SOURCE: Hunt, E. D. “Christians and Christianity in Ammianus Marcellinus.” Classical Quarterly n.s. 35, no. 1 (1985): 186-200.

[In the following essay, Hunt discusses Ammianus's attitudes toward Christianity and Christians.]

Ammianus Marcellinus, by common consent the last great historian of Rome, rounds off his obituary notice of the emperor Constantius II (d. 361) with the following observation:

The plain simplicity of Christianity he obscured by an old woman's superstition; by intricate investigation instead of seriously trying to reconcile, he stirred up very many disputes, and as these spread widely he nourished them with arguments about words; with the result that crowds of bishops rushed hither and thither by means of public mounts on their way to synods (as they call them), and while he tried to make all their worship conform to his own will, he cut the sinews of the public transport service.¹

This is a perceptive judgement of the ecclesiastical politics of the reign of Constantius, remarkable in a pagan writer, and of exceptional significance in that it lies outside those very ‘arguments about words’ which contaminate all the Christian assessments of this emperor. Although Ammianus is unsympathetic to Constantius, he manages succinctly to grasp the basic drift of imperial policy, inherited from Constantine himself, of trying to enforce the emperor's view of doctrinal and ecclesiastical unity by the summoning of repeated episcopal councils and browbeating the bishops into agreement—thus paying lip-service to the independence of the church's judgements.² To the observant outsider, this process was notable above all for the burden it placed on the cursus publicus, as the bishops went about their business around the empire now provided with official evectiones; and Ammianus' comment finds confirmation in the letter issued by eastern bishops attending one of the many councils of Constantius' reign, that at Sardica in 343, who complained of the ‘attrition’ of the transport service caused by the imperial summons.³

Not only is it worthy of note that a non-Christian historian appears informed and observant about the ecclesiastical policies of a Christian emperor; the very fact that the topic is mentioned at all ought to occasion some surprise, when we reflect that Latin historical writers contemporary with Ammianus—Aurelius Victor, Eutropius, Festus—managed to write about Christian emperors without so much as a word about their Christianity. In including this paragraph on Constantius, and (to take another example) in mentioning the Christian conviction of the short-lived emperor Jovian, Ammianus is quite exceptional among his Latin counterparts.⁴ He also stands apart, in this respect, from contemporary historical writing in Greek. Here the principal witness is Eunapius, an author whose work is likely to have been known to Ammianus:⁵ it is clear from his surviving fragments and the dependent narrative of Zosimus that Eunapius did not shrink from comments about Christianity or the religious policy of Christian emperors, but his mentions were shot through with a polemical opposition to the new faith and a hostility to emperors, especially Constantine and Theodosius, on the grounds of their Christianity.⁶ It is evident that Ammianus' reference to Constantius' ecclesiastical interventions is very different from this. Not only is Christianity not the basis of his dislike of the emperor (so much is clear from the whole narrative of the books devoted to Constantius, and the mention of his Christianity comes only at the end of the list of his faults), but what is more it is difficult to read the phrase ‘religionem absolutam et simplicem’ as anything other than commendation: what was at fault in Ammianus' eyes was the ‘anilis superstitio’ with which Constantius perverted the essential merit of Christianity.⁷

In the light of passages like this paragraph on Constantius it used long ago to be asserted that Ammianus Marcellinus was in reality a Christian.⁸ Nowadays it is not possible to doubt his paganism: only a pagan could have written a history so pervaded by the religiosity of omens and fate, and one in which the pagan emperor Julian was the—albeit flawed—hero. Ensslin found in Ammianus a breadth of religious tolerance which he associated with the monotheistic tendency of late antique Neoplatonism, and this interpretation has found many followers;⁹ this same tradition of pagan tolerance, and acquiescence in a multiplicity of faiths, has recently been reaffirmed for Ammianus in a wide-ranging article by J. J. O'Donnell.¹⁰ Yet there has also been in recent years an attempt, despite appearances to the contrary, to see beneath the surface of Ammianus' narrative a submerged hostility to Christianity, and thus to identify him more closely with the kind of pagan polemic found in Eunapius. E. A. Thompson found frank expressions of paganism, such as the major digression on the practice of divination (21.1), but detected a change in the composition of the last six books when, he believed, Ammianus was obliged to muzzle his pagan commitment in deference to the attitudes prevailing at the Christian court of Theodosius I in the early 390s.¹¹ Others have been less ready to suppose that Ammianus' anti-Christian sentiments are suppressed: they see, for example, the contrast between the favourable judgements of leading pagans like Praetextatus or the elder Symmachus and the criticism of the Christian family of the Anicii as an indication of pagan emphasis in the history.¹² The conclusion of this line of argument places the pagan/Christian issue at the heart of the historian's purpose, and is prepared to see Ammianus' work as part and parcel of the last pagan reaction at Rome in the late fourth century.¹³

I do not propose in these pages to undertake a re-examination of Ammianus' general religious viewpoint. I am concerned, rather, with the more specific objective of identifying the attitude expressed in his history towards Christians and Christianity, a task which primarily entails considering in some detail those passages where the topic is actually discussed: we are likely to be on firmer ground arguing from what Ammianus says than speculating from what he does not say. Two classic articles of some twenty years ago have taught us that any examination of the contents of Ammianus' history must begin from the historiographical tradition in which he set himself: we must be clear what to expect and what not to expect in such a context.¹⁴ Self-consciously the classical Roman historian, Ammianus displayed a reticence in discussing contemporary Christian affairs because, whatever their importance for the history of the Roman empire in his day, they were no part of his literary heritage. Moreover, when Christian matters are mentioned, Ammianus may be expected (although he does not always do so) to deploy periphrastic formulae, such as the ‘ut appellant’ expression in the passage quoted about Constantius: this practice does not imply ignorance on his part concerning Christian institutions, but is his way of adhering to classical models¹⁵—Christians and Christianity are thus distanced from the appropriate subject matter of the Roman historian, admitted to the narrative only with an apologetic affectation.

It is surprising, then, just how much reference to Christianity we find in Ammianus' pages—in stark contrast, as we have seen, to others who were recording Roman history in Latin around the same time. Incidental to Ammianus' main narrative we find, for instance, Christian churches referred to as places of refuge for a fallen usurper (Silvanus) or a charioteer convicted of magical practices (Hilarinus); Christian clergy are identified conducting diplomatic negotiations between Roman commanders and enemies in various theatres of war—on the eastern frontier, in north Africa, in the Danube provinces.¹⁶ The celebration of a Christian festival (‘Christiani ritus … sollemnitatem’) at Mainz in 368 exposed the place to an Alamannic raid; while a corrupt court official of Valentinian (the tribunus et notarius Palladius) was able to evade conviction by committing suicide in the absence of his guards, who were ‘spending the night in church’ for a Christian celebration.¹⁷ Then there is the occasion at Vienne at the beginning of 361 when Julian very obviously participated in the festival ‘which the Christians observe in January and call Epiphany’: the usurping Augustus, to conceal his apostasy, solemnly entered the church and prayed to the numen.¹⁸ We may add to the list the group of Christian virgins mentioned by Ammianus as captured by the Persian monarch, but spared by him in a show of ‘lenitudo’ and allowed to pursue their religion unharmed.¹⁹

These are isolated notices of Christian individuals or institutions in the course of narrative about other matters; but there is some more extended coverage of Christian affairs where these impinge on the secular concerns of the Roman historian. Nowhere more so than in the discussions of the administration of the city of Rome, a topic which regularly punctuates Ammianus' narrative as he recounts the regimes of successive prefects of the capital in an (anachronistic) effort to keep urbs Roma at the centre of his history.²⁰ In mentioning the prefecture of Leontius (355-6), Ammianus presents his version of the summons to Constantius' court at Milan of Liberius, the bishop of Rome (‘Christianae legis antistes’), because of his refusal to subscribe to the condemnation of bishop Athanasius (a notorious instance of the kind of ecclesiastical meddling which brought down Ammianus' censure on the Christianity of Constantius).²¹ The episode doubtless qualified for inclusion because of the public outcry which accompanied Liberius' departure (the bishop, we are told, was so popular that he had to be escorted from Rome at dead of night ‘for fear of the people’): the prospect of disorder would have called for the prefect's attention, and probably figured in the official record of his prefecture which would have been Ammianus' source.²² The historian's treatment of this piece of Christian history has some interesting features. Ammianus is again shown to be informed about Christian organisation and procedure, and knows of Athanasius' condemnation and deposition by a synod of bishops (which was re-enacted several times, most recently at Milan in 355), referring to this in a particularly long-winded periphrasis—‘coetus in unum quaesitus eiusdem loci cultorum, synodus ut appellant’; yet the manner of Athanasius' introduction here makes it clear that he had not been mentioned earlier in one of the lost books—he only secures admission to the narrative at this point because of the secular repercussions at Rome. In Ammianus' eyes Athanasius' offence had consisted in getting above himself and meddling in affairs outside his ‘professio’: in particular, he was said to be skilled in the arts of foretelling the future. The bishop was accused of other crimes, too, which were ‘abhorrent to the teaching of which he had charge’ (‘a proposito legis abhorrentia cui praesidebat’). As with Constantius' perversion of the simple truths of Christianity, the tone here is hardly hostile, as Ammianus appears to acknowledge the basic merit of the ‘lex Christiana’, transgressed by the alleged crimes of Athanasius. The historian may have in mind the violent methods of quelling opposition in his diocese of which the bishop's critics had accused him from the earliest years of his episcopate.²³ The charge that Athanasius was expert at divination is also mentioned by Christian writers,²⁴ but it assumes a particular prominence in Ammianus' version, where the bishop is made to appear as another in the line of distinguished citizens in the empire who, by their dabbling in such pursuits, fell foul of the pathologically suspicious Constantius: shortly before, for instance, Ammianus has told of a Pannonian governor hauled before the emperor because of dangerous talk about the future overheard in his company.²⁵ The key point is that in such a context Athanasius' offence is secularised. Of the theological disagreements and arguments about ecclesiastical authority which were integral to the saga of Athanasius Ammianus has no hint: whether he knew much about them or not is less important than the fact that they had no place in the kind of Roman history on which he was engaged.

Similarly Liberius' resistance to the emperor assumes in Ammianus the character of plain insubordination: the bishop of Rome was perceived as ‘defying the orders of the emperor’ (‘tamquam imperatoriis iussis … obsistens’, ‘aperte scilicet recalcitrans imperatoris arbitrio’), a highly placed subject resisting the will of his sovereign. There is no suggestion of an argument about the emperor's right to coerce bishops, nor I believe is the historian making any point about the ecclesiastical standing of the bishop of Rome. True, he does say in a famous phrase that Constantius was anxious for Athanasius' deposition to be confirmed by ‘the more powerful influence of the bishop of the eternal city’ (‘auctoritate quoque potiore aeternae urbis episcopi’), a statement which has been read as an important observation by an outsider of the growing power of the papacy.²⁶ It is to be doubted, though, whether Ammianus had such ecclesiastical authority in mind. For him the ‘auctoritas’ of the urbs aeterna is most unlikely to have been its Christian primacy, but rather its historical and cultural pre-eminence in the Roman world; for, despite the political and military realities of the late empire, Ammianus still has his history revolving round the traditional focus of Rome.²⁷ When Constantius paid a ceremonial visit to the city in 357, Ammianus' celebrated description of that occasion depicts the emperor entering Rome as the ‘true home of empire and of all virtues’,²⁸ and touring the monuments of the city's great past; the fact that this was the visit of a Christian emperor to a city now endowed with Christian holy places is pointedly ignored in this context. Moreover, when Constantius is said by Ammianus to have been ‘delighted’ by the ‘jibes’ (‘dicacitate’) of the Roman populace at the games, there is no mention here of the Christian congregation who clamoured for the restoration of their bishop, Liberius.²⁹

We must conclude, then, that Ammianus largely divests the Liberius episode of its Christian content, preferring to interpret it in accordance with a familiar pattern of the enforcement of the emperor's will against prominent opposition, with bishop Athanasius the victim of his own dangerous interests in futurology. Not indeed that Ammianus is unaware (far from it) of all the Christian implications, but these are not allowed to loom large in a secular history of Rome.

Turning to another Christian episode in Rome, we shall find Ammianus' treatment similar. The urban prefect of 366, Viventius (‘an upright and wise Pannonian’, a rare specimen in Ammianus); had his otherwise peaceful term of office disturbed by civil strife within the Christian congregation in the city, as the rival factions of Damasus and Ursinus laid claim to the succession to Liberius. Viventius could not contain the violence, and was forced to withdraw; one day alone saw 137 people dead in one of the churches in Rome, the ‘basilica Sicinini’.³⁰ Urban violence in itself was a familiar theme for Ammianus' record of events at Rome and the activities of its prefects;³¹ but this Christian slaughter over the possession of the bishopric does provoke some reflections from the historian. He draws a famous contrast between those whose desperate striving to obtain a high position at Rome is excused by all the wealth and luxury which will come their way, and on the other hand ‘some provincial bishops’ whose simplicity of life was more likely to commend them to ‘the eternal god and his true worshippers’ (‘perpetuo numini verisque eius cultoribus’). As with his remarks on the episcopal to-ings and fro-ings of the reign of Constantius, there must be some element here of genuine observation of different styles of Christian clergy: certainly, composing his history in Rome, Ammianus was well placed to take note of the demeanour of the bishop there and his entourage, and his picture of ostentatious extravagance receives some support from the sharp criticisms of the conduct of the Roman clergy to be found in the correspondence of another provincial visitor to the city, Jerome.³² Yet there is more to this passage than a comment on contrasting styles. Ammianus' brief glimpse of ecclesiastical extravagance—wealthy matrons, grand carriages and clothes, lavish feasts—has much the same ring as his more extended criticism of the life-style of the Roman aristocracy elsewhere in the history, where Ammianus adopts the moralising pose of the sober foreigner, outraged by the excesses of the capital.³³ His commendation of the simplicity of ‘provincial’ clerics is in a similar vein. Any specific reference to Christianity, to the massacre incident in the basilica, or to the behaviour of Roman Christians, becomes merged with the same sort of general reflection on the difference between Rome and the provinces: it becomes difficult to believe that when he wrote of the ‘eternal god and his true worshippers’ Ammianus was thinking only, or indeed at all, of the Christian deity.³⁴ What began as an account of a specific incident involving Christians has turned into generalised moral reflection on the code of conduct best suited to the worship of God, and enjoined on pagans and Christians alike.

The Christians also intrude violently into Ammianus' narrative with his account of an incident at Alexandria. On 24 December 361 George, the bishop who had been imposed on the Alexandrian congregation after Constantius' second banishment of Athanasius, was lynched by a wild mob, along with two imperial officials. The bodies of the three of them were torn limb from limb and carried to the shore, where the remains were burnt and the ashes cast into the sea.³⁵ The date of this affair is established from a text known as the Historia Acephala, a near-contemporary, but now fragmentary, chronicle of the life of Athanasius, which makes clear that what sparked off the attack on George and the others was the receipt of the news of the death of Constantius (3 November 361), the emperor to whom George owed his possession of the see of Alexandria.³⁶ Ammianus in fact misplaces the episode to a year later, while Julian and his court were wintering at Antioch, and adds George to a list of individuals condemned to death or exile by Julian's regime in the months leading up to the Persian expedition.³⁷ This misplacement conceals to some degree the true character of the circumstances of George's death, which has to be seen in the context of the violent hostilities prevalent among Alexandrian Christians between supporters and opponents of Athanasius, hostilities which had made George's tenure of the see far from peaceful and now erupted after the death of Constantius, when Julian permitted exiled Christian bishops (Athanasius included) to return to their congregations.³⁸ That George was really the victim of violence meted out by other Christians is all but confirmed by the emphasis with which church historians (Socrates and Sozomen) deny the point, and endeavour to shift the blame on to the pagans: they seek to exonerate fellow-Christians, specifically the followers of Athanasius, by quoting the letter which Julian addressed to the citizens of Alexandria, in which he reprimands them for taking the law into their own hands by the murder of the bishop, but refrains from exacting any punishment.³⁹ In this letter Julian certainly writes as though pagans were responsible for George's death, angry at the behaviour of an ‘enemy of the gods’;⁴⁰ but the text is inconclusive, since in Julian's eyes the whole population of Alexandria were fellow-Hellenes and hence, to his way of thinking, pagans (he addressed the demos of Alexandria as followers of Alexander and ‘above all, of the great god Serapis’).⁴¹ For Julian, Alexandria was a pagan city, and its inhabitants had to be pagans.

In contrast to Julian's letter and to the ecclesiastical historians, the version of Ammianus, in its misplaced setting, does not give priority to the religious context of George's murder. There is the merest hint of the Christians divided among themselves in the remark that the victims could not hope for protection from the Christians because of the universal hatred in which George was held (‘Georgi odio omnes indiscrete flagrabant’): this implies that Christians participated in the lynching, and mirrors the statement in the Hist. Aceph. which incriminates ‘nearly all the people’.⁴² Ammianus does seem to follow Julian in assuming that the responsibility was primarily that of the pagans, and reports the offence caused by George's scoffing at a pagan temple, ‘how long will this sepulchre stand?’ (the church historians have different details, but the same emphasis on the ridicule of paganism).⁴³ Yet for Ammianus such offence was a secondary factor in the bishop's unpopularity, which he attributes first and foremost to George's relationship to the court of Constantius, and to his exploiting his standing with the emperor to inform against citizens of Alexandria: the bishop becomes identified with that species of servile courtier familiar from elsewhere in Ammianus' portrayal of Constantius and his entourage, filling the emperor's ‘wide-open ears’ with a stream of accusations.⁴⁴ The historian thus interprets the bishop's offence in decidedly worldy terms, in much the same way as he had earlier seen the allegations against Athanasius; and he explicitly recalls what he had said of Athanasius in using almost identical language of George, ‘disregarding his professed belief, which demands nothing that is not just and mild’ (‘quae nihil nisi iustum suadet et lene’).⁴⁵ Once more Ammianus encapsulates what to him is the essential merit of Christianity, perverted by the offending bishop. It is some measure of the historian's success in distancing himself from the internal conflicts of church politics that he is in a position to level identical accusations against both Athanasius and one of his principal opponents. It is again clear, though, that this distancing from ecclesiastical affairs, and the secularising of Christian history, is not to be mistaken for ignorance about Christianity; here, for instance, we have one of the two passages in the history where Ammianus alludes to the Christian fondness for the veneration of martyrs, as an explanation of the thoroughness with which the victims' murderers removed all trace of their mortal remains.⁴⁶

Ammianus' faithfulness to the tradition of Roman historiography in ‘playing down’ intrusive Christian material is further illustrated in the account of Julian's attempt to revive the temple of Apollo at Daphne, near Antioch. In the course of the renovation Julian ordered the removal of burials from the neighbourhood of the temple, an action which to Ammianus is nothing more than traditional temple purification, and he cites the model of the Athenians and the island of Delos.⁴⁷ Christian authors, by contrast, saw Julian's order as a deliberately hostile act by the pagan emperor aimed at one particular tomb, that of the martyr Babylas (whose relics had been transferred to Daphne not long before by Julian's brother Gallus), and concentrated on the removal of his remains into Antioch amid much display of devotion. Ammianus mentions only ‘circumhumata corpora’, studiously avoiding any reference to a Christian martyr.⁴⁸ Nor does he make any causal link, other than the vague ‘eodem tempore’, between the removal of the bodies and the fire which subsequently swept through the newly renovated temple, which Julian blamed on the Christians in Antioch: he took harsh reprisals against them, including the closure of the main church in the city.⁴⁹ Christian writers, of course, responded to Julian's accusations by asserting that the fire was divinely instigated as retribution for the attempt to revive the worship of Apollo.⁵⁰ Ammianus, on the other hand, while acknowledging the suspicion which fell upon the Christians, typically does not leave it at that: he is alone in mentioning another possible cause of the fire which he had heard put around (‘licet rumore levissimo’), that it had been started by stray sparks from candles lit by the philosopher Asclepiades when he was on his way to visit the emperor.⁵¹ It may be fanciful to suppose that there is a pointed irony in the suggestion that the zealous pursuit of the worship Julian was seeking to encourage had accidentally set fire to the temple; but what is certain is that Ammianus is not content to label the incident in the terms of a conflict between pagans and Christianity, as it was interpreted both by Julian and by his Christian opponents. The historian stands aside, distanced from their version of events.

Ammianus also takes note of another of Julian's attempts at temple restoration, his ill-fated scheme to rebuild the Jewish Temple in Jerusalem. This project was abortive, brought to a stop by some natural upheaval, most likely an earthquake. Even the sober Ammianus allows himself to mention ‘fearful balls of flame’ bursting out from the foundations, while the Christian historians exult at the miraculous evidence of divine disapproval of the undertaking.⁵² What is striking about Ammianus' account is, once more, his refusal to draw out, even to mention, the religious implications of Julian's intentions: his aim of reviving the Jewish cult in Jerusalem and thus of dislodging Christianity from the central place in the holy city which it had acquired under Constantine.⁵³ For the Roman historian, Julian's efforts were aimed at no more than to ‘spread the memory of his rule through the greatness of his public works’, a thoroughly secular explanation in the tradition of extravagant imperial building schemes. Ammianus implies that Julian deserved to fail in the attempt, and to pay the penalty for his over-reaching ambition.⁵⁴

Ammianus' version of the Temple episode is a good example of his reticence when it comes to the Christian aspects of his subject matter, and of his preference for secular over ecclesiastical interpretations. I began by linking this firmly to the historiographical tradition of which he felt himself a part. But what of his real opinions about Christianity and Christians? Can the veil of literary restraint be lifted sufficiently to reveal anything of what the historian felt about the empire's new faith? As an army officer Ammianus had travelled the empire and cannot fail to have observed contemporary Christianity at every turn: the public life of the church was plain for all to see. In so far as this contemporary expression of Christianity found its way into his history, Ammianus seems to have viewed much of it with the same vein of vivid mockery that he displayed, for example, towards the habits of the Roman aristocracy. This is the tone of his satirical portrayal of the affluent ostentation of some bishops, and of their monopolising of the cursus publicus in the reign of Constantius. There is also the same emphasis about his often quoted remark, in connection with Julian's amnesty for Christian exiles, that ‘no wild animals are as deadly to human beings as are most Christians in their hatred of each other’—a comment which appears to be Ammianus' own.⁵⁵ He might well find the observation justified in the face of the carnage in the Roman basilica in 366, or of the treatment of bishop George of Alexandria (he would also have on record the fate of Constantius' magister equitum Hermogenes, killed in a Christian riot in 342 over the possession of the see of Constantinople: Ammianus had described this incident in one of the lost books, but again gives no indication that he had interpreted it as a Christian affair).⁵⁶ Mockery of Christian practices is also likely to lie behind Ammianus' picture of the general Sabinianus wasting his time ‘amid the tombs of Edessa, as though he had made his peace with the dead’, and delighting in military exercises when he ought to have been moving against the Persian enemy.⁵⁷ There is little doubt that these ‘tombs’ are Christian tombs, and that Ammianus is pouring scorn on Sabinianus' obsession with the veneration of the dead—a familiar topic in contemporary polemic against the Christians.⁵⁸ Edessa was well endowed with the shrines of Christian martyrs, principally that of the apostle Thomas, to satisfy Sabinianus' devotions: later in the century, in 384, the Jerusalem pilgrim Egeria would make a special detour to visit the tombs of Edessa.⁵⁹ But criticism of the Christians, it must be emphasised, is not Ammianus' main point in his remarks about Sabinianus, which is that tombs, any tombs, were ill omened and associated with dangerous magical practices (he says as much elsewhere in the history), and no place for a Roman commander on the eve of a confrontration with the enemy.⁶⁰ In any case, even if Sabinianus' veneration of martyrs is being mocked, it is not his Christianity which is the basis of the hostile tone, but rather that he had supplanted from command the historian's superior officer and patron, Ursicinus.⁶¹

If there is a strain of mockery about Ammianus' presentation of some aspects of the public behaviour of Christians, we have also to reckon with those passages, already noted, where he is respectful of Christianity, and commends the modesty, directness and simplicity of the Christian ‘profession’—not of course as a set of theological doctrines, but as a code of upright moral conduct ‘which demands nothing that is not just and mild’. Sceptical readers of Ammianus might wish to observe that he only mentions the merits of Christianity when its adherents fail to display them: Constantius, George, Athanasius, the bishops of Rome, all in their various ways in Ammianus' eyes perverted the essentials of their faith. Is the historian, then, genuinely appreciative of the good points of Christianity—or is there a hint of cynicism, that Christians do not practise what they preach?

This latter suggestion is not to be pressed, unless it can be seen to be supported by a general undercurrent of hostility to Christianity in Ammianus' pages. Attempts have been made to demonstrate the anti-Christian bias of Ammianus, but it is very elusive.⁶² Christians and Christianity are mentioned, it might be argued, in a context where Roman authority is in some way or other under threat, and are found wanting in the defence of security and law (the age-old accusation of failing to be good citizens):⁶³ Christian churches are identified as places of refuge, albeit short-lived, for a condemned criminal and a usurper to the throne; a Christian festival diverts the attention of Roman soldiers and facilitates a barbarian raid; Christian martyr cult distracts a Roman commander in a frontier emergency. On a number of occasions the mediators between Rome and her enemies are identified as Christian clergymen—a hint, so the argument continues, that they are not to be trusted and might be disloyal. The principal protagonist here is the bishop of the fortress of Bezabde on the Tigris, who played a leading role in the negotiations which preceded the capture of the place by the Persians in 360. Ammianus recounts how the bishop was granted an audience with the Persian king and urged him ‘placido sermone’ to withdraw; his pleas were of no avail and Sapor pressed on with the siege. After this meeting there was a suspicion, which Ammianus dismisses as ‘unfounded, I believe, yet spread by the assertions of many’, that the bishop had secretly revealed the most vulnerable parts of the town's defences, since these now came under heavy attack.⁶⁴ Apart from the fact that the rumour about the bishop is mentioned at all, there is nothing in Ammianus' account to indicate that his purpose was to incriminate him on a charge of treachery: he goes out of his way, in fact, to say that he did not believe the story. Local Christian tradition did not believe it either, and the Syriac martyr literature counts the bishop of Bezabde among the thousands carried off into captivity, and dying in the hands of the Persians.⁶⁵ It must be said, too, that Christian clergy, whose influence might well be effective on both sides of a political and military frontier, were well placed to play a diplomatic role, and are found increasingly operating in that capacity;⁶⁶ the fact that Ammianus identifies them in this role need imply no suspicion of their loyalty. Nor is there any warrant for citing the mention of churches harbouring ‘wanted’ men, or Christian festivals providing opportunities for attack, as evidence of an anti-Christian sentiment on the part of Ammianus: surely they are no more than identifying labels properly included by the observant historian, and do not carry with them any hostile overtones.

On the other hand, it is possible to point to passages where Ammianus avoids incriminating Christians, when he might well have seized the opportunity if this were his intention. The bishop of Bezabde is a case in point. So too is Asclepiades the philosopher: we have seen that Ammianus made a point of inserting, even if only on the basis of a fragile rumour, an alternative explanation of the burning of the temple of Apollo at Daphne which shifted the blame away from the Christians—if he had wanted the accusation against the Christians to stick, he should have kept silent about Asclepiades' candles. And then there is that fatal spear which felled Julian on the battlefield.⁶⁷ By the time that Ammianus was composing his account of Julian's last campaign, stories about the circumstances of his death were rife, as well as rumours that it was no Persian enemy who had been responsible: Libanius hinted at the agency of the Christians (‘the people who were keen to have him killed’), and Christian writers, starting from Ephraim the Syrian and Gregory of Nazianzus, welcomed the suggestion that one of their number may have been the apostate emperor's assassin (it was not long before he was a saint).⁶⁸ Ammianus was not unaware of these allegations about Julian's death, but pointedly refused to make anti-Christian capital out of them.⁶⁹ True, it would hardly suit his generally admiring portrait of Julian to suggest that the emperor might have been killed by one of his own side, yet if his purpose had really been to discredit the Christians as disloyal subjects of the emperor he could hardly have passed this opportunity by.

In characterising individuals in the course of his narrative, Ammianus again misses opportunities to make remarks against Christianity. It is indeed true that he reveals a higher opinion of leading pagans like Praetextatus and the elder Symmachus than of a prominent Christian such as Petronius Probus; but it is Probus' avarice and corruption, and that of his whole family, which gives rise to the historian's sharp censure, and not his Christianity—which is not mentioned.⁷⁰ Probus' religion appears to be one of the few attributes which Ammianus does not hold against him. The same may be said, for example, of Strategius Musonianus, praetorian prefect of the East (354-8): he had been a Christian associate of Constantine's (Ammianus refers to his helping the emperor in the investigation of heretical sects) and had been an imperial representative at the council of Sardica in 343. Yet when Ammianus has faults to find in him, he points to his ‘lucrandi aviditas’, not to his Christianity.⁷¹

Ammianus' portrait of the brief reign of Jovian forms a tailpiece to the narrative of Julian's Persian expedition, and has to be interpreted in that light: the historian seeks to absolve Julian of some of the responsibility for the failure of the enterprise by focusing the odium of surrendered Roman territory and a shameful peace on to his short-lived successor. Yet just as Ammianus had refused to engage in anti-Christian polemic on the subject of Julian's death, so the fact that his successor was a Christian plays no part in the historian's criticism of him, which concentrates rather on his youthful inexperience and indifferent qualities—especially with regard to the episode of the surrender of Nisibis.⁷² Ammianus nowhere suggests that Jovian's Christianity was of any significance in his elevation to the throne, and he passes without comment over the mention of pagan sacrifices being offered on the new emperor's behalf.⁷³ As with other prominent Christians in the history, Jovian's religion does not enter into the reckoning when it comes to identifying his faults; on the contrary, his Christian commitment is one of the few positive commendations Ammianus bestows on him.⁷⁴ Ammianus' Jovian is an unlikely creation for an anti-Christian historian.

Ammianus' view of Julian himself is more problematical: the ambiguities in the portrait of the central figure of our surviving books, passing from model prince in Gaul to failed emperor on the Persian front, make it difficult to determine where the historian really stood.⁷⁵ Yet, on the religious question, it appears again that Ammianus' Julian is not the figure we would expect to be depicted by a historian hostile to, or seeking to discredit, Christianity. When Julian was first proclaimed Caesar and entered the city of Vienne in Gaul, Ammianus tells how a blind old woman, on discovering who the new arrival was, cried out ‘this is the man who will restore the temples of the gods’;⁷⁶ five years later, in the same city, Julian (now a usurping Augustus) was still pretending to be a Christian and publicly celebrating the festival of Epiphany, although according to Ammianus he had secretly long ago deserted the faith.⁷⁷ These two episodes frame the narrative of Julian's remarkable Caesarship in Gaul; but it must be said that the old woman's enthusiasm for Julian's paganism leaves no mark on the record of his successful campaigns against the Germans, where the virtues of the ideal prince are exemplified in government and on the battlefield, but not in his religious persuasions, which are all but ignored. Perhaps this merely reflects the fact that Julian's pagan beliefs were still in their secret phase;⁷⁸ yet even when Julian is sole emperor and his own master, Ammianus' presentation of the pagan revival is sketchy and at best lukewarm in its enthusiasm.⁷⁹ The reopening of the temples comes low down the list of Julian's reforms, after the purge of the court.⁸⁰ as Bowersock has observed, Ammianus gives the impression that Julian's public renunciation of Christianity occurred rather later than was actually the case⁸¹—hardly the method of a historian who attached a particular priority to the attack on the Christians. Moreover, as is well known, some aspects of the pagan revival come in for strong criticism from Ammianus: the excessive indulgence in sacrificing, and the education edict effectively debarring Christian teachers from their posts.⁸² This latter piece of legislation, arguably the most central and potentially far-reaching of Julian's measures against Christianity, is twice condemned by Ammianus as ‘inclemens’, and thought worthy to be ‘obliterated in eternal silence’, again not the judgement we should expect from a historian hostile to the Christians. In his concluding obituary of Julian, Ammianus finds room for few faults, yet he does manage to include his last word on the emperor's paganism in this category: Julian was ‘too much given to the investigation of portents … superstitious rather than a true worshipper of the divine (“superstitiosus magis quam sacrorum legitimus observator”).’⁸³

Julian's paganism is thus exposed to much the same criticism as was Constantius' Christianity in the passage with which I began: the Christian emperor's passion for the investigation of disputes was seen as a perversion of essential Christianity by an old woman's ‘superstitio’, Julian's paganism is judged more superstition than true religion. Evidently what mattered to Ammianus in the assessment of them as emperors was not their religious convictions as such, Christian or pagan, but rather that they should not be seen to descend into ‘superstition’—a failing which contrasted not only with pure ‘religion’ (and Ammianus, we remember, was content to regard Christianity in its essence as ‘religio absoluta et simplex’), but also with the ‘legitimacy’ of proper imperial behaviour.⁸⁴ Julian's excessive superstition was one of the many aspects in which, to Ammianus's eyes, he fell short of the conduct properly to be expected of a Roman emperor. These factors, religion—the true worship of the eternal divinity, whether pagan or Christian—and legitimacy—the correct forms of imperial conduct—were of far more consequence in Ammianus' view of his world than opposition to Christianity, and it is a mistake to suppose that such opposition was at the centre of his thinking. When he did have occasion to find fault with someone's religion, it was the excesses and failings of both pagans and Christians which caught his attention. There is not much to choose between his picture of Julian's soldiers gorged on sacrificial meat having to be carried home through the streets of Antioch⁸⁵ and the tableau of the ostentatious behaviour of church leaders in Rome: both in their different ways were a perversion of true ‘religion’, be it pagan or Christian.

The fact is that, as often as not, Christianity per se was just not an issue for Ammianus: as Momigliano once put it, ‘what matters is virtus, not paganism or Christianity’.⁸⁶ Individuals may be accused of avarice, cruelty, intemperance, and a whole host of vices, but never of Christianity. Even in the case of the emperor Constantius, whose brand of Christianity is censured, it is very much a tailpiece to the historian's wider complaint about the nature of Constantius' regime. Pace E. A. Thompson, there is no perceptible difference in Ammianus' attitude in the later books. In Book XXX occurs what all are agreed is a key passage: in his obituary of the emperor Valentinian I, Ammianus commended the fact that

among the different religions he stood in the middle, and did not disturb anyone nor order them to worship this or that god; he did not use threatening prohibitions to force his subjects to bow down to the god that he himself wished, but he left these matters uninterfered with, as he found them.⁸⁷

Some may wish to see this as a pagan writer pointedly commending a Christian emperor for tolerating paganism, with more than half a glance at the intolerant line being taken by Theodosius, in whose reign it was written;⁸⁸ but surely it is rather approval for a regime which did not make an issue, one way or the other, of religion and stayed out of ecclesiastical arguments—a regime which was mindful of the way an emperor ought to behave in this sphere, and of the proper place of ‘religion’. In Ammianus' view, by distancing himself from the conflicts of churchmen Valentinian did what both the Christian Constantius and the pagan Julian had failed to do; the language he uses here recalls not only his complaint that Constantius had tried to force everyone to his own way of thinking, but also a revealing comment that it had been a weakness of Julian's from time to time when presiding in judgement to ask the parties which god they worshipped.⁸⁹ Such a question from the emperor in court Ammianus found ‘intempestivus’, out of place; and his criticism of Julian for asking it is a reflection of his general complaint that the emperor, illegitimately, allowed his public policies to be dictated by religion. Similarly, what was so objectionable to Ammianus about the prohibition of Christian teachers was that it was driving a wedge of religious difference into the common culture of classical education, and provoking divisions where there were no boundaries between pagan and Christian: this was not the place for religious polemics, nor was Julian's action appropriate conduct for a Roman emperor.

I have tried to argue that two streams flowed into the formulation of Ammianus' opinions about Christianity: the historiographical tradition of which he was a part, and his contemporary experience of the Roman empire (the streams are not, of course, independent, for the literary background against which the historian fashions his work will inevitably also shape his view of the world). The first disposed Ammianus to see Christian matters in a detached light, tending to fit them into the familiar secular patterns of his historical heritage—the relationship between emperor and subjects, the maintenance of public order by Roman officials. The Christian content was, if not washed away entirely, certainly diluted: not, it must be re-emphasised, because Ammianus was ignorant of Christianity, but because a history of Rome had other topics to dwell on. What Ammianus learnt from the traditions of Roman historiography can only have been reinforced by his experience of the ill-fated religious policies, Christian and pagan, of the emperors of his own day. The actions of Constantius, and above all the failure of Julian, had demonstrated to him that the business of government and empire could not successfully be dictated by religious conviction: pagan and Christian arguments were outside the proper sphere of imperial conduct, and the emperor's interventions endangered the purity of true religion. This will recently have been brought home to him in the contrast between the religious policies of Valentinian and Valens:⁹⁰ from close at hand in Antioch Ammianus cannot fail to have reacted unfavourably to Valens' purges of dissenting Christians in the eastern provinces in the years before the disaster at Adrianople, and he will have had his opinions confirmed—certainly about the Christian tendency to internecine strife, but also too about the illegitimacy of emperors allowing religion to determine their conduct of affairs. Indeed Ammianus may have drawn much the same conclusion from Valens' defeat by the Goths as from Julian's failure in Persia. The readers of Ammianus' endorsement of the religious attitude of Valentinian doubtless reflected on its relevance to the vigorous enforcement of Christianity by Theodosius and his successors; but all that we know of Ammianus from his history suggests that his protest was less against the imposition of a particular creed than against the inappropriateness of any imperial meddling in matters of religion.

Notes

1. 21.16.18 (References in these notes where no author is specified are to Ammianus Marcellinus).

2. Even modern discussions of Constantius' church policy tend to set too much store by the ‘opposition’ (Athanasian) version, and do not give the emperor the credit for following the precedents in these matters set by his father: e.g. Karl Baus et al. The Imperial Church from Constantine to the Early Middle Ages (Engl. transl. London, 1980), ch. 3.

3. See appendix to the works of Hilary of Poitiers, Corp. Script. Eccl. Lat. 65, p. 64. On bishops and the imperial post at another council of Constantius', Rimini in 359, see Sulp. Sev. Chron. 2.41.

4. On Jovian, see 25.10.15, and below.

5. T. D. Barnes, The Sources of the Historia Augusta (Brussels, 1978), 117 ff.

6. See the comments of Photius, Bibliotheca, cod. 77 (Henry, 158-9), with R. C. Blockley, The Fragmentary Classicising Historians of the Later Roman Empire (Liverpool, 1981), 18 ff.

7. On ‘religio’ and ‘superstitio’ in Ammianus, see below, p. 199 [in E. D. Hunt, “Christians and Christianity in Ammianus Marcellinus,” Classical Quarterly n.s. 35, no. 1 (1985): 186-200]. I cannot follow P. M. Camus, Ammien Marcellin (Paris, 1967), 249, in supposing that Ammianus is referring specifically to Constantius' Arianism.

8. First, in 1627, by Chifflet: see E. A. Thompson, The Historical Work of Ammianus Marcellinus (Cambridge, 1947), 114. Claudian has been blessed with the same misconception: A. Cameron, Claudian (Oxford, 1970), 214 ff.

9. W. Ensslin, Zur Geschichtsschreibung und Weltanschauung des Ammianus Marcellinus (Klio Beiheft 16, 1923), 96 ff. Cf. G. B. Pighi, RAC [Revista di Archaeologia Cristiana] 1 (1950), 386-94; A. Demandt, Zeitkritik und Geschichtsbild im Werk Ammians (Bonn, 1965), 79 ff.; P. M. Camus, op. cit. 261 ff.; J. Geffcken, The Last Days of Greco-Roman Paganism (transl. S. MacCormack, Amsterdam, 1978), 187-8.

10. ‘The Demise of Paganism’, Traditio 35 (1979), 45-88, esp. 55 ff.

11. E. A. Thompson, op. cit. 111 ff.

12. The main thesis of S. D'Elia, ‘Ammiano Marcellino e il Cristianesimo’, Stud. Romani 10 (1962), 372-90; cf. A. Selem, ‘Considerazione circa Ammiano ed il Cristianesimo’, Riv. di cult. class. e med. 6 (1964), 224-61, and L. Angliviel de la Beaumelle, ‘Remarques sur l'attitude d'Ammien Marcellin à l'égard du christianisme’, Mélanges offerts à W. Seston (Paris, 1974), 15-23. For criticism of this approach see R. C. Blockley, Ammianus Marcellinus, a study of his historiography and political thought (Brussels, 1975), ch. vii.

13. For the Harvard dissertation (1971) of T. G. Elliott, The Pagan Bias of Ammianus Marcellinus xiv-xxv, I am dependent on the synopsis in HSCP [Harvard Studies in Classical Philology ] 79 (1975), 358. For a blast against the notion that literature was a significant part of pagan revival, see A. Cameron, ‘Paganism and literature in late fourth century Rome’, Fondation Hardt Entretiens, 23 (1977), 1-30.

14. A. Momigliano, ‘Pagan and Christian historiography in the fourth century a.d.’, in The Conflict between Paganism and Christianity in the Fourth Century (Oxford, 1963), 79-99, repr. in Essays in Ancient and Modern Historiography (Oxford, 1977), 107-26; A. & A. Cameron, ‘Christianity and tradition in the historiography of the late Empire’, CQ [Classical Quarterly] n.s. 14 (1964), 316-28. Cf. also H. Tränkle, ‘Ammianus Marcellinus als römischer Geschichtsschreiber’, Antike und Abendland, 11 (1962), 21-33.

15. Cf. Tacitus' celebrated ‘quos … vulgus Christianos appellabat’, Ann. 15.44.

16. 15.5.31 (Silvanus), 26.3.3 (Hilarinus); diplomacy—20.7.7 (Bezabde), 29.5.15 (north Africa), 31.12.8-9, 15.6 (Adrianople).

17. 27.10.2 (Mainz), 28.6.27 (Palladius).

18. 21.2.5. Ammianus here secures a place in the history of Christian liturgy, with the first certain evidence for the celebration of Epiphany in the West: E. Pax, RAC 5 (1962), 902.

19. 18.10.4. Ammianus is obviously aware of the reputation of Sapor II as a persecutor of the Christians: J. Labourt, Le Christianisme dans l'Empire Perse sous la Dynastie Sassanide (224-632) (Paris, 1904), ch. iii; cf. A. Christensen, L'Iran sous les Sassanides² (Copenhagen, 1944), 267-8.

20. Cf. F. Paschoud, Roma Aeterna (Rome, 1967), 59-60.

21. 15.7.6 ff. For general background see modern church histories, e.g. W. H. C. Frend, The Early Church (London, 1965), ch. 13; H. Chadwick, The Early Church (Harmondsworth, 1967), 136-45; K. Baus et al., loc. cit. (n. 2). Athanasius' own version of these events is at Hist Arian. 35-41.

22. For Ammianus and the ‘documents préfectoraux’ see G. Sabbah, La méthode d'Ammien Marcellin (Paris, 1978), 183-4. That Leontius was having trouble with the Christians is perhaps signalled by his arrest of the ringleader Petrus Valuomeres (15.7.4).

23. See the revealing papyrus documents discussed by H. I. Bell, Jews and Christians in Egypt (London, 1924), 45 ff., and esp. 53 ff.

24. Sozom. Hist. Eccl. 4.9.10.

25. 15.3.7 ff. On the suspicious nature of Constantius see, among many passages, 16.8.2, 19.12.5, 21.16.8 ff.; with H. Funke, ‘Majestäts- und Magieprozesse bei Ammianus Marcellinus’, Jahrb. für Ant. und Christ. 10 (1967), at pp. 151 ff.

26. So A. & A. Cameron, op. cit., 323: ‘Ammianus refers to papal authority in terms which gladden the heart of the Catholic historian’, cf. Ch. Pietri, Roma Christiana (Rome, 1976), 246.

27. Cf. Paschoud, loc. cit. (n. 20); Tränkle, op. cit. For Ammianus, Rome was the only fit burial-place for his hero Julian: 25.10.5.

28. 16.10.13.

29. As reported by the pamphlet preserved in the Collectio Avellana, CSEL [Corpus Scriptorum Ecclesiasticorum Latinorum] 35, 2; Theodoret, Hist. Eccl. 2.17, has the agitation coming from senators' wives.

30. 27.3.11 ff. On this episode, and other accounts (principally the partisan pamphlet in the Coll. Avell.), see Ch. Pietri, op. cit. 408 ff., with A. Lippold, Historia, 14 (1965), 119 ff. For the ‘basilica Sicinini’, cf. Jer. Chron. an. 366 (p. 244 Helm)—Jerome may himself have been a young eye-witness of the incident. Lippold disputes the traditional identification of the basilica with the church which was later to become S. Maria Maggiore.

31. Cf. his lament at 14.6.2 that the record was ‘nothing but riots, inns, and other such base activities’.

32. Especially his Ep. 22: J. N. D. Kelly, Jerome, his life, writings, and controversies (London, 1975), 108-9. Note also the famous remark allegedly addressed to Damasus by Praetextatus, Viventius' successor as urban prefect: ‘make me bishop of Rome, and I will be a Christian tomorrow’ (Jer. C. Ioh. Hier. 8 = PL 23,361).

33. For the ‘honestus advena’, see 14.6.12.

34. ‘Numen’ is Ammianus' favourite word for ‘divinity’: see, e.g., Ensslin, op. cit. 48 ff.; Camus, op. cit. 134 ff. I fail to see why Blockley, op. cit. 126, should say ‘it would more naturally be taken to indicate the Christian God and Christians alone’.

35. 22.11.3 ff.

36. Hist. Aceph. 8, best read in C. H. Turner, Eccl. Occ. Monum. Iur. Ant. 1.2,666.

37. Ammianus must be wrong in stating that the murder of George followed the execution of the former dux Aegypti Artemius: the latter appears still to have been alive after George's death. See PLRE [Prosopography of the Later Roman Empire] 1, 112.

38. Cf. 22.5.3; with Hist. Aceph. 10, and Julian, Ep. 110 Bidez.

39. I.e. Julian, Ep. 60 Bidez, quoted by Socr. Hist. Eccl. 3.3; cf. also Sozom. Hist. Eccl. 5.7. In the letter as preserved by Socrates Julian contents himself with merely offering ‘advice’ (380c, ‘παραίνεsιν καὶ λόγουs’) to the Alexandrians, whereas Ammianus' summary (22.11.11) has him ‘threatening’ them (‘minatus extrema’).

40. Ep. 60. 379c.

41. Ibid. 378c; cf. Ep. 111 Bidez, 433a: ‘I am ashamed if anyone at all among the Alexandrians admits to being a Galilean’.

42. 22.11.10, Hist. Aceph. 8.

43. According to Socr. and Sozom. (loc. citt.) George tried to build a church to replace, not a contemporary pagan shrine, but a deserted temple of Mithras: Christians mocked the human remains discovered. Note that in the case of George's two fellow-victims, Ammianus, 22.11.9, is clear that it was excessive Christian zeal which was to blame.

44. For the ‘patulae aures’ of Constantius, cf. among other passages 15.3.5, 18.3.6, 4.4.

45. 22.11.5.

46. Cf. the victims of Valentinian venerated by the Christians in Milan, and the quaestor Eupraxius' advice to the emperor not to provide them with more martyrs: 27.7.5-6. At 22.11.9 Ammianus appears to allude, disapprovingly, to another Christian practice, the tonsure: Diodorus was unpopular because he ‘liberally cut off the young men's locks, thinking that even this mattered to the worship of gods’ (I can find no reference to this passage in the standard treatment of the Christian tonsure by P. Gobillot, RHE [Revue d'Histoire Eccsesiastique] 21 [1925], 399-451).

47. 22.12.8; cf. Thuc. 3.104. For Julian and Daphne, G. Downey, A History of Antioch in Syria from Seleucus to the Arab Conquest (Princeton, 1961), 387-8.

48. Cf. Julian himself, Misop. 361 b, and Libanius, Or. 60.5, who contemptuously do not identify the body (merely the ‘corpse’). Christian accounts, Joh. Chrys. In S. Babylam, 15 (PG [Patrologia Graeca] 50, 555), Sozom. 5.19.17, refer to ‘bodies’ close to the oracle, but single out Babylas.

49. 22.13.2.

50. Joh. Chrys. op. cit. (PG 50, 559): the martyr's prayers called down the fire. Cf. Sozom. 5.20.5.

51. For Asclepiades visiting Julian, see his Against the Cynic Heracleius, 224 d.

52. 23.1.2-3, with Rufin. Hist. Eccl. 10.38-40. For the whole episode, in relation to Julian's attitude to the Jews, see M. Avi-Yonah, The Jews of Palestine (Engl. transl. Oxford, 1976), ch. viii. Recent evidence on the date of the building project, in May 363 (on which see G. W. Bowersock, Julian the Apostate [London, 1978], appendix i), suggests that Ammianus has misplaced the episode to the time before Julian's departure for Persia, which took place on 5 March 363.

53. In the words of Avi-Yonah, op. cit. ‘to return Jerusalem to the Jews’; cf. my Holy Land Pilgrimage in the Later Roman Empire AD 312-460 (Oxford, 1982), 156-7 (the authenticity of Julian's letter to the Jews, Ep. 51 Loeb, is a matter of dispute). On Ammianus' treatment of this episode, cf. Tränkle, op. cit. (n. 14), 32.

54. On Julian's ‘ostentation’ and desire for ‘empty renown’, cf. 22.7.3, 14.3.

55. 22.5.4.

56. 14.10.2. On the fate of Hermogenes, killed by partisans of bishop Paul of Constantinople, see Jerome, Chron. an. 342 (p. 235 Helm), Socr. Hist. Eccl. 2.13, Sozom. Hist. Eccl. 3.7.6; the riot is also mentioned in Libanius' autobiographical oration, ch. 44, but again without reference to its Christian context.

57. 18.7.7, cf. 19.3.1. A. Cameron (Claudian, 224 ff.) drew attention to Claudian's similar ridicule of the magister equitum Jacob.

58. Cf. Julian, Misop. 344 a, C. Galil. 335 ff. (p. 414 Loeb ‘you have filled the whole world with tombs and sepulchres, and yet in your scriptures it is nowhere said that you must grovel among tombs and pay them honour’), Eunap. Vit. Sophist. 472; Ammianus' phrase ‘sepulchris haerentem’ is very close to the polemicists’ ‘προsκαλινδει̑sθαι’.

59. It. Eg. 17 ff., with J. B. Segal, Edessa, ‘The Blessed City’ (Oxford, 1970), 172 ff.

60. 18.7.7 ‘ominoso sane et incepto et loco, cum haec et huiusmodi factu dictuque tristia futuros praenuntiantia motus vitare optimum quemque debere’; cf. 19.12.14 (association of tombs and magic), and on Athanasius, see above.

61. See E. A. Thompson, op. cit. (n. 8), 50 ff.

62. For the attempts, see above, nn. 12-13.

63. The main point of L. Angliviel de la Beaumelle, op. cit.

64. 20.7.7-9.

65. J. Labourt, op. cit. (n. 19), 78-9, with Sozom. Hist. Eccl. 2.13.7.

66. See J. F. Matthews, RAC 10 (1978), 673 ff.

67. 25.3.6.

68. Liban. Or. 24.6 ff.; cf. Greg. Naz. Or. 5.13 (PG 35, 680), with I. Hahn, ‘Der ideologische Kampf um den Tod Julians des Abtrünnigen’, Klio, 38 (1960), 225-32, and (for Christian elaboration of the story), N. H. Baynes, ‘Death of Julian the Apostate in a Christian legend’, JRS [Journal of Roman Studies] 27 (1937), 22-9 (= Byzantine Studies and Other Essays [London, 1955], 271-81). According to Bowersock, Julian the Apostate, 117, the testimony of Philostorgius establishes that Julian was in fact killed from the Persian side.

69. 25.6.6 (‘rumore iactato incerto Iulianum telo cecidisse Romano’) makes clear that Ammianus knew the rumour; the usual reading of 25.3.6 (‘incertum unde’) alludes to it, but Fontaine's recent Budé text (1977) removes any hint of the doubts—cf. Sabbah, op. cit. (n. 22), 413-14. D. Conduché, ‘Ammien Marcellin et la mort de Julien’, Latomus, 24 (1965), 359-80, well reveals how Ammianus' whole account of Julian's last campaign smacks more of pagan opposition to him than of any Christian plots.

70. Criticism of Probus: 27.11, 30.5.4 ff.; and of Anicii generally, 16.8.13. Praise of Praetextatus, 22.7.6 (‘praeclarae indolis gravitatisque priscae senator’), 27.9.8; of Symmachus, 27.3.3 (‘inter praecipua nominandus exempla doctrinarum atque modestiae’).

71. 15.13.2, cf. 16.9.2 ‘venalis et flecti a veritate pecunia facilis’; for Musonianus, see PLRE 1, 611-12.

72. 25.9.7 ‘iuveni … quem nullis ante actae vitae insignibus in huiusmodi negotiis cognitum’.

73. 25.6.1.

74. 25.10.15, ‘Christianae legis itidem studiosus et nonnumquam honorificus’.

75. E. A. Thompson, op. cit. ch. v, was the first really to see just how much criticism of Julian there was; for a new view of Ammianus' disillusionment with Julian and its importance to his historical writing see J. F. Matthews, in T. J. Luce (ed.), Ancient Writers: Greece and Rome, 2 (New York, 1982), 1125 ff.

76. 15.8.22.

77. 21.2.5. Julian himself, Ep. 111 Bidez, 435a, traces his pagan conversion back to 351.

78. As emphasised by J. F. Drinkwater, in Carl Deroux (ed.) Studies in Latin Literature and Roman History, 3 (=Collection Latomus, vol. 180, 1983), 354 ff. Nevertheless Julian was openly encouraging pagan philosophers to visit him in Gaul: Liban. Or. 12.55, 18.74.

79. Cf. Thompson, op. cit. 84 ff., who notes that Zosimus has even less to say of Julian's pagan revival: according to F. Paschoud, in his Budé edition of Zosimus, 2 (1979), 100, pagan propagandists were embarrassed by the disaster of Julian's death.

80. 22.5.2. Note the contrast with Libanius, Or. 18.121 ff., who reverses the order, giving priority to the religious reforms.

81. Julian the Apostate, 61-2, citing Julian Ep. 26 Bidez, 415c, as evidence of his sacrificing openly before the death of Constantius. Julian also mentioned in his Ep. ad. Athen. 286d that he had offered sacrifices on his departure from Gaul, but perhaps these had taken place in secret.

82. 22.12.6 (sacrificing); 22.10.7, 25.4.20 (education edict). On the latter, cf. CTh [Classical Theology] 13.3.5 (17 June 362) and Julian, Ep. 61 Bidez; the best indication of the significance of this legislation is the strong Christian reaction, e.g. Greg. Naz. Or. 4.101 ff., accusing Julian of claiming for pagans a monopoly of education and culture—see G. Downey, ‘Julian and the Schools’, The Classical Journal 53 (1957-8), 97-103.

83. 25.4.17; cf. Eutrop. Breviarium, 10.16.3 ‘religionis Christianae nimius insectator’.

84. Ammianus found in the barbaric Huns (‘who were completely ignorant of right and wrong’) neither religion nor superstition, a more extreme state of disapproval: 31.2.11. On Ammianus and ‘legitimacy’, see now Matthews, op. cit. (n. 75), 1135 ff.

85. 22.12.6.

86. Op. cit. (n. 14), 95 (= Essays, 120).

87. 30.9.5; cf. Valentinian's own words to the senate in CTh 9.16.9. For a Christian acknowledgement of Valentinian's refusal to intervene, see Ambrose, Ep. 21.5.

88. As E. A. Thompson, op. cit. 115-16.

89. 22.10.2; nevertheless, Ammianus claims, Julian's sense of justice was not impaired by ‘religio’, nor by anything else.

90. Dwelt upon, of course, by the church historians: Socr. Hist. Eccl. 4.1, Sozom. Hist. Eccl. 6.6.10.

This paper first saw the light of day at a meeting of the Oxford Philological Society in October 1981; I am grateful for the comments made on that occasion, and more recently for the observations of Dr Jill Harries.