumental in shaping the theological settlement referred to as homoiousios (that the Father and Son are “of like essence”) which found favour (albeit briefly) with Constantius in the final years of the 350s. In spring 351, both Basil and Eustathius appear to have formed part of Gallus’ entourage as the Caesar entered Antioch. Details of his association with the two bishops are preserved by Philostorgius who narrates Basil and Eustathius’ attempts to undermine Aetius, a much celebrated figure who had served as a deacon in the Antiochene church under Leontius, the bishop of Antioch, until he had been deposed, although by the time of Gallus’ reign as Caesar, he appears to have returned to Antioch.154 According to Philostorgius’ partisan account of Aetius (see Chapter 3), the teacher of Eunomius was himself tutored by the disciples of the revered Lucian of Antioch – notably, Paulinus of Tyre, Athanasius of Anazarbus,155 the presbyter Antony and Leontius of Antioch – and as we noted earlier, Philostorgius elevated Aetius and Eunomius above even Arius as the unblemished heirs of Lucian’s legacy (Hist. Eccl. 3.14) as exponents of the position that the Son was of a different essence to the Father (characterised as anomoios, that the Son is unlike the Father, or by the preferred synonym, at least among modern commentators, of heteroousios, that the Son is of a different essence to the Father). The Christological positions of Basil and Eustathius on the one hand, and Aetius and Eunomius on the other, have traditionally been considered to reside on a spectrum of “Arian” thought that emerged following Nicaea, but as Richard Paul Vaggione’s analysis of Eunomius has demonstrated, Aetius and Eunomius repudiated Arius’ teaching regarding the incomprehensibility of God,156 and their repudiation of Basil and Eustathius’ position was determined by the idea that God and Christ were different from one another at the level of their essence (ousia).157 The term homoian as opposed to the loaded term Arian currently appears to be the chosen way of referring to those teachers – especially the ones in the eastern half of the empire – who taught in varying ways the subordination of the Son to the Father in the sense that the promotion of “likeness” represented a rejection of the Nicene position that the Father and Son were “of the same essence”.158 As Sara Parvis has noted, “‘Homoian’ … can be meant in a strict or loose theological sense, or in a historical-political sense, or in a combination of these”159; in the strict sense, it refers of course to the theological settlement devised in 359 which sought to prohibit essence or substance-based terminology.160 The strict subordinationism of Aetius, therefore, technically excludes him from participation in homoian as a category, although he is not infrequently referred to as a homoian by commentators.161 Rivalry, theologically driven or otherwise, had led Aetius to clash with Basil and Eustathius (Philostorgius, 3.16) and in response, Basil and Eustathius had approached Gallus for his assistance in reducing or eradicating entirely Aetius’ influence in the theological circles of Antioch. Photius’ epitome of Philostorgius’ history preserves the encounter between the bishops and Gallus (3.27): Basil and Eustathius fabricated accusations about Aetius to Gallus following which the Caesar ordered Aetius’ arrest and torture (his legs were to be broken). At this point, Leontius, the bishop of Antioch, intervened and Aetius’ ordeal was terminated (presumably before his legs were actually broken!). Gallus and Aetius then met and from that point onwards, they developed a friendship, albeit a friendship with a condition: Gallus sent Aetius to Julian, his younger brother, when he learned that Julian was entertaining thoughts of turning from Christianity to worship of the traditional gods (“he was sent in order to save him from false worship if at all possible”: Philostorgius, 5.27) which looks like a fanciful intervention since Julian did not disclose his attachment to the traditional cults publicly before he was in the secure position as a favourite of the army in Gaul in the lead up to what turned out to be the non-event of his confrontation with Constantius (e.g. Letter to the Athenians 277a, 280d, 284b–c, 285c, 286d, 287d).162 (It seems unlikely that he would have disclosed his new allegiance to his half-brother in light of the fact that Gallus was Caesar and that he was becoming increasingly renowned in Antioch and beyond for his enthusiastic patronage of (homoian) Christianity.) Gregory of Nyssa’s Against Eunomius (1.47) preserves an alternative version of Aetius’ initial encounter with Gallus: According to Gregory, writing in 380, “Blemmys Theophilus” – otherwise known as Theophilus “the Indian” (Philostorgius, 3.4) from Divus (modern-day Socotra?), an island in the Gulf of Aden163 – introduced Aetius to Gallus on the basis of his prior connection to the Caesar’s court.164 Theophilus was a key figure in the promotion of homoian-oriented Christianity during the 350s and following the death of Gallus, he played a central role in Constantius’ plans to expand imperial authority into the territories on the Red Sea, the Gulf of Aden and beyond.

A central figure supporting Gallus during his reign in Antioch was Leontius. As noted above, Leontius is identified by Philostorgius as one of Lucian’s disciples and by the year 344, he had been made the bishop of Antioch following his predecessor Stephen’s deposition after the details of the murky plot involving a sex-worker and a party of western bishops came to light following the Council of Serdica (see Chapter 5). Leontius was heavily networked in the theological and didactic life of the Antiochene church even prior to his episcopal appointment since we are informed that Aetius studied the books of the Prophets with Leontius (“especially Ezekiel”, according to Philostorgius, 3.15). Leontius’ association with Lucian indicates that his own theology lay somewhere on the homoian spectrum which emphasised – to varying degrees – the difference between the Father and the Son. The Christian church in Antioch was split for many years following the deposition of Eustathius, the Nicene bishop of the city (to be distinguished from Basil’s ascetically inclined associate of Sebaste!), in 330/331, after whom a succession of homoian appointments had been made.165 Leontius’ alliance with Gallus was, therefore, in keeping with the dominance of non-Nicene, homoian-inclined clergy in the city, although it is important remind ourselves that a Nicene presence did in fact remain in Antioch during these years.

Christine Shepardson’s study from 2014 on Antioch has highlighted the manifold tensions underlying the city’s Christian factionalism and in particular the role Christian topography played in shaping those tensions. In this regard, Leontius led on one of the most notable changes to the sacred landscape of the area by overseeing the translation of the relics of his third-century episcopal predecessor, a figure named Babylas martyred during the reign of Decius (r. 249–251; Eusebius, Church History 6.39.4), from the Christian cemetery outside Antioch to the suburb of Daphne which lay c. six kilometres to the south-east of the city.166 A temple of Apollo was situated there, the original sanctuary was established by Seleucus I in c. 300 BCE167 and it housed a colossal acrolithic statue of Apollo which issued spoken oracles.168 The importance of the site to the religious and political events of the fourth century cannot be underestimated.169 It is perhaps best known as a result of Julian’s attempts to re-establish the prophetic vitality of the temple and its attendant Castalian spring when he ordered the removal of bodies interred close to the temple in 362. According to Christian accounts, Babylas’ presence was regarded by the emperor as inhibiting the daemon at the site from speaking his oracles and as a result Julian ordered his relics to be returned to the Christian cemetery.170 By contrast, Ammianus’ account (22.12.8) emphasises that Julian’s decision derived from a tradition which sought to purify sacred sites following their pollution by the burial of corpses close by (the author references the cleansing of Delos by Pisistratus, Herodotus 1.64). Following Ammianus’ narrative (22.13.1–3), in October of the same year, the temple and statue of Apollo burnt to the ground: Julian blamed the Christians of Antioch for the incident but, following an investigation, the fire was found to have been caused inadvertently by the philosopher Ascelpiades who had left some wax tapers burning near the composite statue, following his dedication of an image of the Heavenly Goddess (dea caelestis; in all likelihood, a reference to dea Syria171), which sparked and set the statue and sanctuary alight. Julian held the Christians of the city accountable for the incident and ordered the closure of the Great Church of Antioch, the church inaugurated in January 341.172

A less clearly acknowledged contribution of Apollo at Daphne to the history of the fourth century has been teased from the sources by Elizabeth DePalma Digeser. Constantine I’s letter “to the Provincials of the East”, sent following his defeat of Licinius in 324 and preserved in Eusebius’ Life (2.48–60) of the emperor, refers, so Digeser establishes, to the initial response of Diocletian c. 299 to an oracle declared by Apollo at Daphne that “the righteous on earth prevented him [sc. Apollo] from speaking truly, and that that was why he was composing false oracles from the tripods” (2.50). The Tetrarch enquired as to the identity of the “righteous on earth”, and was told by a haruspex that they were “Christians, I suppose” (2.51). Diocletian reacted by writing, “the edicts of carnage, and urged the magistrates to apply their native ingenuity to the invention of unprecedented tortures” (2.51),173 which led, in turn, to the purge of Christians at the imperial court and in the army. As Digeser notes, Constantine’s account is frequently conflated with the account in Lactantius’ On the Deaths of the Persecutors (11.6–8) detailing Diocletian’s despatch of a haruspex to the temple of Apollo at Didyma (Miletus) in 303 which presaged the “Great Persecution” in February of that year, but Constantine’s own (partially eye-witness) version refers to the Apollo at Daphne, near Antioch, and to a time some four years previously.

Evidence for the influence of the declaration of the Apollo at Daphne on Diocletian’s decision to persecute Christians, first in his immediate vicinity of the palace and the army (cf. On the Deaths 10.4–5), and then later following pressure from Galerius and the hostile utterance of the Apollo at Didyma (On the Deaths 11.7–8) in launching a general persecution of Christians within the empire, thereby offers a more complex context for the translation of Babylas’ relics to the suburb outside Antioch during the reign of Gallus. The decision of the Caesar and Leontius to move Babylas’ body from the Christian cemetery to a purpose-built martyrion takes on the guise of a calculated act of revenge to punish Apollo – or at least, the sanctuary associated with the God – for his role in the century’s formative persecution of Christianity. Writing in the fifth century, Sozomen adds a further piece of evidence by way of his comment (Church History 5.20), following his description of the fire at the temple in late 362, that martyrs’ had also been interred (at some point) near the sanctuary of Apollo at Didyma. As Digeser notes, “both sites [paid] for their involvement in the persecution”.174 Leontius’ role in the initiative may have been downplayed, according to Susanna Elm,175 as a result of his homoian leanings, and indeed Gallus tends to be credited with the act of translation by Nicene-leaning writers and ecclesiastical historians. However, memories of the persecution under Diocletian and Galerius extended far beyond the reign of Constantine I, particularly it seems into later generations of the Constantinian family. According to the homily delivered c. 378 prior to the restoration of Babylas’ body to a martyrion under construction (completed, c. 380/381176) by the Nicene bishop of Antioch, Meletius, John Chrysostom tells how “it was God who inspired” Gallus to transfer the body from the Christian cemetery to the sanctuary of Apollo in order to eradicate the “debauchery” for which the pagan site had become famous (Discourse on Blessed Babylas and Against the Hellenes 67).177 In this regard, the role of Gallus should not be downplayed. In contrast to the position of Hanns Christof Brennecke, who has argued that Gallus’ role was one of a mere state functionary who simply facilitated Leontius’ grand idea,178 placing the Caesar front and centre of the decision to promote the worship of the martyred Babylas at Daphne is in keeping with what we know about the Caesar’s familial predecessors’ numerous acts of intervention and subsequent promotion of a new sacred Christian topography across the empire. It also projects an image of Gallus as a dutiful Christian emperor and a vehement opponent of the empire’s recent “pagan” past.

Another central figure in Gallus’ homoian entourage was Theophilus “the Indian”, the individual mentioned earlier. He is credited with serving as a mediator between Gallus and Constantius during the period when the Caesar was recalled to Milan, and despite receiving a sentence of exile for his loyalty to Gallus (Philostorgius, 4.1), he survived the purge of Gallus’ enablers,179 was later recalled and continued to serve Constantius. His exile was commuted as his services were required by Constantius to cure Eusebia of the hysteria afflicting her (Philostorgius, 4.7) and to which, according to Zonaras (13.11), she eventually succumbed. In light of the significance that the epitome of Philostorgius’ work plays in the reconstruction of the career of Theophilus, it is important to recognise that the homoian historian’s portrayal of Theophilus, Aetius and Eunomius is an authentic reflection of the self-identities of these figures as “true philosophers” who established their authority “from their divinely authorized ‘quest for truth’ [Philostorgius, 4.12] made manifest by ‘divine signs’ [Philostorgius 9.1] and conveyed through the ‘power of words’ [Eunomius, Apology 27]”.180 As Elm adds: “Highly trained, they had received a divine calling to philosophy, which they understood as active participation in the life of the politeia as ambassadors, physicians, teachers, and bishops”.181 Theophilus’ episcopal status is moot. Ordained a deacon by Eusebius of Nicomedia (so, no later than 341), Philostorgius (3.4) notes that on his return from his embassy to locations along the Red Sea and the Gulf of Aden in 356–357 (see below), Theophilus “received the dignity of bishop from those who shared his beliefs” (3.4), and while he was not awarded a see of his own, he “belonged to all in common” (3.6, and 6a (Suda, a Greek encyclopaedia from the 900s)). It seems, therefore, that Theophilus was an episcopal adviser, holy man182 and “fixer” for both Gallus and Constantius during the 350s, and his elevation illustrates that both courts prized figures who combined holiness with political and diplomatic acumen. As a passage from the Suda which drew on Philostorgius memorably claims:

the man was greater than words can describe: he was like an image of the apostles (Gk. o-s an tis to-n apostolo-n eiko-n). It is in fact said that he once revived a dead Jewish woman in Antioch. Thalassius [PPO in the East] says so, and he is one of those who spent a great deal of time in the man’s company and is the last person to be suspected of lying about such matters.


In the period following his return from exile, Theophilus was chosen by Constantius to serve as his ambassador to the kingdom of Himyar, a territory situated on the southwestern Arabian Peninsula in the area that is modern-day Yemen.183 Philostorgius’ epitomised account (3.4 and 4a (from Simeon Metaphrastes’ Martyrium Arethae); 4b) presents the principal rationale for the embassy (presbeia) as a religious mission: Constantius’ main interest was the conversion of the Himyarites to Christianity (according to the epitome, this was anhomoian in character, 3.5) achieved by way of largesse presented to the monarch, Ta’ran Yuhan‘im,184 and the construction of a church for Romans travelling in the region. In addition to the financial resources needed to build a new church, the embassy was also given two hundred of the finest Cappadocian horses in order “to strike wonder at their sumptuousness and to enchant the beholder” (3.4). Once Theophilus arrived, he set about convincing the ruler (Gk. ethnarchēs185) of the Himyarites to worship Christ and in the process bested in debate Jewish leaders in the kingdom; the presence of Jewish communities in the peninsula stretched back to the first century CE, and towards the end of the fourth century, the Himyarite monarch, Abikarib As‘ad, converted to Judaism.186 The embassy was successful: The ruler converted to Christianity and built – out of his own pocket rather than from the funds brought by Theophilus since it is said he was so impressed by Theophilus’ deeds – three churches: One in the kingdom’s capital (Zafar), another in the Roman market in the port city of Aden and another in an unnamed Persian market at the mouth of the Persian Gulf (perhaps in the port city of Qāna’(?)).187 Following his time among the Himyarites, Theophilus travelled on to his homeland of Divus, and from there to the “other India” (allē Indikē = referring to Malabar, thereby following an established trading route) where he corrected the practices of Christians there, including the habit of listening to readings from the Gospel while remaining seated. The epitome then abruptly188 situates Theophilus back in Great Arabia from where he sailed to Aksum in northern Ethiopia where he “took care of matters there, and then returned to the Roman Empire” (3.6).

There is a great deal to unpack in this account of Constantius’ commission to Theophilus to missionise in Himyar, Malabar and Aksum. A recent study by Walter Stevenson has ably demonstrated that Constantius, in the period following the defeat of Magnentius and thereby freed from the burden of civil conflict, undertook a series of grand schemes to enhance and expand Roman interests along the Red Sea in East Africa and Arabia and beyond towards the Indian Ocean. The mission of Theophilus, which Stevenson dates to 356–357, took place around the same time that the most important trading emporia (Koptos, modern-day Qift) and ports on the western side of the Red Sea – notably Berenike – were undergoing significant infrastructural renewal.189 The emerging economic importance of the kingdoms of Aksum and Himyar190 in the period cemented an existing relationship between Roman traders and the ports of the Red Sea, and consolidated the spread of Christianity to those areas which the embassy led by Theophilus was looking to develop still further by encouraging the rulers of those regions to construct churches that would, as the epitome indicates, serve Roman travellers and facilitate the conversion of the local population (3.4). In addition to these economic and religious interests, Shahîd had earlier highlighted the important role which the embassy played in shoring up Roman interests in southern Arabia in light of rising Sasanian activity, beginning around 355, which looked to test the limits of the “fragile and uneasy”191 truce that had held for around five years between the Roman and Persian governments. Shahîd proposes that the third church constructed in the Himyarite kingdom mentioned by the epitome, located somewhere near the mouth of the Persian Gulf, “could reflect a bold and imaginative attempt on the part of Constantius to watch the enemy from its own backyard”.192 Stevenson’s recent analysis of Theophilus’ embassy has augmented the idea that Constantius had strategic concerns in mind beyond simply commerce and conversion which he establishes by focusing on key items of evidence (see below), although it is an error to compartmentalise these issues since they all served the same end, the expansion of Rome’s imperium into areas of contested political and cultural importance – for example, Christianity indeed represented “the most direct channel for introducing Roman political influence”193 to the area, but religion was also in the mind of Constantius the sine qua non for guaranteeing the divinely sanctioned success of the imperial project.

Taking these items out of the order handled by Stevenson, we begin with the constitution (Cod. Theod. 12.12.2) in the names of Constantius and Julian from 15 January 357194 addressed to the Praetorian Prefect of the East, Strategius Musonianus. The ruling notes that: “No one commanded to visit the peoples of Aksum and Himyar should delay more than one year in Alexandria nor should he draw travel allowances beyond one year” (trans. Stevenson 2012: 72). The recipient of this letter was a hugely significant figure. Strategius had been engaged by Constantine I to investigate “superstitious sects”, including Manichaeans “and the like” (whose role in this regard was preserved in a “bon mot195 by Ammianus about the corruptibility of Roman officials),196 and had been Constantius II’s representative at the Council of Serdica in 343, and then later Proconsul of Constantinople and Proconsul of Achaea. He was appointed Praetorian Prefect following the death of Domitianus in 354.197 As Stevenson establishes, one of his many duties in this role included co-ordinating the activities of imperial agents in the eastern provinces of the empire at a time when traffic of these officials began to increase significantly in light of the thawing of the stalemate with Shapur II. Libanus’ reference (ep. 430.6–7) to his close associate Clematius,198 an agens in rebus, and his mission to gather intelligence about Persian activity beyond the Euphrates in the summer of 355, illustrates that agents were encouraged to report intelligence directly to Strategius.199 Clematius’ journey into Persian territory was almost certainly linked with the covert embassy of 356 which approached the Sasanian commander, Tamshapur,200 with the proposal of a more formal peace between the two powers (Ammianus, 16.9.3–4), and which, in turn, led Constantius to respond defensively in his reply to Shapur II’s imperial missive from 357 (Ammianus, 17.5.12; see Chapter 6). Stevenson holds that Strategius was running, on behalf of the emperor, a network of agents drawn from a variety of roles and positions in the empire and that in the case of the constitution from January 357, Constantius was calling on the Praetorian Prefect to address some of the abuses pursued by individuals in this network. Thus, “no one commanded to visit the peoples of Aksum and Himyar should delay more than one year in Alexandria” is taken to indicate that government officials had exploited the associations and contacts they had made during their journeys along the Red Sea by, for instance, establishing their own patronage networks out of Alexandria which risked interfering with the business of the imperial government.201 Furthermore, Stevenson proposes that the constitution was addressed to Strategius Musonianus rather than the governor of Alexandria because the emperor could not be certain of the Alexandrian administration’s loyalty to the imperial government in light of the level of unpopularity with the removal of Athanasius in February 356.202

Aksum and Himyar together203 “created a very strong power block”204 that had the potential to dominate trade in frankincense and myrrh and to control the Bab el-Mandeb, the strait connecting the Red Sea to the Gulf of Aden.205 It is plausible, therefore, to regard Roman maritime oversight of the southern Arabian trade route as ranking highly among Constantius’ strategic priorities during the ratcheting up of the phony war between Rome and Persia between 355 and 357, since it would undercut Persian trading activities in the region. (The economy of the Sasanian Empire was much more dependent on external international trade than the economy of the Roman Empire.206) However, Stevenson regards this idea as an anachronism in light of the vastly more important role played for Sasanian trade by the “burgeoning Silk Route” and the ports along the Persian Gulf and the south-eastern areas of the Arabian Peninsula.207 While this observation is debatable, it would appear that factors other than conflict with Persia played a role in Theophilus’ journey to Himyar and Aksum. This brings us on to another key item of evidence for Stevenson’s argument: Constantius II’s letter to the monarch of Aksum, Aezanas, and his sibling and co-regent, Sazanas.

This text, preserved in Athanasius’ Defence before Constantius – a work originally composed between 353 and 355, but edited with the addition of supporting documentation in 357208 – derives from late 356 or early 357 and sits in the Defence alongside another imperial missive by Constantius, this time to the people of Alexandria (Defence 30), which dates from around the same time (both letters cannot be later than summer 357 when they were incorporated by Athanasius into his work209). In broad terms, both letters condemn Athanasius, who had fled his see earlier in 356, and promote the new regime of George of Cappadocia in the city. The late date for George’s arrival in Alexandria on 24 February 357 (“the thirtieth day of Mechir”: Historia Acephala 2.1) does present a chronological conundrum in relation to the letters being “leaked” to Athanasius in time for their inclusion in the redacted Defence.210 Instead, the letters may very well belong to the period between February 356 and February 357 when there was no episcopal presence in the city and during which there were some serious civil disturbances, notably the riot in the Caesarion’s Great Church in June 356 led by the city’s pagan population, unrest which may have arisen ultimately as a result of there being no ecclesiastical leadership to quell “Alexandria’s precariously balanced intercommunal relations”.211 That George’s presence in the city is alluded to in both letters may reflect Constantius’ belief that the bishop was there but that the strained situation in Alexandria had prevented him from arriving. Thus, he was only sent when it was safe to do so. The letter to Aezanas and Sazanas is a remarkable document, a letter from one ancient Christian ruler to another. However, this is where the similarities end. The power differential between the two monarchs is evident from Constantius’ epistolary tone, which conveys clearly and forcefully his priorities for Roman influence in the Red Sea region, a territory which lay over a thousand miles outside Roman imperial territory.212 The letter opens with Constantius announcing his duty “to extend knowledge of the supreme God” to the “whole race of humankind”. He then considers how this duty applies to the people of Aksum: Since they are under the provident care of the empire, the doctrine taught in Aksumite churches should be the same as the one taught in Roman ones. The emphasis on the empire as an agent for religious change is not hard to miss in these opening sentiments, and Constantius’ concern with universalist expansion would strike us as remarkable were it not for the fact that, as it transpires in the letter, Christianity already had a presence in Aksum through the efforts of Frumentius, the principal bishop in the kingdom. Indeed, it is the pedigree of Frumentius which formed the basis for the communication. Offering Aezanas and Sazanas a history lesson, Constantius’ reminds the rulers that Frumentius was appointed to his present role by Athanasius, “a man guilty of ten thousand crimes”, and one who “has not been able to fairly clear himself of any of the charges brought against him”. Constantius fears that Frumentius, if unchecked in his character and teachings, will corrupt the Aksumites with impieties and blasphemies about God learnt from the condemned Athanasius. As such, he requests that the monarchs send Frumentius to Egypt for investigation by George, the bishop of Alexandria, and by “the others” who have jurisdiction in a case such as this. If it is found that he holds nothing contrary to “the laws of the Church and the established faith”, he will be appointed bishop by those sitting on judgement at his trial, a clear sign that Constantius held Frumentius’ appointment by Athanasius to be invalid. Bland sounding grounds as the basis for measuring Frumentius’ moral and doctrinal rectitude were a central feature of imperial bureaucratese, and as such it is necessary to dig a little deeper.

Who was Frumentius? A romantic story in the Nicene author Rufinus’ Ecclesiastical History (10.9–10) situates Frumentius’ youth within a narrative of a dispossessed childhood, an account that is ultimately intended to highlight the paternal nature of his relationship with Athanasius. As a Roman boy in “Further India” (Rufinus’ term for Aksum), God commanded Frumentius to build places of assembly in the kingdom for Roman merchants who were Christians (10.9). On returning to the eastern empire (his relative, Meropius, is said to have been from Tyre) Frumentius travelled to Alexandria to inform the bishop (Alexander?) about God’s achievements in India. It is Athanasius who argues for Frumentius’ appointment and confers on him the priesthood (10.10).213 As a result, the Christian origins of Aksum in the fourth century are portrayed as impeccably Nicene, and Frumentius is said to have acted, from the moment of his return to the kingdom, as if “the signs of the apostles were worked by him” (10.10). A contrary state of affairs is presented by Constantius’ letter where it is Frumentius’ association with Athanasius that demands investigation by George, the bishop of Alexandria. George’s appointment as Athanasius’ replacement marked the end point of Constantius’ efforts in the 350s to unseat Athanasius from Alexandria, a process that began with his condemnation at the Council of Sirmium in 351,214 a predominantly eastern oriented affair that also condemned Photinus, formerly of Sirmium, and Marcellus of Ancyra, and issued a creed – the same as the Fourth Creed of Antioch (from 341) – and twenty-six anathemas215 which clearly took aim at the “of the same essence” heritage of Nicaea but which were otherwise not overtly subordinationist.216 It is worthwhile pausing the discussion of Theophilus, Athanasius and Frumentius for a short time in order to provide some additional background for events in the Red Sea during 356/357 and the episcopacy of George in Alexandria.

The details for what eventually became Athanasius’ third exile – his flight from Alexandria took place on 8 February 356 – are very much connected to the broader political events of the time, not least the downfall of Magnentius, to whom it is suspected rightly that Athanasius made overtures, and the ascent of Constantius to the position of sole Augustus. Freed from the constraints put in his way by Constans during the 340s (see Chapter 5), Constantius could pursue Athanasius’ deposition from Alexandria in a more systematic manner than had previously been the case, notably in the run-up to, during and after the Councils of Arles (353/354) and Milan (355). Constantius’ weapon(s) of choice against Athanasius in the period between 351 and 356 has been a matter of considerable debate. Hanson argued that most western bishops had no great loyalty to Athanasius, and as such would not have found it a difficult task to “disown Athanasius”.217 Since so many of them did find it difficult to condemn the bishop of Alexandria in line with the decision of the Council of Sirmium, it is presumed that clergy were also being required to subscribe to a doctrinal formula “which was uncongenial to them”.218 As Barnes notes, there is

direct evidence [Athanasius, On his Flight 4.2 and History of the Arians 31.3–6; Liberius, Letter to Eusebius of Vercelli 1.1.2; Sulpicius Severus, Chronicle 2.39] [which] makes it clear that there was an imperial edict requiring all bishops to accept the decisions of a council of eastern bishops which contained both a condemnation of Athanasius and a creed.219

Barnes develops the highly plausible thesis that the doctrinal ruling to which Constantius required Nicene-leaning bishops to subscribe to was the synodal letter published at Sirmium in 351 (see above).220 Athanasius himself offers a suitably jaundiced account of Constantius’ actions in the wake of synods of Arles and Milan in his History of the Arians (31.1–6), a relatively late work (357?), in which, compared to the discussion of the same period in his Letter to the Bishops of Egypt and Libya (from 356 following his ousting from Alexandria by Syrianus) where his public support for Constantius is still evident (e.g. Ad episcopos Aegypti 1.5), his patience with Constantius had finally evaporated. In the History, the persecution of Athanasius was a “multi-agency” affair which was predictably only about him! Orders were sent to Strategius Musonianus as Praetorian Prefect of the East to remove the grain dole from Athanasius and his associates and give it instead to those who supported the teachings of Arius. Imperial agents – notarii and palatini – accompanied by Ursacius and Valens were despatched to cities to obtain the signatures of bishops condemning Athanasius. Those who refused were to be exiled. Laity who supported Athanasius were subject to intimidation and violence. In light of the ongoing difficulties facing the imperial administration with regard to participatory involvement in city and town councils across the empire (see Chapter 5), it is very telling that Athanasius also notes that decurions “in each city” (31.6) were also threatened with fines should they not be able to convince their own bishops to subscribe to the imperial orders.

Clearly, this is a compressed account of the events connected with the Councils of Arles and Milan, although its emphasis on the role of imperial and episcopal agents in enforcing the “imperial will” to subscribe to Athanasius’ condemnation, and the ongoing violence which had become a central feature of the divided Christian community following the Council of Serdica in 343, are a genuine reflection of the known details of the years from 353 to 355.221 The Council of Milan, which met during the high summer months of 355, attempted to break the western Church’s recalcitrance led by Liberius, the bishop of Rome, to malign and denounce Athanasius. Although Liberius did not attend the meeting, his companions and supporters maintained their opposition to Athanasius’ condemnation and were instrumental in conflating Athanasius’ status with the creed of Nicaea. Thus, the infamous story related by Hilary of Poitiers (ed. Feder 1916, CSEL 65.186.19–187.19) about the episode at the Council when Eusebius, the bishop of Vercelli, was brought into the meeting to subscribe against Athanasius. He responded by turning the tables against that eastern bishops by stating that all those assembled should first subscribe to the Nicene Creed as the measure of their faith, since “certain of those present had been found by him to be stained with heresy”.222 Dionysius of Milan was the first to offer his signature to the Creed but was stopped from doing so by Valens of Mursa “roughly wrenching pen and paper from his hands, crying out that no business could happen from then on”.223 Eusebius, Dionysius and Lucifer, the bishop of Cagliari, all refused to assent to the Council’s demands and were exiled to relatively remote locations where, in the case of Eusebius, he was first forced to reside Scythopolis in Palestine under the watchful eye of a vehement anti-Athanasian bishop, Patrophilus.224 As Stevenson notes: “These exiles, all implicated in Constantius’ central goal of banishing Athanasius, would signal the first intensive phase of Constantius’ strategy for exiling troublesome bishops”.225 Liberius’ exile to Beroea in Thrace followed soon afterwards, certainly in the period immediately following the Council of Milan,226 and his heroic defence of Nicene orthodoxy before Constantius and his court was imagined in the fifth century by Theodoret (Church History 2.13).227

However, it appears that the more the pressure grew on Athanasius, the more ambitious he became and the greater his popularity grew in Alexandria. His pastoral and literary work-rates during the first half of the 350s suggest that he knew he was a man living on borrowed time, a fact which can also be seen in his supervision of Frumentius’ episcopacy during this period. Athanasius’ initial contact with Frumentius is likely to have been soon after his own rise to the episcopacy in 328 when Frumentius was in Alexandria. Aezenas’ conversion from the worship of Mahrem (the Ge‘ez appellation for Ares)228 to Christianity could have occurred in the early 330s, and Frumentius could very well have played a major role in it.229 The negus Aezenas’ own religious biography is traceable through a series of inscriptions in Ethiopic (Ge‘ez) and Greek.230 While much uncertainty remains about the date, the agents and the mechanisms for the conversion of the Aksumite monarchy to Roman Christianity,231 it is clear that by the time of Constantius II’s intervention into the influence of Athanasius on Frumentius in 356/357, Frumentius was regarded as an extension of Athanasius’ pernicious character and teachings. His “re-education” in Alexandria was thereby urgently required. The Greek inscriptions recording the martial achievements of Aezenas from the Enda Semon site of Aksum, found close to the remains of a royal palace in the 1960s, reflect the embedded Trinitarianism, ergo anti-Arianism, of Aksum’s Christianity in the fourth century. Lines 1–2 of the stele invoke three distinct members of the Trinity, “by faith in God and by the power of the Father, Son and the Holy Spirit”; and, in line 28 with Aezenas’ claim to have risen up against his enemy, the Nōba, through his belief “in the power of the God Christ” (en tē dunami tou theou christou), we see a striking expression whereby the two nouns serve as syntactically parallel terms (both in the genitive case). The date of the inscription from this stele is moot, but it is clear that the formulae reflect a theology that has been updated in light of the reception of Nicaea and its creed during the 340s and 350s, but which was also beginning to be portrayed as a caricature of Nicaea by opponents (notably in the list of anathemas produced by delegates at the Council of Sirmium in 351, among which was the condemnation of those who hold that the Father and Son are two gods232). This, in turn, strongly implies that Athanasius had kept in communication with his protégé in Aksum during this time.

Plans to end Athanasius’ time in Alexandria had been fermenting for many years, although the imperial government only began to act once the bishop’s network of supporters in the west had begun to dissipate. The Council of Milan saw to that. The initial advances on Alexandria were measured and very cautious. A notarius named Diogenes arrived first, followed by a general (dux Aegypti) named Syrianus.233 An exchange of epistolary etiquette was performed between Athanasius and officials (Defence before Constantius 22–25; History of the Arians 52.1) prior to an escalation of events by Syrianus, the notarius Hilarius,234 and the strategos of Alexandria, Gorgonius. Syrianus and troops drawn from Egypt and Libya attacked the church of St. Theonas235 on the night of the 8/9 February (a Friday) during the midnight service.236 A remarkable document, an appeal (Gk. diamarturia) which formed part of a petition addressed to the Prefect of Egypt, a certain Maximus,237 and sent three days after the assault, details the incident from the point of view of the Nicene community (History of the Arians 81.1–14); the petition is especially fascinating since as an earlier document the complicity of Constantius in Syrianus’ actions is not assumed, in contrast to Athanasius’ charge that the emperor gave his approval for the assault (History of the Arians 48.2). Syrianus entered the church at midnight during prayer, breaking down the doors with “legions of soldiers”, all carrying weapons and with helmets on their heads (81.6). Attendees were attacked and some killed (81.7). Athanasius himself was dragged from the episcopal throne and beaten (81.8). Looting of church and private property also took place (81.10). Athanasius himself was spirited away to safety (Defence before Constantius 25): His whereabouts remained unknown from the point of view of the petitioners writing in the immediate aftermath of the attack.

In spite of this show of intimation and brutality, there was considerable public outrage and resistance to the acts of Syrianus, as the appeal addressed to Maximus highlights. Catholic clergy and laity in fact continued to occupy the churches for nearly four months before they were seized by the new Prefect of Egypt, Cataphronius238 and the comes Heraclius who arrived in Alexandria on 10 June 356.239 The churches were very soon (only four days after their arrival, according to the chronologically precise Historia Acephala240 (2.2)) transferred to the religious partisans of the new bishop of Alexandria named George, a former military contractor (History of the Arians 75.1)241 who derived originally from Cappadocia242 and was an established supporter of the homoian cause, who arrived in the city on 24 February 357, over a year after the assault on the church of St. Theonas. The interregnum between the departure of Athanasius and the arrival of George witnessed some remarkable episodes of manipulation of social relations in Alexandria spearheaded by Heraclius, Cataphronius, Faustinus and Bithynus (History of the Arians 55.2; Faustinus occupied the role of katholikos, otherwise the rationalis rei privatae, the chief financial official in Egypt243). The resistance to Syrianus’ initial attempt to seize Athanasius’ centres of support probably demonstrated to Egyptian officials the need to conduct the dissipation of the bishop’s influence with greater subterfuge. As a result, Heraclius drove compliance across the city by targeting key parties with orders prescribed in a letter from the emperor, a text which had likely been prompted by a petition from the same officials to address the complaints made in the catholic appeal from 12 February 356 (see above). The parties identified in the emperor’s ruling comprised the populace at large whose bread dole was threatened with suspension, the city’s decurions who were threatened with imprisonment and the city’s pagans (Gk. Ellēnes) whose rights of worship were to be curtailed, sanctions to be applied to all parties should they fail to subscribe to Constantius’ orders as relayed by Heraclius (History of the Arians 54.1–4). The demands in question reinforced the banishment of Athanasius and mandated the transfer of the city’s (Nicene) churches “to the Arians”. Forced compliance had the desired effect; the parties subscribed to the ruling, including members of the city’s trade guilds (ergasia). The question in this regard is whether the desired outcome included the sparking of more mob violence this time directed against the Great Church in the city’s Caesarion complex (History of the Arians 55.2) which occurred very soon after the arrival of the new Prefect and comes in June 356.244 At the end of a service, pagans attacked worshippers departing from the Great Church, subjecting the parthenoi to especially vile physical and verbal abuse. The episodes of violence “had clear religious overtones”245 and were certainly stylised in order to reflect Alexandria’s historic pagan heritage. Church property, including the altar and curtains, were removed, sprinkled with incense and burnt. The sacrifice of a bull used for irrigating the Caesarion was attempted in imitation of the worship of Serapis, the city’s tutelary god.246

Athanasius’ account portrays the mob as acting under the charge of the recently arrived administrative cabal (56.1), who were, in turn, following the orders of the emperor. In the background lay the Arians of Alexandria who – in the words of Athanasius – “went along with the Hellenes … because they thought it would harm and injure us” (57.1). Popular slogans filled the air: “Constantius has become a Hellene and the Arians have given their approval to our beliefs” (56.2). One recent commentator has seen the events of June 356 as a riot prompted by the flight of Athanasius from the city: The absence of a strong episcopal presence in Alexandria led to the rising of long-standing inter-communal grievances.247 The pagans had grounds to be incensed as the Great Church now dominated a previously revered pagan complex but they may also have been aware that the emperor was beginning a fresh effort to undermine the viability of pagan worship by striking at the legality of sacrifice and image-worship (notably, Cod. Theod. 16.10.6).248 It has been proposed that in this incident, we have an example of a “Church and King” riot,

a violent demonstration in support of traditional political and religious usages that are perceived as being threatened … these riots are often carried out in the name of a traditional political authority and with the tacit backing of local administrators.249

Thus, while the principal aim of the officials who arrived in Alexandria in June 356 was the dismantling of Athanasius’ network of supporters, it is also apparent that their mission was conducted with a fair degree of opportunism in terms of mobilising historic grievances in order to realise their goals.

The imperial party which preceded George’s arrival was highly successful in exploiting the city’s historic religious tensions and undermining the Nicene cause both within and beyond Alexandria. They appear in fact to have been more effective than George who ultimately proved incapable of marshalling the competing forces in the city to the advantage of the government. His ability to cause offence to the pagans of Alexandria was legendary: One incident when George passed by the fabled Tychaeon has him proclaim, “How long will this tomb remain standing?”250 On 29 August 358, he was assaulted in the church of St. Dionysius “by the common people” and fled the city, to return two and a half years later.251 In the interim, the notarius Paul arrived in June 359 with an imperial order in support of George, and continue the work of coercing the Christians of the city into complying with the imperial will.252 Following his return, and after news of the death of Constantius reached the city in late November 361, George was imprisoned and subsequently lynched on 24 December by “all the people” (Ammianus, 22.11.8) – such was his universal lack of appeal. However, the party driving the rebellion against George was almost certainly headed by the city’s pagans and arose as a result of his assaults on the pagan populace and pagan centres of worship, including the Serapeum.253 The crowd that dragged George from prison254 proceeded to torture him along with two associates; one of whom (Dracontius) had desecrated the sacred altar of the Alexandrian mint for which he was responsible, and the other (Diodorus: “possibly comes et architectus”, PLRE 1: 255 (Diodorus 2)) had assaulted some boys by cutting off their curls in the belief that they were adherents of a pagan god,255 after which their lifeless bodies were burnt and their ashes thrown into the sea in order to prevent their remains being transformed into relics of martyrs killed for a religious cause.256

For a brief period in late summer 358 during the initial stage of George’s absence, Athanasius’ supporters reoccupied the churches for just over two months having driven out “the men of George”, which could in all likelihood refer to the comes Heraclius et al.257 The repression of Athanasius’ supporters was ably conducted by another dux Aegypti – a military commander in charge of troops in Egypt, the Thebaid and Upper and Lower Libya258 – named Sebastianus who proved to be successful in breaking up Nicene networks in Alexandria and across Egypt. During the years 357–358, Sebastianus wrote letters to the military commanders (praepositi) of Egypt (History of the Arians 72.1) with the aim of unseating Nicene bishops across the region. Therefore, the exiles which followed were initiated and overseen by one of the most senior military commanders in the region and, as we have seen, the involvement of the army was a characteristic feature of this phase of events in Alexandria. The authors of the appeal from 12 February 356 even sought to characterise Syrianus’ assault on the church of St. Theonas as “a war (polemos) against the churches” (History of the Arians 81.5). Intimidation was likely at the forefront of the minds of the imperial authorities when the army was deployed against clerics and laity in the manner described by Athanasius and others. For instance, Sebastianus had been the lead figure during the events of the Sunday of Pentecost in 357,259 when a large contingent of troops (three thousand is claimed) was sent to disperse Christians praying close to a cemetery (Defence before Constantius 27; Defence of His Flight 6).260 However, there were also clear security anxieties about the possibility of civil unrest whenever plans were made to unseat a cleric with significant popular appeal.261 This was evidently the case with those bishops exiled during the purge of Egyptian sees conducted by Sebastianus during 357/358. The quarries for this initiative were highly significant: They were bishops who owed their positions to bishop Alexander (d. 17 April 328), the original target of Arius’ criticisms in 318,262 or to the early years of Athanasius’ episcopate. Athanasius made known the names of those banished through Sebastianus’ efforts (History of the Arians 72.2–4), and we can cross-reference them against documents from key stages of the factional conflict.263 Ammonius, Anagamphus, Muis, Psenosiris and Neilammon were all signatories to the encyclical of the western Council of Serdica in 343.264 Some were also signatories of Athanasius’ Tome to the Antiochenes, composed in 362 following the bishop’s brief return to Alexandria between February and October of that year. The letter to the Antiochene church records the doctrinal decisions reached at the synod of Alexandria which had been called following Athanasius’ return to the city.265 Among its signatories were Marcus of Philae and Marcus of Zygra, two bishops who had been exiled in the purges of 357/358 but who had been permitted to return to their sees following Julian’s cancellation of exilic sentences from Constantius’ reign (e.g. Julian to Aetius, Letter 15 (Wright 1923)).266 Some were also mentioned by Athanasius in his Festal Letter from 347 (Letter 19), for example, Neilammon who is identified as the bishop of Syene (he shared the name with his predecessor) and Psenosiris whose see was the important trading zone of Koptos (see above). All were elderly and Athanasius indicates that their mistreatment as they were forced into exile was intended to exploit their infirmity and hasten their demise. The exilic status of many of the names is corroborated by Athanasius’ On the Defence of His Flight (7.4), including Dracontius of Hermupolis Parva, exiled to the deserts around Clysma at the head of the Hieroopoliticus Sinus (the northern mouth of the Suez); Philo, one of the two bishops with that name who signed the western encyclical of Serdica (Apology against the Arians 49.3), was exiled to the Babylon Fortress on the Nile Delta; Adelphius, bishop of Onuphis in Lychni, was exiled to Psinabla in the Thebaid; and the presbyters Hierax and Dioscorus from the Mareotis were banished to Syene.267 Both Dracontius and Philo were visited by Hilarion, the Palestinian monastic who was the subject of a Life composed (c. 391) by Jerome of Stridon. The Life of Hilarion (30) records that Dracontius was confined at a fortress (castrum) called Thaubastus, and Philo as noted at the Fortress Babylon. Adelphius’ banishment to Psinabla was also to a fortified location. This form of “fortress banishment”, so styled by one recent commentator, was undoubtedly linked to the central role which the dux Aegypti played in the affair: “One may suspect, therefore, that these three bishops were singled out as in need of closer control than the other dissidents on which banishment Athanasius reported on the same occasion and who were mostly simply expelled from cities”.268

Other bishops sent into exile at this point, also ordained during the early Nicene phase of the Alexandrian episcopate, included Ammonius, Agathus, Agathodaimon, Apollonius, Eulogius, Apollo, Paphnutius, Gaius and Flavius (“very old bishops”), Dioscorus, another Ammonius, Heracleides and Psais. In total, more than thirty bishops (On the Defence of his Flight 7.4) were deposed in this period.

A number of features stand out from the period of Sebastianus’ involvement: The pragmatic extension of authority for sentences of clerical exile to the military figure, the dux Aegypti; the number of bishops removed from their sees and the locations of some of those exiled (for instance, the so-called “fortress banishment”); and the systematic effort to demolish the influence of Alexander and Athanasius on the Egyptian episcopal scene. Returning again to the case of Frumentius, Constantius’ demand that he be sent to George should therefore be seen in the context of the campaign against Nicene-Athanasian influences in Alexandria and across Egypt between 356 and 361. In the letter to the Alexandrians from 356 to 357 (Defence 30), Constantius portrays Athanasius as a violent popularist whose lack of regard for the city’s rich intellectual heritage was a direct threat to the Alexandria’s standing in the empire. A different approach is taken by contrast in the letter to Aezanas and Sazanas (Defence 31). Here, Constantius’ will is explicitly stated: He desires to protect the Aksumites from the corrupting influence of “impious statements” deriving from Athanasius and to promote a united religious settlement whereby the Christianity of the Aksumites corresponds to the Christianity of the empire. Diplomatic coercion therefore replaced legislative and military coercion. The concerns in both texts can be understood in a variety of ways: The need to engender social cohesiveness across a divided city, a desire for religious rectitude and the realisation of imperial unity. Therefore, Constantius’ management of affairs in Alexandria, Egypt and the Red Sea was “most likely formed from a complex combination of motives that reflected the world he inhabited”.269 These motives comprised the above desiderata, but were also informed by knowledge of Sasanian plans for their western frontier, and the need to secure and enhance access to the main routes for trade and commerce down the Red Sea and out into the Indian Ocean. However, in fundamental terms, it was and had always been about who exercised power and influence in one of the empire’s most strategically important areas. In this regard, Stevenson should have the final say:

[Constantius] had witnessed with alarm Athanasius building an ecclesiastic empire south into the Thebaid, down the Nile, and also southeast down the Red Sea. In fact, judging by Athanasius’ successor Theophilus’ [Patriarch of Alexandria, 385–412] building program, the bishop of Alexandria enjoyed considerably more income than the average bishop, money that would pour in from diocesan contributions as well as extensive ecclesiastic land holdings. And it was Athanasius himself who had created, or more accurately chanced upon, an extra-imperial link to the Aksumites, the very link that Constantius used to exert his power in the letter to their kings demanding that the Aksumite bishop Frumentius return to Alexandria for reeducation.270


1. Notably Barnes 1993: 102; for a more cautious analysis, see Didu 1977.

2. See the survey by Didu 1977. On the importance of milestones and the issue of imperial loyalty, see Sauer 2014.

3. So, Didu 1977: 39–40.

4. As related by Ammianus (14.5.6), the allegation against Magnentius’ supporters was that they “dared to conspire” with Magnentius against Constantius.

5. Firmicus Maternus, On the Error 28.6; see Woudhuysen 2018: 176–177. On the possible explanation for Constans’ expedition of 343, see Mattingly 2006: 234–235.

6. Amm. Marc. 14.5.8.

7. PLRE 1: 683–684 (Paulus 4).

8. On chronology, see Barnes 1989: 419.

9. Amm. Marc. 14.5.6.

10. Identified as Vulcacius Rufinus, PLRE 1: 783; Barnes 1992a: 256; on the date of the constitution addressed to Rufinus (Cod. Theod. 9.23.1), see Cuneo 1997: 233–235.

11. Amm. Marc. 14.5.8–9.

12. Julian, Or. 1: 41c–d; Didu 1977: 37.

13. Didu 1977: 35–36.

14. Notably CIL 8.2252; CIL 8 22558; AE 1933.105; CIL 8.7012; CIL 8.7013; Didu 1977: 42.

15. Sauer 2014: 261–263.

16. Sauer 2014: 296.

17. CIL 2.4791.

18. See esp. Birley 1958.

19. For example, RIB 2308 and 2314, also 930.

20. See esp. Fagan 2011: 471–476.

21. Amm. Marc. 14.5.9.

22. Webster 1983: 245.

23. Discussed by Webster 1983: 247–250.

24. Cf. the more cautious assessment by de la Bédoyère 2013: 74.

25. Cons. Const. s.a. 351.

26. RIC 8 Siscia 372, nos. 318, 319 (plate 16; and commentary, pp. 345–346); Šašel 1971: 213.

27. Trans. Ridley 2017.

28. See DiMaio 1988: 244–247.

29. Cf. Barnes 1993: 105, who places Constantius in Sirmium, awaiting “the outcome in safety”.

30. Julian, Or. 2.71c. See the discussion by Šašel 1971: 213–214.

31. Humphries 2020.

32. Humphries 2020: 174–175. On the relationship between the accounts of these two authors and their likely sources, see esp. DiMaio 1988: 245–247.

33. Cf. Šašel 1971: 211.

34. Mateos, Pizzo and Ventura 2017.

35. Jerome, Chron. s.a. 350.

36. See Hanson 1988: 313; Barnes 1993: 97.

37. Trans. Roberts 1894.

38. Socrates, Hist. eccl. 2.32.2–5; trans. Zenos 1890. Cf. Sozomen, Hist. eccl. 4.7.1–2; see the comments by Leppin 2015: 201–203.

39. Julian, Or. 1.38b–d; Maraval 2013: 111.

40. Julian, Or. 2.71c–d. The date is proposed by Šašel 1971: 214.

41. Julian, Or. 1.39c–d; Maraval 2013: 112.

42. See esp. the comments by Humphries 2020: 170–176.

43. CIL 6.1158 (ILS 731); for trans. and commentary LSA 838 (Carlos Machado); see also Machado 2019: 102–103 for a discussion of the broader context of honorific dedications in the Forum.

44. Cod. Theod. 15.14.5; Cuneo 1997: 206.

45. See Holt 2003: 72.

46. Notably by Sozomen, Hist. eccl. 4.7.3, and Socrates, Hist. eccl. 2.32; on these sources, see DiMaio 1988: 246–247.

47. Julian, Or. 1.40a–b.

48. Zonaras 13.9.

49. DiMaio 1998: 246–247; PLRE 1: 249–250 (Desiderius I).

50. See the commentary by Bleckmann and Gross 2018: 309.

51. Eutropius, 10.12.2.

52. Cf. Holt 2003: 70–71.

53. Holt 2006; also, Holt 2003.

54. Drinkwater 2007: 213.

55. RIC 8 Trier 164–165, nos. 329–331 (plate 3).

56. Holt 2003: 72.

57. Holt 2003: 63–67.

58. RIC 8 Trier 165, nos. 332–337 (plate 3).

59. Holt 2003: 68. See Cuneo 1997: 233–235 for a discussion of the law.

60. Cuneo 1997: 220.

61. Trans. Wright 1913: 255.

62. Humphries 2015: 163.

63. CIL 1166 = LSA 1281; CIL 1167 = LSA 1284; Chastagnol 1962: 110–111.

64. Unlike Chastagnol 1962: 111, and Maraval 2013: 115, I find no indication in Julian’s Or. 2 that Titianus’ property was confiscated by Constantius.

65. Amm. Marc. 14.5.1; cf. Kienast 2017: 301 who corrects Ammianus’ date from October to November, and also Barnes 1993: 314, nt. 32. (“Perhaps Octobres in the text of Ammianus should be emended to Novembres”.)

66. Amm. Marc. 14.5.1–9.

67. Leppin 2015: 203.

68. Amm. Marc. 14.5.1; PLRE I: 393 (Gerontius 1).

69. Leppin 2015: 203.

70. Cf. Maraval 2013: 116 who counts Justus as among the victims of the period.

71. Drijvers 2015: 482.

72. Amm. Marc. 30.7.2–3; PLRE 1: 400–401 (Gratianus 1). See also the commentary by Matthews 2007: 75, and esp. Drijvers 2015.

73. CIL 8.7014=ILS 758; see the commentary by Drijvers 2015: 486 (nt. 47).

74. Themistius, Or. 6.81d (PLRE 1: 401); also Drijvers 2015: 486.

75. See esp. the brilliant study by Cramer 2011 (esp. pp. 15–86).

76. See esp. Ross 2016a: 52–95.

77. Barnes 1998: 6.

78. PLRE 1: 262 (Domitianus 3).

79. PLRE 1: 535–536 (Montius Magnus 11).

80. Cf. Barnes 1989: 421.

81. Thus, Cod. Theod. 10.21.3, from 16 January 424 to the Count of the Sacred Largesse by Theodosius II; see the excellent analysis of purple production by Hall 2004: 229–234.

82. See esp. Woods 1993 on the importance of this detail in the Passion.

83. See Harries 1988.

84. Suspected by Thompson 1947: 64–65.

85. Jones 1964: 834–835.

86. Amm. Marc. 14.9.3.

87. See Thompson 1947: 60–61; Cf. Matthews 2007: 406–409.

88. PLRE 1: 907 (Theophilus 1).

89. PLRE 1: 287–288 (Eubulus 2).

90. On the campaign of spring 354, see esp. Drinkwater 2007: 204–205.

91. See Moser 2018: 76.

92. Julian, Letter to the Athenians 272d, 274a.

93. Drinkwater 2007: 205.

94. Amm. Marc. 18.4.3; on Eusebius, see esp. Tougher 2008: 36–37, 2021: 83–89.

95. PLRE 1: 751.

96. Drinkwater 2000: 153.

97. Julian, Letter to the Athenians 272d; see Chausson 2007: 114–115.

98. Chausson 2007: 138–141.

99. On the inscriptions, see Machado 2011; Machado 2019: 190–191.

100. Hillner 2017: 87.

101. PLRE 1: 687 (Pentadius 2).

102. PLRE 1: 539.

103. PLRE 1: 503 (Leontius 22).

104. PLRE 1: 517–518 (Lucillianus 3).

105. PLRE 1: 145 (Bainobaudes 1).

106. PLRE 1: 146–147.

107. PLRE 1: 82 (Apodemius 1).

108. Kristensen 2015: 337.

109. To be distinguished from Theophilus, the Consular Governor of Syria (consularis Syriae) in 354, PLRE 1 907. Hall 2018: 81, nt. 21, confuses the governor Theophilus with Theophilus “the Indian”.

110. Crawford 2016: 90.

111. Cf. PLRE 1: 633.

112. Dodgeon and Lieu 1991: 388, nt. 4.

113. Cf. Dodgeon and Lieu 1991: 338, nt. 4.

114. Barnes 1989: 421.

115. Crawford 2016: 89 suggests that there was a connection.

116. Barnes 1989: 418–419.

117. Hopwood 1999: 225–226. See also the overview of Isauria by Matthews 2007: 355–367.

118. Cf. Matthews 2007: 362–363.

119. Hopwood 1999: 227.

120. Matthews 2007: 363.

121. See the insightful comments by Fisher 2011: 1–3 on the terms “Saracen” and “Arab” in ancient historiography.

122. Shahîd 1984: 74–86.

123. See Shahîd 1984: 330–334.

124. Shahîd 1984: 76.

125. Parvis 2006: 133.

126. Cf. Shahîd 1984: 78.

127. Shahîd 1984: 78.

128. For a critical revision of Shahîd’s approach, see the comments by Fisher 2011: 10–14.

129. Matthews 2007: 361.

130. Shahîd 1984: 79.

131. It is assumed to have been covered by Ammianus in his lost thirteenth book, see esp. Geiger 1979a; cf. Matthews 2007: 484, nt. 2.

132. Aurelius Victor, On the Caesars 42.11; cf. Soc. Hist. eccl. 2.33; Soz. Hist. eccl. 4.7.

133. So, Jerome, Chron. s.a. 352.

134. See the helpful summary by Bijovsky 2007: 193–194, 196–197.

135. See Russell 1980.

136. See Neville 2018: 162–168.

137. Geiger 1979b: 250.

138. So, Geiger 1979b: 255.

139. Geiger 1979b: 255.

140. Geiger 1979b: 253.

141. Geiger 1979b: 254.

142. Geiger 1979b: 255.

143. For a complete list of references to Ursicinus in rabbinic sources, see Geiger 1979a.

144. Geiger 1979b: 256.

145. Barnes 1993: 316, nt. 55.

146. Irshai 2009: 403.

147. Irshai 2009: 407.

148. Irshai 2009: 406.

149. On this notorious constitution, see Cuneo 1997: 50–53.

150. See esp. Schwartz 2004: 203–214; and the comments by Bijovsky 2004: 193.

151. The argument made by Geiger 1979a.

152. Irshai 2009: 409.

153. Cf. Vaggione 2000: 158.

154. Following the reconstruction by Hanson 1988: 599–600.

155. See Hanson 1988: 41–43.

156. Vaggione 2000: 167–168.

157. See esp. the discussion in Vaggione 2000: 161–179.

158. See esp. Shepardson 2014: 12–19.

159. Parvis 2014: 50.

160. Notably, Heil 2014.

161. See the careful explication between homoian and neo-Arian (anhomoian) theology by Hanson 1988: 557–572.

162. A complex question with no straightforward answer, although neatly summarised by Tougher 2007: 54–55.

163. Stevenson 2021: 20–21, 34, nt.6. Cf. Fiaccadori 1984: 314–328, who proposes the Maldives as the location of Theophilus’ homeland.

164. See Hall 2018: 81.

165. For an insightful overview of the Antiochene church during the fourth century, see Shepardson 2014: 1–29. Cf., Vaggione 2000: 18, nt. 32.

166. Cf. Lieu 1989: 46–47.

167. Schatkin and Harkins 1985: 30–31.

168. See esp. Digeser 2004.

169. See esp. the discussion by Shepardson 2014: 58–91.

170. For example, Rufinus, Eccl. Hist. 10.36; Philostorgius, Hist. Eccl. 7.8; Soc., Hist. Eccl. 3.18; Soz. Hist. Eccl. 5.19; Theodoret, Hist. Eccl. 3.6.

171. See den Boeft, Drijvers, den Hengst and Teitler 1995: 232.

172. On Ammianus’ explanation for the fire, see den Boeft, Drijvers, den Hengst and Teitler 1995: 228–233.

173. Trans. Cameron and Hall 1999: 112.

174. Digeser 2004: 76.

175. Elm 2012: 277–281.

176. Shepardson 2014: 80.

177. Trans. Schatkin and Harkins 1985: 113–114.

178. Brennecke 1988: 137.

179. See the note by Stevenson 2021: 34, nt. 2, regarding Theophilus’ career beyond the downfall of Gallus following the murders of Montius and Domitianus.

180. Elm 2012: 241.

181. Elm 2012: 243. See also the comments by Ferguson 2005: 152–162.

182. See the comments by Elm 2012: 242–243.

183. See Zellentin 2018.

184. Amidon 2007: 41, nt. 9.

185. See Shahîd 1984: 88.

186. Hoyland 2001: 146–150; Bowersock 2013: 78–91.

187. Cf. Shahîd 1984: 89, nt 59

(… Qāna does not answer to the description of Philostorgius as an emporium close to the mouth of the Persian Gulf since it is situated far from it. The exact location of this emporium is difficult to determine. What is more important is its description as being close to the mouth of the Persian Gulf, so far to the east.)

188. See the comments by Fiaccadori 1984: 290–292.

189. Sidebotham 2011: 261–262; Stevenson 2021: 22–23.

190. Sidebotham 2011: 278.

191. Shahîd 1984: 94.

192. Shahîd 1984: 95.

193. Fowden 1993: 103.

194. See Cuneo 1997: 286 for a summary of the controversy over the constitution’s date; and, Stevenson 2021: 86–88, for a more recent assessment proposing 357 rather than 356 for the law.

195. Stevenson 2021: 75.

196. Amm. Marc. 15.13.1–4 (16.9.2); interpretations about the characterisation of Strategius in the passage from bk. 15 vary considerably; see Drijvers 1996; Woods 2001; and more recently, Stevenson 2021: 73–78.

197. PLRE 1: 611–613.

198. PLRE 1: 213–214 (Clematius 2).

199. Stevenson 2021: 76–77.

200. Wiesehöfer 2018.

201. Stevenson 2021: 87–89.

202. Stevenson 2021: 78.

203. On the legendary and historic links between Aksum and Himyar, see esp. Bowersock 2013: 44–77.

204. Sidebotham 2011: 278.

205. So, Sidebotham 2011: 278; Stevenson 2021: 22–23.

206. See esp. Banaji 2016: 127–140.

207. Stevenson 2021: 33.

208. Barnes 1993: 196–197; also, Barnes 2007.

209. See Barnes 1993: 197 for the date of the revised version of the work; also, Stevenson 2021: 31.

210. Stevenson 2021: 30–31.

211. So, Haas 1997: 285.

212. Procopius, Wars I.19.27–33; Bowersock 2013: 71–72; also, Stevenson 2021: 26.

213. See the commentary by Amidon 1997: 47.

214. Outlined by Barnes 1993: 109–110.

215. See Lienhard 1999: 178–181.

216. See Kelly 1960: 281–282; Hanson 1988: 325–329; Barnes 1993: 109–110.

217. Hanson 1988: 330.

218. Hanson 1988: 330.

219. Barnes 1993: 273, nt. 9; cf. Hanson 1988: 330–331.

220. Although, cf. Robertson 1892: 222.

221. Outlined and narrated by Barnes 1993: 115–118.

222. Trans. Wickham 1997: 69.

223. Trans. Wickham 1997: 69. See Barnes 1993: 117–118.

224. See esp. Washburn 2009.

225. Stevenson 2021: 51; see also his wider discussion, 51–53.

226. Barnes 1993: 118; Stevenson 2021: 54.

227. On the dialogue between Constantius and Liberius, see esp. Stevenson 2021: 53–55.

228. Black 2008: 98–99.

229. Cf. the caution struck by Bowersock 2013: 72–73.

230. Detailed in Black 2008.

231. Cf. Black 2008: 95–96.

232. Hanson 1988: 326.

233. PLRE 1: 872.

234. PLRE 1: 434 (Hilarius 2; also Hilarius 6(?)).

235. See Kiss 2007.

236. Historia Acephala 1.10–11 (Martin 1985: 143–145).

237. PLRE 1: 582 (Maximus 13).

238. PLRE 1: 186 (Cataphronius 1).

239. PLRE 1: 418–419 (Heraclius 3). Historia Acephala 2.1–2 (Martin 1985: 145).

240. On this work, see Martin 1985: 11–67.

241. See Haas 1997: 286.

242. History of the Arians 75.1; cf. Amm. Marc. 22.11.4.

243. PLRE 1: 326–327 (Faustinus 2). On his role, see Jones 1964: 412–413; Kelly 2004: 202.

244. Historia Acephala 2.2 (Martin 1985: 145); also Martin 1985: 184 for the date of the episode at the Great Church.

245. Haas 1997: 284.

246. Haas 1997: 284.

247. Haas 1997: 283–286.

248. Cf. Cuneo 1997: 288; Maraval 2013: 218–223; Stenger 2020.

249. Haas 1997: 285.

250. Amm. Marc. 22.11.7; den Boeft, Drijvers, den Hengst and Teitler 1995: 206; see also p. 196 and discussion of the fact that Ammianus’ narrative about George’s last days is chronologically out of place in the events of book twenty-two.

251. Historia Acephala 2.6 (Martin 1985: 147).

252. Historia Acephala 2.5 (Martin 1985: 147); also see Martin 1985: 186–187, nt. 54.

253. Haas 1997: 287–288.

254. Historia Acephala 2.10 (Martin 1985: 149).

255. See the fascinating commentarial note by den Boeft, Drijvers, den Hengst and Teitler 1995: 209–210.

256. Ammianus Marcellinus 22.11.9–10; Historia Acephala 2.10 (Martin 1985: 149); see Martin 1985: 189, nt. 66; and for detailed commentary on the episode, see den Boeft, Drijvers, den Hengst and Teitler 1995: 208–212.

257. Historia Acephala 2.3–4 (Martin 1985: 147).

258. Jones 1964: 101; Palme 2007.

259. Regarding the date of this episode, see Martin 1985: 186, nt. 52.

260. On the involvement of Sebastianus, see esp. Martin 1985: 186, nt. 52.

261. See Hillner 2015: 221–232.

262. See esp. Hanson 1988: 129–145.

263. See the useful table compiled by Hillner 2015: 361.

264. Athanasius, Apology against the Arians 49.3; see Flower 2016: 101, nt. 259.

265. See Hanson 1988: 639–653.

266. See Teitler 2017: 37; 164, nt. 5.

267. See Flower 2016: 102, nt. 260.

268. Hillner 2015: 227.

269. Stevenson 2021: 91.

270. Stevenson 2021: 91.

If you find an error or have any questions, please email us at Thank you!