* This paper has benefited from its discussion with a variety of students, friends, and colleagues. Some of its conclusions were presented in the spring of 1977 to perceptive and helpful audiences at Scripps College (Claremont), Cornell University, and the Universities of Michigan and Maryland.

[[1]] The most influential modern account of these events is H. Bloch, 'A New Document of the Last Pagan Revival in the West, 393-394 A.D.,' Harvard Theological Review 38 (1945) 199-244 [cited below as 'Bloch (1945)']; Bloch repeated his arguments in a more accessible form in his 'The Pagan Revival in the West at the End of the Fourth Century, ' in A. D. Momigliano, ed., The Conflict Between Paganism and Christianity in the Fourth Century (London 1963) 193-218. The same story, in its essentials, had already appeared in J. Geffcken, Der Ausgang des griechisch-römischen Heidentums (Heidelberg 1929; repr. Darmstadt 1972) esp. 141-62; it recurs most recently, with little variation, in J. Wytzes, Der letzte Kampf des Heidentums in Rom (Leiden 1977) and J. H. Smith, The Death of Classical Paganism (New York 1976).

[[2]] Amm. 22.10.7.

[[3]] Amm. 25.4.20. Cf. also Amm. 25.4.17: '[Iulianus] . . . superstitiosus magis quam sacrorum legitimus observator.' Compare the pagan Eutropius, who ilourished in a Christian court, Brev. 10.16.3: ' [Iulianus] . . . religionis Christianae nimius insectator.' Non-Christian sensitivity on this point is probably indicated by Ambrose, Ep. 17.4 (384): 'Petunt etiam, ut illis privilegia deieras, qui loquendi et docendi nostris communem usum Iuliani lege proxima denegarunt.' This is the first charge with which Ambrose tried to smear his opposition, before even reading Symmachus' relatio.

[[4]] Note especially the not untypical way in which the influential depiction of fourth-century polytheism by Samuel Dill (in his Roman Society in the Last (Century of the Western Empire [London 1899] 5-10) drew upon sources as anachronistic as Plutarch and Apuleius.

[[5]] On Secundus Salutius see the first volume of the Prosopography of the Later Roman Empire (Cambridge 1971) 814-17 (all references are to the first volume of PLRE only). He is now generally accepted as the dedicatee of the fourth, and subject of the eighth, of Julian's orations. If, as many think, he is also the author of the treatise De diis et mundo, the semi-official tract of Julianic paganism, his actions are the more remarkable; see note 99 below.

[[6]] Greg. Naz., Or. 4.91; Socrates, HE 5.10; Sozomen, HE 3.19.

[[7]] i Rufinus, HE 10.37.

[[8]] Amm. 25.5.1-5. Zos. 3.36.1-2 put Secundus Salutius' diffidence a year later, upon the death of Jovian, and added the detail that he also prevented his own son's elevation. But on the date at least Ammianus (who was with the army at the time) is to be preferred, unless we hypothesize a second refusal a few months after the first.

[[9]] The case of Nerva in the first century is a suggestive parallel.

[[10]] The only anomalies generally accepted are Manichaeism and Judaism, which are, along with paganism and Christianity, the only categories of religious identification recognized by PLRE.

[[11]] See now my study, 'Paganus,' Classical Folia 31 (1977) 163-9.

[[12]] E.g. Bloch (1945) 217-18. The opposite transformation (of Christianity into an eastern mystery cult) is equally well-known and equally illegitimate.

[[13]] Pliny, Epp. 10.96-97. Sherwin-White's commentary provides the best guide to the extensive literature and debate on these texts.

[[14]] Pliny, Ep. 10.97.2: 'Conquirendi non sunt; si deierantur et arguantur, puniendi sunt, ita tamen ut, qui negaverit se Christlanum esse idque re ipsa maniiestum iecerit, id est supplicando dis nostris, quamvis suspectus in praeteritum, veniam ex paenitentia impetret.' The last phrase probably contains a sly dig at Christianity and shows that Trajan's legal staff knew what they were dealing with.

[[15]] Pliny, Ep. 10.96.3.

[[16]] Opponents regularly attacked this feature of Christianity as leading to irresponsible indifference to ethical standards. Augustine, Conf. 1.11.18, shares the same concern in connection with delayed baptism as conducive to moral laxity.

[[17]] Viz. Mardonius; cf. conveniently A. Piganiol, L'Empire chrétien2 (Paris 1972) 124-25, or any standard life of Julian. The best of these is now G. W. Bowersock, Julian the Apostate (Cambridge, Mass. 1978), on which I comment in a review in the Catholic Historical Review 66(1980) 471-72.

[[18]] H.M. Gwatkin, Studies of Arianism2 (Cambridge 1900) 26; C.N. Cochrane, Christianity and Classical Culture (Oxford 1940) 234: 'The question raised by Arianism was whether the substance of paganism was to survive under Christian forms.' Constans and Constantine II were probably anti-Arian, but their influence was limited in both space and time.

[[19]] In addition to the familiar evidence for the policies of Gratian and Theodosius, see CTh 16.7.1-5 (ranging from 381 to 393), the first laws against apostates from Christianity who lapsed back into paganism. The Catholic emperors also carried out a policy of removing pagan solemnities from the list of holidays and Christianizing the official calendar (CTh 2.8.18-25).

[[20]] See P. Camus, Ammien Marcellin (Paris 1967), for the standard treatment of Ammianus' intellectual preoccupations.

[[21]] See Alan and Averil Cameron, 'Christianity and Tradition in the Historiography of the Later Empire,' Classical Quarterly 58 (1964) 316-28.

[[22]] E.g. Amm. 14.6.3, 14.11.24-26, 15.28, et saep.

[[23]] E.g. Amm. 15.8.9-10, 21.13.14.

[[24]] Amm. 25.6.1.

[[25]] As in the famous contest over the papacy in 366: Amm. 27.3.12-15.

[[26]] Amm. 28.1.29.

[[27]] See Amm. 28.1 in its entirety; cf. A. Alföldi, A Conflict of Ideas in the Late Roman Empire (Oxford 1952) esp. 28-47. See also Amm. 29.1.28-33; 21.2.6-14.

[[28]] Good philosophers: Amm. 29.1.36-42. A fraud: Amm. 14.9.5, on 'Epigonus . . . amictu tenus philosophus.'

[[29]] Amm. 21.1, 23.5.

[[30]] Amm. 21.13, 22.14.

[[31]] It is good to bear in mind that apart from local acts of Christian fanaticism, non-Christians were not persecuted and were even protected, as by CTh 16.10.24 (8 June 423): 'Sed hoc Christianis, qui vel vere sunt vel esse dicuntur, specialiter demandamus, ut Iudaeis ac paganis in quiete degentibus nihilque temptantibus turbulentum legibusque contrarium non audeant manus inferre religionis auctoritate abusi.'

[[32]] Amm. 27.3.15, of prelates 'quos tenuitas edendi potandique parcissime, vilitas etiam indumentorum, et supercilia humum spectantia, perpetuo numini, verisque eius cultoribus, ut puros commendant et verecundos.'

[[33]] Amm. 20.5.9.

[[34]] Amm. 15.7.6-10.

[[35]] Amm. 21.16.18.

[[36]] Amm. 30.9.5: 'Postremo hoc moderamine principatus inclaruit, quod inter religionum diversitates medius stetit, nec quemquam inquietavit, neque, ut hoc coleretur, imperavit aut illud: nec interdictis minacibus subiectorum cervicem ad id, quod ipse coluit, inclinabat, sed intemeratas reliquit has partes ut repperit.'

[[37]] This is probably the significance of Amm. 28.4.24.

[[38]] Again see the Camerons' article (note 21 above).

[[39]] Amm. 25.4.17: 'praesagiorum sciscitationi nimium deditus, . . . superstitiosus magis quam sacrorum legitimus observator.'

[[40]] Amm. 23.5.13-14.

[[41]] J. F. Matthews, Western Aristocracies and Imperial Court, A.D. 364--425 (Oxford 1975).

[[42]] Matthews, xi, identifies a central theme of his book to be 'the manner in which the government of the western empire seems progressively in these years to fall from public into private hands.'

[[43]] In the years covered by O. Seeck, Regesten der Kaiser und Päpste für die Jahre 311 bis 476 n. Chr. (Stuttgart 1919), no reigning Augustus (with trivial exceptions) ever set foot in Britain, Gaul south of the Loire and west of the Rhône, the Iberian peninsula, Africa including Egypt, the Levant south of Antioch, Greece and the Aegean coast of Asia Minor, Italy south of the Apennines (except in the fourth century for purely ceremonial visits to Rome by Constantine [312-13, 315], Constantius [357], and Theodosius [389]), or the Mediterranean islands. Note that the omissions include the strongholds of senatorial privilege: Campania, Sicily, Africa. Imperial and senatorial spheres of interest scarcely overlapped.

[[44]] Matthews 76-81 et saep.

[[45]] E. Stein, 'La disparition du sénat de Rome à la fin du VIe siecle,' in his Opera minora selecta (Amsterdam 1968) 386-400. Mention should be made of P. Wormald,' The Decline of the Western Empire and the Survival of Its Aristocracy,' Journal of Roman Studies 66 (1976) 217-26, in which the author, reviewing Matthews' book, traces many of these same developments and in doing so gives more importance to the aristocracy than I find justifiable.

[[46]] On Maternus Cynegius, see Matthews, 140-42.

[[47]] Almost three times as many laws of the Theodosian Code are directed against heresy (66, in CTh 16.5) as against paganism (24, in CTh 16.10).

[[48]] CTh 16.10.21 (7 Dec. 416).

[[49]] On Themistius, see G. Dagron, 'L'Empire Romain de l'Orient au IVe siècle et les traditions politiques de l'hellénisme: le témoignage de Themistius,' Travaux et mémoires 3 (1968) 1-242; and Matthews, 116-18.

[[50]] A. Cameron, Claudian (Oxford 1970) 189-227.

[[51]] The traditional view of Synesius is embodied in C. Lacombrade, Synésius de Cyrène: hellène et chrétien (Paris 1951). Compare H.-I. Marrou, 'Synesius of Cyrene and Alexandrian Neoplatonism,' in Momigliano, ed., Conflict Between Paganism and Christianity 126-50.

[[52]] One should not forget Cyrus of Panopolis, discussed below, poet and courtier at Constantinople in the 430s, a 'Hellene' by faith. Disgraced at court ca. 442, he ended his life perforce as a bishop.

[[53]] A. D. Nock, Conversion (Oxford 1933) 156-57.

[[54]] On these figures, see PLRE 332 (Felix 3), 415 (Helpidius 6), and 470-71 (Iulianus 12).

[[55]] PLRE 409 (Hecebolius 1). The chronicler of the sophists, Eunapius, was willing to admit the possibility of a Christian sophist: most notably Proaeresius, teacher of Basil and Gregory Nazianzen (PLRE 731); and of course even Libanius' most famous pupil was none other than John Chrysostom.

[[56]] Augustine, Serm. 1 Morin (Miscellanea Agostiniana 1 [1930] 589-93).

[[57]] PLRE 900-902 (Theodorus 27).

[[58]] Aug., Retractationes 1.2.

[[59]] CIL 8.450 and 8.10516.

[[60]] On Volusianus and his family, see the masterly article of A. Chastagnol, 'Le sénateur Volusien et la conversion d'une famille de l'aristocratie romaine au Bas-Empire,' Revue des éudes anciennes 58 (1956) 240-53.

[[61]] Aug., Epp. 132, 135, 136, 137, 138, on which see M. Moreau, 'Le dossier Marcellinus dans la correspondance de saint Augustin,' Recherches Augustiniennes 9 (1973) 3-181, esp. 52-77 (Moreau's work is also published separately).

[[62]] Rutilius, De reditu 1.167-76. On Rutilius' attitudes, see further below.

[[63]] Chastagnol, 253.

[[64]] See the stemmata of the Ceionii Rufii (PLRE 1138), of Melania (PLRE 1142), and St. Paula (PLRE 1143).

[[65]] Jerome, Ep. 107 (written ca. 400).

[[66]] Jerome, Ep. 107.1.

[[67]] Ambrose, In Ps. 36.61 (CSEL 64.118).

[[68]] Conf. 1.11.17.

[[69]] Conf. 3.4.7.

[[70]] Conf. 5.14.24: 'statui ergo tamdiu esse catechumenus in catholica ecclesia mihi a parentibus commendata, donec aliquid certi eluceret, quo cursum dirigerem.'

[[71]] The Contra Academicos was written at Cassiciacum and bears the dramatic date of November 386, three months after Augustine's conversion in the garden at Milan. The Academic school had long been a convenient umbrella protecting intellectual pagans from the logical consequences of their agnosticism, as in the case of Caecilius in Minucius Felix's Octavius; cf. esp. chapter 13.

[[72]] As by P. R. L. Brown, Augustine of Hippo (London 1967) 41: 'Paganism meant nothing to Augustine.'

[[73]] The somewhat parallel case of Paulinus of Nola, depicted for us in his poignantly uncommunicative exchange of letters with Ausonius, has been sensitively studied in C. Witke, Numen litterarum (Leiden 1971). Ausonius himself (whose works are singularly empty of religion) is variously represented as pagan or Christian by many modern scholars; in the way made possible by our redefinition above, he was probably a bit of both. Augustine also resembled the distinguished rhetorician Marius Victorinus (PLRE 964), as much for the tardiness as for the eventual enthusiasm of his conversion.

[[74]] It does not seem to be merely an illusion fostered by conventions in the surviving evidence that the mystery cults in general had passed their peak of attraction by the late fourth century. One might have expected that the hidebound senatorial class would only begin to toy with such pastimes when their original vitality had for the most part already waned.

[[75]] Amm. 23.1.4.

[[76]] CIL 9.2566 (ILS 1253).

[[77]] Jerome, Chron. s.a. 371; Amm. 29.3.4. Pagans were known to accept Christian sanctuary more than once: Symmachus in 388, according to Socrates, HE 5.14.6; many pagans at the time of the sack of Rome in 410, according to Augustine, De civ. Dei 1.1: 'Sic evaserunt multi, qui nunc Christianis temporibus detrahunt et mala, quae illa civitas pertulit, Christo imputant.'

[[78]] PLRE 812-13 (Sebastianus 2).

[[79]] Athanasius, Hist. Ar. 61, followed by Theodoret, NE 2.13-14, and Socrates, HE 2.28. Pagans and heretics often made common cause; Augustine knew of the notorious Nicomachus Flavianus, who in his days as vicar of Africa (377) had been a supporter of Donatism (Aug., Ep. 87.8 [ca. 405]). Cf. also Aug., Serm. 62.14 (probably before 399): 'Haeretici, Iudaei et pagani unitatem fecerunt contra unitatem [nostram].' A Christianity rent by heresy and schism would be harder pressed to say how it differed from tolerant pagan cults.

[[80]] Libanius, Ep. 318 (A.D. 357).

[[81]] Amm. 23.3.5 et saep.; Eunapius, fr. 47 (in FHG 4).

[[82]] Amm. 30.10.3.

[[83]] Aug., Ep. 29-9: after so many persecutions and such violence, when peace came, mobs of gentiles wanting to become Christian were slowed by one thing: that they were accustomed to pass their feast days with their idols in banqueting and drinking, and they couldn't easily abstain from these pernicious and deeply ingrained pleasures. It seemed wise to our predecessors to spare this part of their weakness for a little while and to establish new feast days (for the ones they left behind) -- in honor of the holy martyrs, to be celebrated not with the same sacrilege but with a kind of similar lavishness. Now that they have been brought under the name of Christ and subdued the yoke of such great power, the health-giving commands of sobriety can be handed on to them, which they won't be able to resist out of respect and fear for the one who commands them. So now it's time that those who don't dare deny that they are Christian begin to live according to the will of Christ, that those things which were conceded to them so they would become Christians should, now that they are Christian, be spat out and rejected.

[[84]] The revised date of composition of the Saturnalia (ca. 431) proposed by A. Cameron, 'The Date and Identity of Macrobius,' Journal of Roman Studies 56 (1966) 25-38, is now widely accepted. Cameron's observation (ibid. 34) that the letters of Symmachus (published by 408) may have been a source for Macrobius could be borne out by a detailed comparison of rare late words used by both authors. Note, e.g., edecumo (Symm., Ep. 5.81 et alibi, used by Macrobius at Sat. 1.5.17 of a speech by Symmachus himself), anteloquium (Symm., Ep. 8.23, used by Macrobius at Sat. 1.24.21, in a speech by Praetextatus), and discussor (Symm., Ep. 5.76, also by Macrobius, In Somnium Scipionis 1.21.8). S. Döpp, 'Zur Datierung von Macrobius' "Saturnalia",' Hermes 106 (1978) 619-32, does not succeed in his attempt to disprove Cameron's arguments (largely through excessive reliance on arguments from silence), but may be said to have shown that further research on Macrobius (and Servius) might well bear fruit.

[[85]] This can be seen in the chart appended to Bloch (1945) after p. 244.

[[86]] CIL 6.1778. 6.1779 (= ILS 1259).

[[87]] Mac., Sat. 1.17.2-1.23.22.

[[88]] Plotinus is cited expressly by Praetextatus at Sat. 1.17.3.

[[89]] Sat. 1.17.2.

[[90]] J. F. Matthews, 'Symmachus and the Oriental Cults,' Journal of Roman Studies 63 (1973) 175-95. Cf. two invectives against unidentified pagan targets from the same period: the Carmen cod. Par. lat. 8084 and the Carmen adversus quendam senatorem, both exhibiting a wide variety of cults.

[[91]] Sat. 1.24.16; the long speech is at Sat. 3.1.1-3.9.16.

[[92]] Praetextatus is probably not identical with the hierophant of the same name who is alleged to have taken part in the dedication of Constantinople (an identification suggested by PLRE); if he were, his age and the dates of his cursus would be strangely unsynchronized.

[[93]] Amm. 22.7.6 records the appointment, noting that Praetextatus was in Constantinople on private business when Julian chose him personally for the proconsulship.

[[94]] Might Praetextatus have spent some time in Athens in his youth studying philosophy and imbibing the religious atmosphere, just as Julian did? It is not even inconceivable that he might have known Julian there. If we assume that his age upon attaining the post of urban prefect was a not untypical 40-45, he would have been only some five to ten years older than Julian.

[[95]] Himerius, Or. 51 (text now lost).

[[96]]CTh 9.16.7; the story of the law's origin is in Zos. 4.3.3.

[[97]] Amm. 27.9.9; Coll. Avell. 5-6; Sozomen, HE 6.23.2. A genial anecdote of this period (reported by Jerome, Contra Ioannem Hieros. 8) recounts that the pagan prefect remarked on the enthusiasm of the partisans in the papal election and the perquisites of the office, then averred he would convert to Christianity himself if he could be bishop of Rome. We can grasp that he may not have spoken altogether in jest, if we but think of Synesius of Cyrene and Cyrus of Panopolis. We also know of an undoubtedly pagan urban prefect of 365, C. Ceionius Rufius Volusianus, who had his name carved on a Christian basilica at Ostia in order to take credit for building it (see R. Meiggs, Roman Ostia [Oxford 1960] 388).

[[98]] Amm. 27-9-10.

[[99]] CIL 6.102 (= ILS 4003). Bloch (1945) 208 prudently connected this interest in the Dei Consentes to the syncretism of the speech attributed to Praetextatus in the Saturnalia and also to ideas of Sallustius, De diis et mundo (ed. A. D. Nock; Oxford 1926); this latter work, the paradigmatic statement of Julianic religion, was probably written by the consul of 363 (for discussion and references see PLRE 796 [Sallustius 1]).

[[100]] CTh 6.5.2 (21 May 384) and CJ 1.54.5 (9 Sept. 384) are the earliest and latest firm dates For his time in office. His scheduled consulship is mentioned both in Symm., Rel. 12.4, and in his inscriptions (CIL 6.1777 [= ILS 1258] and 6.1779 [= ILS 1259]).

[[101]] Symm., Rel. 21.5.

[[102]] A. Cameron, 'The Roman Friends of Ammianus,' Journal of Roman Studies 54 (1964) 15-28.

[[103]] Again it is necessary to state that there is no evidence of any censorship designed to please Christian authorities on matters of religion. The proof of this assertion has been set out by J. A. McGeachy, 'The Editing of the Letters of Symmachus,' Classical Philology 44 (1949) 222-29.

[[104]] Symmachus' substantive allusions to religion have often been catalogued before, most recently by J. F. Matthews, Journal of Roman Studies 63 (1973) 192-93. But even where a complete catalogue is given, one is left by earlier scholars with the impression that the collection is only a sampler of Symmachus' opinions. What is significant for our purposes is precisely that the complete catalogue is so brief and circumscribed.

[[105]] The use of religious terminology in matters of amicitia is noteworthy; see J. F. Matthews, Journal of Roman Studies 63 (1973) 177. The force which inspires Symmachus to some action called for by friendship is always religio (a dozen times in the 91 letters of book 3, for example); one friend commending another is acting as mystagogus (e.g., Ep. 5.64, where religio also appears); and compare the use of pontificium in Ep. 7.27. It is worth noting that Symmachus is a man more devoted to friendship than to friends. While almost all his correspondents are addressed as friends, they are worthy correspondents for the most part only so long as they hold office and have the power to use their influence on Symmachus' behalf (Ausonius is the most striking case of someone with whom Symmachus had nothing to do after his term of office expired). I would submit the following tentatively complete list of real friends (i.e., people who were more than potentially useful acquaintances) out of all the horde of correspondents: Symmachus' father (Epp. 1.1-12), Praetextatus (Epp. 1.44-55), Symmachus' brother Titianus (Epp. 1.62-74), Nicomachus Flavianus (book 2), Marinianus (Epp. 3.2329), Symmachus' son-in-law Flavianus the younger (book 6), his own son (Epp, 7.1-14), and Caecina Decius Albinus (Epp. 7.35-41), who was the son of the Caecina Albinus mentioned in Mac., Sat. 1.2.15 as a particular friend of Symmachus -- the same Caecina Albinus whom Jerome recounted as dandling a Christian granddaughter on his knee (see above, pp. 62-63). One might be tempted to add the brethren from Gaul (Protadius, Minervius, and Florentinus) whom Symmachus would have met when he was with the court at Trier in 369/370 and kept in vague touch with later -- but perhaps in this case they were exploiting him (Epp. 4.1757); and Magnillus, who may have been simply a neighbor of one of Symmachus' many great houses (Epp. 5.17-33). However generously compiled, the list is brief.

[[106]] Amm. 28.4.18 lampooned the aristocracy's posturings about the burdensomeness of travel, of which Symmachus' earnest remarks must be considered a part.

[[107]] Epp. 7.58, 9.28. There are perhaps ten such passages in all.

[[108]] E.g., Ep. 5.15: 'mecumque laeteris, quod d.n. Theodosii sacro divinoque iudicio merui consulatum.'

[[109]] Only one of the sixteen escapes these categories. Ep. 5.85 (A.D. 395 ?) is an invitation to Helpidius to celebrate a festival with Symmachus: 'oro iam venias et praesentia tua augeas honorem iestorum dierum.' But the tone of the letter indicates that the house party envisaged would be a holiday rather than a holy day.

[[110]] This text is evidence that non-Christians shared the suspicions of Christians of the sincerity of some of the new converts to the popular and powerful creed. Evidence that some Christians were not whole-hearted converts is thus bipartisan.

[[111]] It should be noted that the younger Flavianus is not elsewhere documented to have had a strong interest in religion, but one may assume that, for the benefit of his father's friends, he was at the very least willing to seem to share his father's enthusiasms. The popular assumption (based on Aug., De civ. Dei 5.26) that he made a politic conversion to Christianity is utterly baseless.

[[112]] Three important documents of earlier generations have not survived intact: the larger works against Christianity of Celsus (known through Origen's Contra Celsum), Porphyry, and Julian.

[[113]] Rel. 3.10. This sentence may have been a sly dig at the saying of Jesus, 'Ego sum via' (Jn. 14.6). At least the aged Augustine regretted having said something very like Symmachus' remark ('Ad sapientiae coniunctionem non una via perveniri' -- Soliloquia 1.13.23) In the late 380s, for just that reason (Retractationes 1.4.3).

[[114]] It was also while he was urban prefect, in 384, that Symmachus appointed the young Augustine to a chair of rhetoric at Milan (Conf. 5.13.23), thinking him still a Manichee -- an action consonant with the pagan tendency to encourage heresy (see note 79 above).

[[115]] This may be indicated by Rel. 3.1. and Amb., Ep. 17.10; the author in the earlier case need not have been Symmachus himself.

[[116]] Even though Damasus had been instrumental in opposing the first attempt to restore the Altar of Victory: Amb., Ep. 17.10.

[[117]] Rel. 21.5.

[[118]] Cameron, 'Roman Friends of Ammianus,' 15-28.

[[119]] T. D. Barnes, 'The Historical Setting of Prudentius' Contra Symmachum,' American Journal of Philology 97 (1976) 373-86. There is another text which has commonly been interpreted as representing yet another (a fourth ?) Symmachan embassy about the Altar of Victory, this one during his consulship in 391. The text is Quodvultdeus (Ps.-Prosper), Liber de promissionibus Dei 3.38.41 (PL 15.834), probably written during Quodvultdeus' episcopacy at Carthage (437-453). The text (as corrected by A. Chastagnol, Les Fastes de la Préfecture de Rome au Bas-Empire [Paris 1962] 227) is as follows: 'Cui Symmachus ille, mirabili eloquio et scientia praeditus, tamen paganus, praeconia laudum in consistorio recitans, subtili arte qua valuit, aram Victoriae in senatu restitui, Christiano, ut noverat, principi intimavit. Quem statim a suis aspectibus pulsum, in centesimo lapide, rhedae non stratae ea die, manere praecepit.' Chastagnol suggests the text refers to the gratiarum actio Symmachus would have delivered at Milan on 1 January 391 to begin his consulship, at which time he would have been hoping to profit from a momentary quarrel between Theodosius and Ambrose (on which see below). But there is no evidence for this, and it is difficult to see how an African bishop, writing fifty years after the fact, could have come by such an important fragment of information which was not preserved elsewhere (e.g., in the letters or life of Ambrose). It is far likelier, therefore, that Quodvultdeus is retailing the story of the famous Relatio 3, misinterpreting it as an oration delivered before the emperor (the text lends itself to that reading), and embroidering upon the emperor's negative response. The difficulty with taking Quodvultdeus' text literally, as Chastagnol noted, is that relegation to the hundredth milestone is a punishment attested at Rome, but not at Milan.

[[120]] Amb., Epp. 17, 18.

[[121]] Ambrose insisted on the point, which is central to my own argument, that even professed Christians could be pagan in their attitude toward religion. Ep. 17.8: 'Quod si aliqui nomine Christiani tale aliquid decernendum putant, mentem tuam vocabula nuda non capiant, nonmina cassa non fallant. Quisquis hoc suadet, sacrificat, et quisquis hoc statuit.'

[[122]] Symm., Ep. 2.36; CIL 6.2145 (= ILS 1261).

[[123]] Ep. 2.36.3. If this hypothesis is correct, who succeeded Praetextatus? Symmachus? Is that why his opinion matters in the controversy over the statue? (See Dessau's note to ILS 1261.)

[[124]] Bloch (1945) 217-19.

[[125]] Symmachus did not attempt to persuade Nicomachus Flavianus on this point but obviously expected him to share his attitude, a point neglected by Bloch. Note that the abdication of the pontificate, however, was the one point in Gratian's anti-pagan program which Symmachus did not deplore in Rel. 3: out of deference to his mentor's thereby enhanced dignity, surely.

[[126]] Symmachus' desire to step down and his later rustication have already been noted. Nicomachus Flavianus had served as quaestor sacri palatii in 381/2-383, then did not again enter public life until 389 as praetorian prefect (for these dates, see my article, 'The Career of Virius Nicomachus Flavianus,' Phoenix 32 [1978] 129-43).

[[127]] Amb., Ep. 41.27.

[[128]] Still recounted by J. Matthews, Western Aristocracies 227, 232-36.

[[129]] A. Alföldi, A Festival of Isis under the Christian Emperors of the IVth Century (Budapest -- Leipzig 1937); and Die Kontorniaten: Ein verkanntes Propagandamittel der stadtrömischen Aristokratie in ihrem Kampf gegen das christliche Kaisertum (Budapest -- Leipzig 1942-1943).

[[130]] Especially noteworthy is the neglected review of Alföldi's work on the contorniates by J. M. C. Toynbee, Journal of Roman Studies 35 (1945) 115-21.

[[131]] See again my article on Flavianus' career (note 126 above) for the sources of what follows.

[[132]] We know this from Symmachus' letters, from Flavianus' role in Macrobius' Saturnalia (where augury is his special interest: 1.24.17), and from his having translated Philostratus' Life of Apollonius of Tyana (Sidonius Apollinaris, Ep. 8.3.1).

[[133]] These allegations stem from the controversial Carmen cod. Par. lat. 8084, whose subject is thought by many to have been Nicomachus Flavianus (see most recently J. F. Matthews, Historia 20 [1970] 464-79); but I show in my article on Flavianus (note 126 above) that other candidates -- Praetextatus perhaps the most likely among them -- can be as convincingly proposed as subjects for the poem.

[[134]] Note that Bloch (1945) 232-33 wrongly states that Augustine attributed to Flavianus the theory that Christianity would collapse after 365 years; Bloch was followed hesitantly in this by Matthews, Western Aristocracies 245-46. The passage of Augustine cited in support of this idea (De civ. Dei 18.53) does not identify the person or persons promoting this notion. Porphyry may very well have been intended; see J. J. O'Meara, Porphyry's Philosophy from Oracles in Augustine (Paris 1959) 67-72, and cf. De civ. Dei 10.32.

[[135]] In an imperial decree of A.D. 431 recorded in stone by the son: CIL 6.1783 (= ILS 2948).

[[136]] Zos. 4.59.

[[137]] Zosimus' story would then offer a parallel to the distortion implicit in the account by Quodvultdeus (see above note 119). (S. Döpp, 'Theodosius I: Ein zweites Mal in Rom?' Apophoreta für Uvo Hölscher [Bonn 1975], is conclusive in the negative on this point.)

[[138]] Information in the next three paragraphs relies heavily on the resources of PLRE. I have chosen to follow their attributions of religious affiliation, where it is disputed, rather than those of A. Chastagnol, Fastes (note 119 above), since the latter scholar is quicker to assume a subject was Christian even when the evidence is open to doubt, following the assumption that the brother or even the father (e.g., Petronius Probinus, urban prefect in 345-346) of a Christian must have been a Christian as well; see also Chastagnol's over-eager assumption of Christianity for Viventius (urban prefect 366-367), Flavius Eupraxius (urban prefect 374), and Gabinius Vettius Probianus (urban prefect 377).

[[139]] He may have been a secret pagan (i.e., an apparent Christian) under Constantius, then openly pagan under Julian, and finally an Arian Christian at some time under Valens (PLRE 605-608). As always, paganism and Arianism are close to each other in this case, along with a willingness to change religions whenever convenient.

[[140]] I omit from my calculations cases where the prefect's religion is simply unknown. Chastagnol has a count of twenty Christians to twelve pagans; this is still a more sizable bloc of non-Christians than in any of the other posts.

[[141]] CTh 16.10.22 (9 April 423).

[[142]] A. Chastagnol, La Préfecture urbaine à Rome sous le bas-empire (Paris 1960) 165-66. But see A. Cameron, Claudian (Oxford 1970) 237-40. This détente is thought to have ended about 405/406, when Stilicho drew closer to Alaric and the Visigoths; CTh 16.10.19 (15 Nov. 407) is misinterpreted as evidence of this.

[[143]] Zos 5.41, Soz. HE 9.6; perhaps both authors follow the lost account of Olympiodorus. The sacrifices never took place.

[[144]] J. B. Bury, The Later Roman Empire (repr. New York 1958) I 394-402.

[[145]] Rutilius, De reditu 1.383-98, esp. 398: 'Victoresque suos natio victa premit.'

[[146]] Ibid. 1.441-52. Non-Christians (e.g., Julian and Eunapius) often found monks easy marks for satire.

[[147]] A trend culminating in A. Cameron, 'Rutilius Namatianus, St. Augustine, and the Date of the De Reditu,' Journal of Roman Studies 57 (1967) 31-39.

[[148]] Seneca's De superstitione, quoted verbatim by Augustine, De civ. Dei 6.11, perhaps echoed by Rutilius, 1.398. Seneca was not much read in non-Christian circles at this date; hence the assumption that Rutilius may depend here on Augustine.

[[149]] Rutilius also criticized Stilicho (De reditu 2.52-56) for burning the Sibylline books.

[[150]] Nonnus' poetical oeuvre (the Dionysiaca and a metrical version of the fourth gospel) may also represent broad-mindedness rather than a career leading from early paganism through conversion to Christianity.

[[151]] John Malalas 14.361. See J. B. Bury, Later Roman Empire I 227-28.

[[152]] P. R. L. Brown, 'Aspects of the Christianization of the Roman Aristocracy,' Journal of Roman Studies 51 (1961) 1-11, reprinted in his Religion and Society in the Age o! Saint Augustine (London 1972) 161-82.

[[153]] Matthews, Western Aristocracies 370.

[[154]] J. Sundwall, Abhandlungen zur Geschichte des ausgehenden Römertums (Helsinki 1919) 114-15.

[[155]] The publication of the Ordo generis Cassiodororum (H. Usener, Anecdoton Holderi [Bonn 1877]) proved that Boethius had written at least some of the theological tractates circulated under his name. V. Schurr, Die Trinitätslehre des Boethius im Lichte der 'skythischen Kontroversen' (Paderborn 1935), is the best study of the theological preoccupations of the class as a whole.

[[156]] On the former, see now my Cassiodorus (Berkeley 1979).

[[157]] P. Courcelle, Les Lettres grecques en occident de Macrobe à Cassiodore2 (Paris 1948) 257-312, esp. 299-300.

[[158]] E.g., the distinction between the perpetuity of the created world and the eternity of God, Consolatio Philosophiae (ed. Bieler; CCL 94) 5, prosa 6, lines 28-52.

[[159]] This aim is enunciated in the preface to the second book of the second edition of his commentary on Aristotle's De interpretatione (ed. C. Meiser; Leipzig 1877-1880).

[[160]] See A. W. J. Holleman, Pope Gelasius and the Lupercalia (Amsterdam 1974). The text of Gelasius' letter on the subject is edited and annotated by G. Pomares in Sources Chrétiennes 65 (1960).

[[161]] The variety of these trivial survivals is considerable. The pagan riots at Calama in 408 (Aug., Ep. 91) belong to this category. See also CJ 1.11.8 (A.D. 472), which ordered the confiscation of land on which pagan rites were carried out. There is evidence for statues of Minerva being restored as late as 483 (CIL 6.526, 6.1664). And see Eugippius, Vita sancti Severini 11.2 (on which see G. Alföldy, Noricum [London 1974] 212); Ennodius, Dictio 2 (CSEL 6.430-33); Greg. Mag. Ep. 5.38 (MGH ed.); and Cassiodorus, Var. 8.33.

[[162]] Christianity as a success religion becomes implicit in authors from Eusebius (and a host of imperial panegyrists), through Prudentius and Orosius, right down into the Middle Ages. See C. N. Cochrane, Christianity and Classical Culture (Oxford 1940) 184-86, 336-37. Augustine did not share this idea, but even his De civitate Dei facilitated its spread; see my article,' The Inspiration for Augustine's De civitate Dei,' Augustinian Studies 10(1979) 75-79.