^{page 316 note 1} Cf. Ensslin, W., Die Geschichtsschreibung und Weltanschauung des Ammianus Marcellinus (Klio, Beiheft xvi [1923], 91 f.)Google Scholar, and Thompson, E. A., The Historical Work of Ammianus Marcellinus (1947), pp. 111 f.Google Scholar See also Pighi, G. B., Reall. f. Antike u. Christentum i (1950), 386 f.,Google Scholar for a useful collection of all the passages in Ammianus relating to Christianity. We are grateful to Professor A. D. Momigliano for reading an earlier draft of this paper, which was read in a slightly different form before the Glasgow University Alexandrian Society on 28 November 1963.

^{page 316 note 2} Gibbon's warm praise of Ammianus' impartiality in this respect evoked from one contemporary critic the observation that ‘Mr. Gibbon shews, it is true, so strong a dislike to Christianity as visibly disqualifies him from that society of which he has created Ammianus Marcellinus president' (Porson, , Letters to Travis [1790], pref. xxviii).Google Scholar See also below p. 322.

^{page 316 note 3} Cf. Ensslin, , op. cit. (n. 1), p. 96:Google Scholar ‘wo er vom Christentum und von Christen erzählen muβte, tut er das in einer Form, die deutlich macht, daβ er es als etwas ihm fremdes betrachtet wissen will, dies besonders in der Art, wie er Fachausdrücke anführt.’

^{page 316 note 4} Phil, and Hist. Commentary on Ammianus Marcellinus, 15. 6–13 (1953), p. 22:Google Scholar though he does add that the tendency ‘to avoid technical expressions (as does Tacitus) for stylistic reasons, plays a part’. It will be seen that we go much farther than this: see below.

^{page 316 note 5} Life and Letters in the Fourth Century (1901), pp. 40 f.Google Scholar Cf. also Labriolle, P. de, La réaction païenne (1934) p. 436Google Scholar: ‘Remarquable est la gaucherie, la vague dont il se sert dès qu'il en parle; on dirait qu'il ignore les termes techniques …‘.

^{page 316 note 6} The Latin Language (1954), p. 205,Google Scholar in a discussion of the spread of ‘Christian Latin’—on which see below, p. 323.

^{page 316 note 7} That is to say writers of history in the manner of the ‘classical’ historians, Herodotus, Thucydides, Sallust, Tacitus—to be sharply distinguished from writers of epitomes, chronicles, and ecclesiastical histories, who all had different aims and different techniques: see the discussion of Momigliano, A. D., in The Conflict between Paganism and Christianity in the Fourth Century, edited by Momigliano, A. D. (1963), pp. 79–99.Google Scholar

^{page 317 note 1} See the famous passage in Gregory of Nyssa's, Oratio de deitate Filii et Spiritus Sancti, P.G. xlvi col. 557, corroborated for the sixth century by Agathias, 2. 29.Google Scholar

^{page 317 note 2} is surely not, as Rubin maintains (R.-E. xxiii. 1. 329Google Scholar), ironic. The word is used in its regular meaning of ‘temperate’, ‘chaste’ (L.S.J., s.v. 11): Theophylact, loc. cit., talks of the monks'

^{page 318 note 1} See below, pp. 319 f.

^{page 318 note 2} Vossius, , de hist. gr. p. 324Google Scholar ed. Westermann, cited by Teuffel, W. S., ‘Agathias von Myrine’, Philologus i (1846), 507 n. 52.Google Scholar

^{page 318 note 3} Though M. Apostolopoulos, (Diss. Athens, 1894), p. 33 points out a reminiscence of Matt. xvi. 26 at 3. 12, p. 165. 22, it is the more striking for being exceptional.

^{page 318 note 4} Costanza, S., ‘Orientamenti Cristiani della Storiografia di Agatia’, Helikon ii (1962), p. 93Google Scholar seems misguided in citing this story as evidence for Agathias' religious orthodoxy. He takes the description of Hagia Sophia (5. 9, p. 295. 8, cited above) in the same way. Apostolopoulos however (op. cit. [n. 3], pp. 38–39) rightly attributes the artificiality of the passage to Agathi’ being one of the class of ‘Profanhistoriker’,

^{page 319 note 1} Bury, , Latin Roman Empire, ii² (1923), p. 418 n. 2,Google Scholar and R. Laqueur, R.-E. s.v. Malchos, 856, followed by Stein, , Bos-Empire, ii (1949), P. 708n. 1.Google Scholar

^{page 319 note 2} Saffrey, H.-D., R.É.G. lxvii (1954), 396 n. 2.Google Scholar

^{page 319 note 3} Or at any rate all that was written, for Photius indicates that the work was unfinished. The last sentence of fr. 4 (H.G.M. i. 390. 20) surely has a distinctly Christian character: Pagans found the Christian doctrine of a God of wrath particularly repugnant. The Apostate Julian, for example, exclaimed indignantly (Contra Christianas, ed. Neumann [ 1880], p. 190. 5 f.). Early Christian writers found the point so awkward that they allegorized it away: Lactantius, however, in his De ira Dei made the Christian point of view quite clear (cf. de Labriolle, P., Hist, de la lilt. lat. chrét. I³ [1947], pp. 307 f.)Google Scholar. The recognition of Malchus' Christianity drives one more nail in the coffin of Bury's theory that the works of the fifth-century historians were allowed to perish because of the paganism of their authors: cf. Stein, loc. cit.

^{page 319 note 4} For Procopius' Christianity cf. Rubin, , op. cit. (p. 317, n. 2), 329 f.Google Scholar

^{page 319 note 5} Cf. Dahn, F., Procopius von Caesarea (Berlin, 1865), pp. 189–90.Google Scholar The scribes of the Mediceus of Eunapius and the Vossianus of Julian had more justification for their indignant marginalia.

^{page 319 note 6} In fact the apparent indifference shown to both paganism and Christianity in such passages led some older critics to the con clusion that Procopius was a Jew: cf. Dahn, , op. cit. (n. 5), p. 193,Google Scholar and also Stein, , Bas-Empire, ii. 716 n. 1.Google Scholar

^{page 320 note 1} Further exx. in Downey, G., ‘Paganism and Christianity in Procopius’, Church History xviii (1949), 91 nn.Google Scholar 11 and 12. Cf. 3. 4. 9, 3. 19. 6, 7. 21. 10.

^{page 320 note 2} Cf. Apostolopoulos, , op. cit. (p. 318 n. 3), pp. 33 f.Google Scholar, and more recently, Cameron, Averil, ‘Agathias and Cedrenus on Julian’, J.R.S. liii (1963), 92 n. 9.Google Scholar

^{page 320 note 3} Cf. Krumbacher, K., Gesck. d. Byz. Lit. ² (1897), p. 250 n. 1Google Scholar ‘er ist Christ, und lebt in einem völlig christlichen Zeitalter, schreibt aber von christlichen Dingen wie ein Heide’.

^{page 320 note 4} As also does Paul the Silentiary; cf. Friedländer, P., Jo. v. Gaza u. Paul Sil. (1912), p. 124 n. 3.Google Scholar

^{page 320 note 5} Agathias himself complained that he had no time to study the ancient (3. I, p. 138. 12). On (imitatio) as one of the of history cf. Peter, H., Gesch. Lit. über die röm. Kaiserzeit, ii. 188.Google Scholar In its stylistic aspect it took the form exemplified in the words of Aristides, Rhet. 2. 10 The epigrams of Agathias and his circle are similarly backward-looking —cf. Waltz, P., L'Acropole vi (1931), 1–21,Google ScholarKeydell, R., Reallexikonf. Ant. u. Christentum, v (1963) 549.Google Scholar

^{page 321 note 1} Cf. e.g. fr. II, p. 25. 16 D, fr. 57, p. 111 D. At fr. 11, p. 24. 25 and 27 D he makes his own Christianity plain.

^{page 321 note 2} For which cf. e.g. Syncellus p. 316 (Cephalion).

^{page 321 note 3} See above, p. 316 n. 7.

^{page 321 note 4} Dexippus and Herodian, for example. Cf. Stein, Franz Joseph, Dexippus et Herodianus rerum scriptores quatenus Thucydidem secuti sint (Diss. Bonn, 1955)Google Scholar, with detailed bibliography.

^{page 321 note 5} Costanza, , op. cit. (p. 318 n. 4), p. 108.Google ScholarDowney, Similarly, op. cit. (p. 320 n. 1), p. 98Google Scholar sees Procopius' conception of as Platonic. One could add that it is in fact Neo-platonic as well (cf. Courcelle, P., Les Lettres grecques en Occident de Macrobe à Cassiodore² [Paris, 1948], pp. 17f.Google Scholar on Macr. Sat. 5. 16. 8, p. 313. 5, and 288 on Proclus, , De prov. et fato ed. Cousin, p. 158. 8Google Scholar) but it is more consistent with the rest of the practice of these two writers to explain their frequent references to as yet another example of this time probably of Polybius, rather than by reference to philosophical systems.

^{page 321 note 6} Veh, O., ‘Der Geschichtsschreiber Agathias von Myrine’, Wiss. Beilage zum Jahresbericht 1952/3 des Gymnasiums Christian-Ernestinum Bayreuth, p. 27.Google ScholarPertusi, A., ‘L'atteggiamento spirituale della più antica storiografia bizantina’, Aevum xxx (1956), 151–2Google Scholar, expresses a similar opinion.

^{page 321 note 7} Cf. Stein, , Bas-Empire, ii. 715.Google Scholar

^{page 321 note 8} Bury, , Later Roman Empire ii² (1923), 427 n. 2.Google Scholar

^{page 321 note 9} 2. 29, p. 127. 15 f.

^{page 322 note 1} For Procopius cf. Braun, H., Procopius Caesariensis quatenus imitatus sit Thiuydidem (Diss. Erlangen, 1885Google Scholar [ = Acta seminarii Erlangeniis iv (1886), 161–221]Google Scholar). Braun's examples are by no means wholly vitiated by the objections of J. Haury, Zur Bewteilung des Geschichtsschreibers Procopius von Cäsarea, Progr. Wilhelm-Gymn, K.. (München, 1896/1897)Google Scholarinit. For Agathias: Franke, G., ‘Quaestiones Agathianae’, Breslauer Philol. Abhandl xlvii (1914), 28 f.;Google ScholarCameron, Averil, ‘Herodotus and Thucydides in Agathias’, Byz. Zeit lvii (1964).Google Scholar

^{page 322 note 2} C.Q. xxxix (1945), 92f.Google Scholar In fact Thompson argues that Priscus copied Dexippus' imitation of Thucydides as well as Thucydides himself. Similarly Agathias derives some of his Herodotean imitations from Procopius as well as from Herodotus himself.

^{page 322 note 3} On battle pieces as a see Peter, , op. cit. (p. 320 n. 5), ii. 312 f.Google Scholar, and Naudé, C. P. T., Acta Classica i (1958), 92 f.Google Scholar

^{page 322 note 4} Especially in the case of Procopius: cf. Braun, H., Die Nachahmung Herodots durch Prokop (Progr. Nürnburg, 1894), pp. 40f.;Google ScholarRubin, B., R.-E. xxiii. 1 (1957), 330 f.;Google Scholar cf. also Downey, , op. cit. (p. 320 n. 1), pp. 89 f.;Google ScholarStein, , Bas-Empire, ii. 716, n. 1.Google Scholar

^{page 322 note 5} Agathias, for example, says not a word of the Ecumenical Council contemporary with the events he is describing: contrast Cedrenus i, pp. 659 f. Agathias and Procopius carried out to the full the policy outlined by Apostolopoulos, , op. cit. (p. 318 n. 3), p. 35;Google Scholar they were writing ‘Profangeschichte’,

^{page 322 note 6} See especially the careful study of Pighi, G. B., ‘Latinitá cristiana negli scrittori pagani del IV secolo’, i, Studi dedicati alia memoria P. Ubaldi (1937), pp. 41 f.,Google Scholarsummarized in Reall. f. Ant. u. Christentum i (1950), 394.Google Scholar

^{page 323 note 1} A phrase that recalls Tertullian's observation that the common pagan mis-spelling ‘Chrestus’ for ‘Christus’ was, though incor rect, at least a name ‘de suavitate et benignitate compositum’ (Apol. 3).

^{page 323 note 2} Cf. Mohrman, C., Études sur le latin des chrétiens i (1958), pp. 255 f.;Google Scholar cf. also Lietzmann, H., History of the Early Church, iii (Eng. Tr. 1949), 314 f.Google Scholar

^{page 323 note 3} Compare the curiously similar periphrasis with which Procopius introduces the same word (5. 22. 10): Procopius' reason, of course, was that the word was Latin.

^{page 323 note 4} Historical Work (p. 316 n. 1), p. 114.

^{page 324 note 1} ‘Aulus Gellius und Ammianus Marcellinus’, Hermes viii (1874), 257–302;Google Scholar one or two additions by Fletcher, G. B. A., Rev. de Phil. xi ³ (1937), 393.Google Scholar

^{page 324 note 2} Studia Ammianea (Milan, 1935), p. 83.Google Scholar

^{page 324 note 3} Pp. vi f. of vol. i of his edition (1910).

^{page 324 note 4} Cf. Clark, loc. cit., or Rolfe, Loeb edition, i, pref. xxii f.

^{page 324 note 5} Like most Latin writers; see Townend, G. B., Hermes Ixxxviii (1960), 98–99.Google Scholar

^{page 324 note 6} Op. cit. (p. 322 n. 6), p. 50.

^{page 324 note 7} Op. cit. (p. 316 n. 1), p. 17.

^{page 324 note 8} Die griech. Lit. des Altertums; Kultur d. Gegenwart, i. 8² (1907), p. 201.Google Scholar

^{page 324 note 9} Many examples cited by Brok, M. F. A., De Perzische Expeditie van Keizer Julianus volgens Ammianus Marcellinus (Diss. Leiden, 1959), p. 179.Google Scholar As for Ammianus' transliteration into Latin of Greek medical terms (Thompson, p. 17 n. 1), this was the practice of his contemporary, the famous physician Vindicianus Afer (R.-E. ix A 1. 30) and other Latin writers of technical treatises faced with Greek terms for which there was no Latin equivalent.

^{page 325 note 1} Rolfe, Loeb edition pref. xix, writes that Ammianus' ‘readers and hearers were, of course, utriusque linguae periti’. But Q,.Aurelius Symmachus, one of the most distinguished figures in the literary world of late fourth-century Rome, had very little Greek (cf. Cameron, Alan, J.R.S. liv [1964])Google Scholar, and the same is probably true of most of his contemporaries; see on the whole question Courcelle's, P.Lettres grecques (p. 321 n. 5) ch. i passim.Google Scholar

^{page 325 note 2} Macrobius, on the other hand, alway introduces Greek words with the formula ‘u Graeci appellant’ (Sat. 1. 4. 9, 7. 9. 6).

^{page 326 note 1} Kroll, W., Glotta xv (1927), 299 n. 2.Google Scholar

^{page 326 note 2} Syme, R., Tacitus, i (1958), pp. 343 f.Google Scholar

^{page 326 note 3} For Ammianus' imitations of Tacitus cf. Fletcher, G. B. A., Rev. de Phil. xi ³ (1937), 389 f.Google Scholar

^{page 326 note 4} Studien zum Verständnis der römischen Literatur (1924), p. 113.Google Scholar For further remarks on Ammianus' unreliability in matters of technical terminology cf. Ensslin, , Klio xxxii (1939), 93–94.Google Scholar

^{page 326 note 5} Müller, A., ‘Militaria aus Ammianus Marcellinus’, Philologus lxiv (1905), 575.Google Scholar

^{page 326 note 6} Agathias' inexactitude in such matters gives rise to several prosopographical prob lems: cf. Stein, , Bas-Empire, ii. 814 f.Google Scholar, Excursus M. By contrast Olympiodorus of Egyptian Thebes, who was writing not history in the strict sense, but as his title indicates, , admitted all the Latin official titles, even (fr. 13 init.) and (fr. 23): cf. Thompson, E. A., ‘Olympiodorus of Thebes’, C.Q. xxxviii (1944), 48.Google Scholar

^{page 326 note 7} Historia iv (1955), 52Google Scholar ‘the Virgins of Vesta’.

^{page 326 note 8} Though many scholars maintain that what Tacitus wrote here is ‘Chrestianos’ and that by ‘quos vulgus appellabat’ he was contrasting the ignorance of the common people with his own knowledge that die name was really ‘Christian!’ (e.g. most recently Fuchs, H., Vigiliae Chrisiianae iv [1950], 72 f.Google Scholar, and Hanslik, R., Wiener Studien lxxvi [1963], 100f.)Google Scholar. Perhaps Ammianus' usage clarifies that of Tacitus: no particular credit would accrue to a Roman senator from knowing the exact name of this obnoxious Oriental sect. He might well, like our modern judges, affect a certain ignorance on such a point, Köster-man's latest Teubner edition rightly gives ‘Christianos’.

^{page 327 note 1} For examples see Goelzer, H., Grammaticae in Sulpicium Severum observationes (Paris, 1883), p. vi and n. 5.Google Scholar

^{page 327 note 2} Laistner, M. L. W., The intellectual heritage of the early Middle Ages (1957), p. 12.Google Scholar It seems, however, to have been known in medieval Ireland; see Bieler, L., J.T.S. x ² (1959), 412.Google Scholar

^{page 327 note 3} Stein, , Bas-Empire, i. 215.Google Scholar

^{page 327 note 4} Ammianus' position in the Graeco-Roman historiographical tradition has recently been examined in detail by Naudé, C. P. T., Ammianus Marcellinus in die lig van die antieke geskiedskrywing (Diss. Leiden, 1956)Google Scholar—with the reservations of Browning, R., J.R.S. xlvii (1956), 209 f.Google Scholar—and in Acta Classica i (1958), 92f.Google Scholar

^{page 327 note 5} See Cameron, Averil, Byz. Zeit. lvii (1964).Google Scholar

^{page 327 note 6} Cf. Ensslin, , op. cit. (p. 316 n. 1), pp. 35–36,Google ScholarLaistner, M. L. W., The Greater Roman Historians (1947), pp. 146–8.Google Scholar

^{page 327 note 7} For the traces in Ammianus' work of ‘sermo castrensis’ cf. Pighi, , Studia Ammianea (1935), pp. 65 f., 173 §7.Google Scholar

^{page 328 note 1} Noted and discussed as they occur in de Jonge's Commentary on Books 14 and 15(16 forthcoming), and works there cited.

^{page 328 note 2} Thompson, , C.Q. xxxviii (1944), 43 n. 2CrossRefGoogle Scholar, following Mendelssohn's edition of Zosimus, p. xxxv n. 1, denies the claim of von Gutschmid, A., Kleine Schriften, v. 412,Google Scholar that Eunapius was the founder of this school. However this may be, as at least a substantial part of Eunapius' history may well have been published before Ammianus' history (Chalmers, W. R., C.Q. N.S. X [1960], 152 f.Google Scholar, cf. Cameron, Alan, C.Q. N.s. xiii [1963], 216)Google Scholar, it is surely very likely that it was Eunapius who started the fashion of referring to Christianity in the manner discussed in this article—an hypothesis that would account for the striking parallelism between Ammianus and the fifth- and sixth-century Greek historians in this respect. Such periphrases do in fact occur in the curtailed version of Eunapius preserved by Zosimus, e.g. at 4. 40. 5, 5. 19. 5, and cf. Eunapius, Vit. Soph. 476.Google ScholarThompson, , op. cit. 52,Google Scholar suggests that it was through the medium of Olympiodorus that Ammianus influenced the Greek historians, but hardly in this respect, for Olympiodorus eschewed such purism (see above, p. 326 n. 6).