¹ Evidence for the sixth century is discussed by Cameron, Averil, Flavius Cresconius Corippus, In laudem Iustini Augusti minoris (London, 1976), pp. 199f.Google Scholar, and Stache, U. J., Flavius Cresconius Corippus, In laudem lustini Augusti minoris, Ein Kommentar (Berlin, 1976), pp. 7–19Google Scholar. Recent discussion on the seventh century by Baldwin, Barry, ‘Theophylact's Knowledge of Latin’, Byzantion 47 (1977), 357–60Google Scholar, with Whitby, L. M., ‘Theophylact's Knowledge of Languages’, Byzantion 52 (1982), 425–8Google Scholar. On an earlier period, see Fisher, Elizabeth A., ‘Greek Translations of Latin Literature in the Fourth Century a.d.’, YCS 27 (1982), 173–215Google Scholar, with bibliography at p. 174, n. 3.

² Whitby, loc. cit., p. 428, ‘One has to conclude that there is no firm evidence for or against Theophylact's knowledge of Latin’. Likewise Pertusi, A., Giorgio di Pisidia, Poemi (Ettal, 1959), i. 38. n. 1Google Scholar, suggests that similarities between the poems of George of Pisidia and Claudian (on which see Nissen, T., ‘Historisches Epos and Panegyrikos in der Spätantike’, Hermes 75, 1940, 298–325Google Scholar) are more likely to be due to a common debt to rhetorical tradition than to direct influence; cf. Antès, S., Corippe, Éloge de l'empereur Justin II (Paris, 1981), p. lii, n. 2Google Scholar, on George's independence of Corippus.

³ Cameron, Averil, ‘A Nativity Poem of the Sixth Century a.d.’, CP 74 (1979), 227Google Scholar, ‘…it is becoming clear that in general there was more Latin surviving in Constantinople even at the end of the sixth century than is commonly supposed…’; cf. JRS 71 (1981), 183–6Google Scholar. Much of the evidence, however, relates to the western exiles who resided in the eastern capital at this time, and to the high-ranking recipients of Pope Gregory's letters, who may have relied on interpreters. S. Antès (op. cit. in note 2, pp. xxxii–xxxv) is less sanguine, believing that Corippus' Latin poem ‘…était destinée à un milieu bien précis, celui du personnel politique et militaire de haut rang et des fonctionnaires de la chancellerie: its étaient en effet les seuls, avec les universitaires à, pouvoir, dans la partie orientale de l'Empire, comprendre et utiliser le latin… ’ (pp. xxxivf.).

⁴ E.g. Gibbon, 's oft-quoted note (Decline and Fall, ch. II, ed. Bury, J. B., 2nd ed., London, 1909, i. 42, n. 45)Google Scholar ‘There is not, I believe, from Dionysius to Libanius, a single Greek critic who mentions Virgil or Horace. They seem ignorant that the Romans had any good writers’; Maas, P. in BZ 35 (1935), 385Google Scholar ‘Einfluss lateinischer Dichtung auf griechische ist, soviet ich weiss, für die Zeit vor dem XIII. Jahrh. nirgends erwiesen’. This view is contested with equal vehemence by Keydell, R., for example in Gnomon 11 (1935), 605Google Scholar, ‘Überhaupt haben wir nun für jeden spätgriechischen Dichter — zu erinnern ist besonders an Paulus Silentiarius — mit der Kenntnis lateinischer Literaturwerke zu rechnen.’

⁵ So, for example, Vian, F., REG 77 (1964), 370Google Scholar, ‘Le problème concerne toute la littérature grecque d'époque impériale; mais disons tout de suite qu'on ne peut le résoudre sans tenir compte des époques ni des individus’; Schulz-Vanheyden, E., Properz und das griechische Epigramm (Diss., Münster, 1969/1970), p. 159Google Scholar, ‘Die Frage, ob Kenntnis der lateinischen Dichtung vorliegt, ist für jeden spätgriechischen Dichter gesondert zu behandeln.’

⁶ Vian, F., Recherches sur les Posthomerica de Quintus de Smyrne (Paris, 1959), pp. 95–101Google Scholar, and Quintus de Smyrne, La suite d' Homère (Paris, 1963), i. xxxii–xxxvGoogle Scholar, criticised by Keydell, R. in Gnomon 33 (1961), 278–84Google Scholar, but followed by Campbell, Malcolm, A Commentary on Quintus Smyrnaeus Posthomerica XII, Mnemosyne Supplement 71 (Brill, 1981), 117f. and n. 51Google Scholar; Gerlaud, B., Triphiodore, La prise d' Ilion (Paris, 1982), pp. 41–7Google Scholar.

⁷ Schott, G., Hero und Leander bei Musaios und Ovid (Diss., Köln, 1957)Google Scholar; Kost, K., Musaios, Hero und Leander (Bonn, 1971), pp. 21–3Google Scholar; Gelzer, T. and Whitman, C., Musaeus, Hero and Leander (Loeb, London, 1975), pp. 304–7Google Scholar.

⁸ Braun, J., Nonnos und Ovid (Greifswald, 1935)Google Scholar, and ‘Nonno e Claudiano’, Maia 1 (1948), 176–93Google Scholar, with Keydell, R. in Gnomon 11 (1935), 597–605Google Scholar and Maas, P. in BZ 35 (1935), 385–7Google Scholar; d'Ippolito, G., Studi nonniani (Palermo, 1964)Google Scholar, with Keydell, in Gnomon 38 (1966), 25–9Google Scholar and Vian, in REG 77 (1964), 369–71Google Scholar; Vian, F., Nonnos de Panopolis, Les Dionysiaques (Paris, 1976), i. xv and xlvif.Google Scholar; Diggle, J., Euripides, Phaethon (Cambridge, 1970), 180–200Google Scholar.

⁹ Viansino, G., Paolo Silenziario, Epigrammi (Turin, 1963), pp. xii–xvGoogle Scholar, with Cameron, Averil in JHS 86 (1966), 210f.CrossRefGoogle Scholar; E. Schulz-Vanheyden, op. cit. in note 5 above, pp. 156–69, supported by Kenney, E. J. in CR 86 (1972), 111Google Scholar, but criticised by Yardley, J. C., ‘Paulus Silentiarius, Ovid, and Properties’, CQ n.s. 30 (1980), 239–43CrossRefGoogle Scholar. Cf. Cameron, Alan, Porphyrius the Charioteer (Oxford, 1973), p. 88, n. 1Google Scholar, ‘I am not myself at all convinced that any of the Agathian poets were familiar with Latin literature’ (with bibliography).

¹⁰ For Paul's aristocratic background and devotion to culture, see Agathias, , Historiae 5.9.7 (ed. Keydell, R., Berlin, 1967)Google Scholar. As a silentiary he was a member of the imperial cubiculum and would have officiated at meetings of the imperial consistory and on ceremonial occasions; see Jones, A. H. M., The Later Roman Empire (Oxford, 1964), pp. 571f.Google Scholar; Guilland, R., Titres et fonctions de l'empire byzantin (Paris, 1976), XVIIGoogle Scholar; Bury, J. B., The Imperial Administrative System in the Ninth Century (London, 1911), pp. 24fGoogle Scholar.

¹¹ Edited by Friedländer, Paul, Johannes von Gaza und Paulus Silentiarius (Leipzig and Berlin, 1912; repr. Olms, Hildesheim and New York, 1969)Google Scholar. Friedländer's text is reproduced, and a German translation added, by Veh, O., in appendix to Procopius, de Aedificiis (Munich, 1977)Google Scholar.

¹² A detailed account of the earthquake is given by Agathias, , Hist. 5.3–9Google Scholar; cf. Malalas, J., Chronographia (ed. Dindorf, L., Bonn, 1831), pp. 488.20–489.10 and 489.19–490.5Google Scholar, Theophanes, , Chronographia (ed. de Boor, C., Leipzig, 1883), pp. 231.14–232.6 and 232.27–233.3Google Scholar.

¹³ Agathias, perhaps under the influence of Paul, likewise stresses the Emperor's primary concern for Sophia, S. (Hist. 5.9.2)Google Scholar.

¹⁴ On Sidonius, see further below, note 32. Claudian's use of Roma was probably influenced by Symmachus, Rel. 3.9f.Google Scholar, where She appeals to the Emperor to allow the continuation of her cult; see Cameron, Alan, Claudian (Oxford, 1970), p. 365Google Scholar. Other later Latin writers also make use of personified Roma, e.g. Prudentius, Contra Symm. 2.80–90, 634–768Google Scholar (where She champions Christianity, in answer to Symmachus' portrait of her); Namatianus, Rutilius, de reditu suo 1.47–164Google Scholar; Ennodius, , Pan. Theod. (ed. Hartel, , C.S.E.L., VI, Vindob., 1882), pp. 274.11–13, 276.20f.Google Scholar, Libell. pro synodo, p. 326.24 ad fin. (where a speech of appeal is put into her mouth). But these passages are not closely analogous to Paul's scene. The Latin writers refer to Old Roma, although New Roma is invoked by Sidonius, at Pan. Anth. 30ffGoogle Scholar. After the mid-fourth century, New Roma is frequently depicted in the East with the same attributes as Old Roma; see Toynbee, J. M. C., ‘Roma and Constantinopolis in late-antique art’, JRS 37 (1947), 135–44Google Scholar and (part II) in Studies presented to D. M. Robinson (Missouri, 1953), ii. 261–77Google Scholar.

¹⁵ The portrayal of Roma in distress goes back to Lucan 1.185ff., where She appeals to Caesar at the Rubicon in a nocturnal vision. There, and in Claud. Gild., her distress is associated with old age (Lucan 1.188 canos…crines; cf. Gild. 25; Roma is old also in Symm. Rel. 3.9f., Prudent. Contra Symm. 2.81, Rut. 1.115f. and Ennod. Pan. Theod. locc. citt.). This would, however, be inappropriate for Paul's New Roma, who is considered to be the daughter of Old Roma (P. Sil. Descr. 151, 166f.; cf. Libanius, , Ep. 972.5, 11.107.16f.Google Scholar Foerster, etc.).

¹⁶ Claud. Stil. 2.270–77 has a briefer reference to Roma's arms. In Corippus, , Laud. Just. 1.288–90Google Scholar (written a few years after Paul's poem) Old Roma is depicted in a traditional pose on Justinian's funeral pall; see Averil Cameron ad loc. for bibliography of this characteristic depiction.

¹⁷ The closest parallel is Sidon. Pan. Mai. 13ff., where the scene with Roma follows immediately upon a prefatory celebration of Maiorianus' consulship.

¹⁸ Cf. Il. 5.290, Nonn. D. 4.307, al.

¹⁹ Cf. also Claud. Stil. 2.407 haec dum Roma refert, iam Fama loquacibus alis… and Gild. 213, where the theme is abruptly changed, but without a verb of saying.

²⁰ P. Sil. Descr. 226f. πάντα κυβερνητ⋯ρι τεῷ διέπουσα χαλινῷ | ὑμετέροις ὑπέθηκα τροπαιοɸόροισι θριάμβοις; Claud. Eutr. 1.391f. quantum to principe possim,| non longinqua docent; cf. 384f., on Roma's maternal emotions at the sight of Honorius negotiating with the Germans.

²¹ Cf. also Claud. Gild. 31ff. and Sid. Pan. Avit. 72ff. (Roma details her previous conquests by contrast with her present plight); Sid. Pan. Anth. 440ff. (Roma reassures Aurora that She has not come to reclaim her eastern conquests); also Symm. Rel. 3.9 (the cult of Roma gave her world empire), Rut. 1.55ff. (eulogy, addressed to Roma, of her peaceful world empire). Paul's Roma mentions specifically Medes and Celts (228 ἠρεμέει κα⋯ Μ⋯δος ἄναξ κα⋯ Κελτ⋯ς ⋯μοκλή), India (229f. κα⋯ ξίɸος ὑμετέροις ɸιλοτήσιον ⋯πασε θώκοις | Ἰνδ⋯ς ⋯νήρ,…) and Carthage (231), in phrases reminiscent of Claud. Stil. 2.286 cecidit Maurus, Germania cessit or VI Cons. 415f….cum foedera Medus et Indus | hinc peteret pacemque mea speraret ab arce. The linguistic similarity cannot, however, be pressed. Paul's account of Justinian's peaceful world empire (cf. also Descr. 11–16, 135–60, 922f., 935–7) is the characteristically rosy view of panegyrists of the 560s; cf. A.P. 4.3B.1–100, 9.641 (both Agath.), 16.72 (Anon.), Cor. Laud. Just. praef. 1–36, with Averil Cameron's note, pp. 118f.

²² Roma uses the word ⋯λβον (234), thereby neatly returning to the theme from which her survey of Justinian's empire began, a reference to the hand of the Emperor as ῥυηɸενέος τροɸ⋯ν ⋯λβου (225).

²³ Malalas p. 488.6–14, Theophanes p. 230.17–25.

²⁴ Malalas p. 492.11–16, Theophanes p. 237.7–12. Cf. Proc. Aed. 5.1.7–16 on the need for a favourable south wind for the grain ships to pass through the Hellespont. Paul's poem was delivered in January 563, so the subject was topical.

²⁵ Other passages in Paul's poem celebrate Justinian's world empire (see note 21 above), but make no mention of the theme of Constantinople as a centre of trade, which is elsewhere only obliquely suggested in a description of how sailors steer towards Constantinople by the lights of S. Sophia (906–20).

²⁶ Gild. 26f. attigit ut tandem caelum genibusque Tonantis | procubuit; Sidon. Pan. Avit. 50 utque pii genibus primum est adfusa Tonantis.

²⁷ See Alföldi, A., ‘Die Ausgestaltung des monarchischen Zeremoniells am römischen Kaiserhofe’, Mitt. deutsch. arch. Inst., röm. abt. 49 (1934), 3–118Google Scholar, repr. in Alföldi, , Die monarchische Repräsentation im römischen Kaiserreiche (Darmstadt, 1970)Google Scholar, with comment and criticism by Avery, W. T., ‘The Adoratio Purpurae and the Importance of the Imperial Purple’, Mem. Amer. Acad. Rome 17 (1940), 66–80CrossRefGoogle Scholar; also Guilland, R., Recherches sur les institutions byzantines (Berliner byz. Arbeiten, 35, Berlin, 1967), i. 144–50Google Scholar. The prostration of Roma is a mark of the greatest extremity, since Alföldi notes (p. 43) that in allegorical representations, Roma alone among divine personifications retained throughout the empire the exceptional privilege of remaining seated in the Emperor's presence.

²⁸ The practice is also alluded to at Proc. Arc. 15.15, Cor. Laud. Just. 1.157f., 3.235f., Const. Porph. de Caer. 1. 84 (p. 387.8, 13f., Bonn), 1. 89 (p. 406.12; cf. p. 407.17f.), 2. 51 (p. 700.12, 16, al.), etc. (These chapters of the de Caer. all date from the sixth century; the first is certainly, the second probably and the third possibly from the περ⋯ πολιτικ⋯ς καταστάσεως of Peter the Patrician; see Bury, J. B., ‘The Ceremonial Book of Constantine Porphyrogennetos’, EHR 22, 1907, 212f., 216f.Google Scholar)

²⁹ P. Sil. Descr. 244f. ⋯ δ' ἵλαον ἠθάδι Ῥώμῃ | δεξιτερ⋯ν ⋯ρεξεν ὑποκλάζουσαν ⋯γείρων; Claud. Gild. 132f. genitor iam corde remitti | coeperit et sacrum dextra sedare tumultum.

³⁰ By the sixth century, γαλήνη and cognate terms are regularly used in connection with the Emperor: an early example is Themistius, , Or. 19Google Scholar (1.330.18 Downey); for other instances, see Lampe, G. W. H., A Patristic Greek Lexicon (Oxford, 1968)Google Scholar, s.vv. γαληναῖος 2, γαλήνη 4, γαληνότης 2; Sophocles, E. A., Greek Lexicon of the Roman and Byzantine Periods (Camb. Mass., London and Oxford, 1914)Google Scholar, s.vv. γαλήνη, γαληνός, γαληνότης 2; du Cange, C., Glossarium ad scriptores mediae et infimae Graecitatis (Lugdunum, 1688)Google Scholar, s.v. γαληνότης; McCail, R. C. in JHS 98 (1978), 51CrossRefGoogle Scholar.

³¹ There are other similarities between this passage of Themistius and Paul; see below.

³² Paul may or may not have been familiar with Sidonius' imitations of Claudian: Sidonius has no scene between Roma and the Emperor and there is nothing in Paul which could derive from Sidonius alone, except the possible parallel for his abrupt opening in Pan. Mai. (see n. 17 above), which may simply be a sign that both writers are lifting material from elsewhere.

³³ E.g. Descr. 246 ἦκα δ⋯ μειδήσας, ἵνα μυρίον ἄλγος ⋯λάσσῃ: cf. A. R. 2.61 ἦκα δ⋯ μειδήσας, simil. 3.107, Hes. Th. 547, all in eadem sede; Il. 1.2 μυρί' …ἄλγε', 18.88 πένθος …μυρίον, A. R. 1.259 ἄλγεα μυρία θείη, A.P. 11.401.3 (Luc.), Opp. H. 2.503f.; Nonn. D. 32.111 λύσσαν ⋯λάσσαι (at line–end); etc.

³⁴ On Themistius' treatment of the relationship between Constantinople and Emperor, see further Fenster, E., Laudes Constantinopolitanae (Misc. Byz. Monacensia 9, Munich, 1968), pp. 28–35Google Scholar.

³⁵ Ed. Heitsch, E., Die griechischen Dichterfragmente der römischen Kaiserzeit (2nd ed., Göttingen, 1963), i. 97Google Scholar.

³⁶ Ed. Mai, A., Scriptorum veterum nova collectio (Rome, 1827), ii. 598Google Scholar. Aspects of the work are discussed by Fotiou, A. S. in JÖB 27 (1978), 1–10Google Scholar (including reference to this passage, p. 3) and in Byzantion 51 (1981), 533–47Google Scholar, and by Behr, C. in AJP 95 (1974), 141–9Google Scholar.

³⁷ See J. M. C. Toynbee, locc. citt. in note 14 above.

³⁸ See Cameron, Alan, Claudian, pp. 273–6 and 363–6Google Scholar. Cf. Friedländer's note on P. Sil. Descr. 145 (p. 271).

³⁹ Ed. Heitsch, op. cit. in note 35 above, pp. 120–4. The parallel is noted by Cameron, Alan, Claudian, p. 255Google Scholar.

⁴⁰ It is more likely that the Egyptian panegyrist knew the works of his fellow-countryman Claudian; cf. Cameron, Alan, ‘Wandering Poets: a Literary Movement in Byzantine Egypt’, Historia 14 (1965) at pp. 494–7Google Scholar, on knowledge of Latin among Egyptian poets. It is noteworthy that Keydell, 's supplement of the Encomium 4.8Google Scholar⋯πορ[ρίψαντες ⋯νί]ην gives the same line-end as P. Sil. Descr. 169. Beazley, however, supplemented ⋯π'οἴ[κων τρέψας ⋯νί]ην. Even if Keydell is right, Paul did not necessarily derive his verse-end from the papyrus Encomium; cf. A.R. 2.884 ⋯πορρίψαντες ⋯νίας in eadem sede.

⁴¹ Averil, and Cameron, Alan, ‘The Cycle of Agathias’, JHS 86 (1966), at p. 14Google Scholar, suggest a Constantinopolitan provenance and a date of c. a.d. 555 for this epigram.

⁴² Pertusi (ad loc.) compares Claud. Still. 3.1ff., but see note 2 above on George of Pisidia and Claudian.

⁴³ Suidas s.v. Χριστόδωρος; cf. Cameron, ‘Wandering Poets’, p. 489.

⁴⁴ Justinian's public image was probably in need of support by the time Paul wrote in a.d. 562–3, when the great military achievements of the earlier years of the reign were beginning to seem remote, and administrative, economic and financial problems loomed; cf. Proc. Arc. 24, Cor. Laud. Just. 2.260f. (with Cameron's note), Agath. Hist. 5.13.5–14.4. (It should, however, be observed that all three sources are prejudiced against Justinian.) See also Jones, , Later Roman Empire, pp. 298–302Google Scholar, Stein, E., Histoire du Bas-Empire (Paris, 1949, repr. 1968), ii. 777–80Google Scholar, Lamma, P., Ricerche sulla storia e la cultura del VI^o secolo (Brescia, 1950), pp. 47–52Google Scholar and passim.

⁴⁵ See MacCormack, S., ‘Roma, Constantinopolis, the Emperor, and his Genius’, CQ n.s. 25 (1975), 131–50, esp. 139ffCrossRefGoogle Scholar.

⁴⁶ Cf. Cameron, , Claudian, p. 364Google Scholar ‘She [Roma] appears so frequently in Claudian, …because she was dramatically the most appropriate figure to exhort Stilico to save the East from Eutropius or Africa from Gildo or (above all) to officiate in a consular panegyric’.

⁴⁷ Cf. Friedländer, 's note on Descr. 145 (p. 271)Google Scholar.

⁴⁸ The notion of Old Roma's rejuvenation which is found in Latin poets (e.g. Claud. Gild. 208–12) is replaced in eastern writers by the idea that the New Roma is the daughter of the Old; cf. note 15 above, and see Dölger, F., ‘Rom in der Gedankenwelt der Byzantiner’, repr. in Byzanz und die europäische Staatenwelt (Darmstadt, 1964), at pp. 93–8Google Scholar, on the ideology of renovatio, in which the young and vital New Roma is contrasted with the ageing Roma of the West; Paul provides an early example.

⁴⁹ This classicising ideal as practised by the historians Procopius and Agathias has been expounded by Cameron, Averil, ‘The “scepticism” of Procopius’, Historia 15 (1966), 466–82Google Scholar, ‘Procopius and the Church of St. Sophia’, Harvard Theological Review 58 (1965), 161–3Google Scholar, and Agathias (Oxford, 1970), pp. 75–88Google Scholar.

⁵⁰ Paul does permit himself to name Christ directly (e.g. Descr. 193), for which he had precedent in Nonnus, Paraphrasis S. Evangelii Ioannei (e.g. 1.68 and passim) and ps. Apollinaris of Laodicea, Metaphrasis Psa1morum (e.g. proth. 65).

⁵¹ Old Roma already champions Christianity in Prudentius, Contra Symm. 2.634ff., and is portrayed in the fifth-century mosaics on the triumphal arch of Santa Maria Maggiore in Rome. Cf. Cameron, , Claudian, pp. 365f.Google Scholar, Toynbee, in JRS 37 (1947), 135fGoogle Scholar.

⁵² Paul likewise attributes the collapse of the dome of S. Sophia not to the Devil, but to the Telchines (Descr. 195), Baskania and Megaera (221), whose affiliations, like Roma's, are ambiguous. At the end of the poem, when he turns towards the explicitly Christian subject of the Patriarch Eutychius, Paul is less fastidious, mentioning Christ as Justinian's guide in similar terms to Roma (Descr. 960, cf. 226) and referring more explicitly to the Devil (975).

⁵³ Alan Cameron, ‘Wandering Poets’, p. 503, and Claudian, p. 244.

⁵⁴ Argued for by Lacombrade, C., ‘Notes sur deux panégyriques’, Pallas 4 (1956), 15–26CrossRefGoogle Scholar; accepted by Cameron, ‘Wandering Poets’, p. 503, but argued against in Claudian, pp. 321–3.

⁵⁵ Birt, Th., Claudii Claudiani carmina (Mon. Germ. Hist., Auct. Ant. X, Berlin, 1892), pp, iii, lxxviiif.Google Scholar, esp. lxxix Quodsi Sedulius in Achaia Claudianum legebat, etiam Constantinopoli saeculo VI apographa eius fuisse non mirum; quibus usum esse credo Priscianum (…) usus esse potest etiam Laurentius Lydus; Cameron, Alan, Claudian, pp. 243–5, 419, and p. 3Google Scholar on John Lydus' reference to Claudian as ⋯ Παɸλαγών.

⁵⁶ S. Antès (op. cit. in note 2 above, pp. lxxivf.) puts Claudian third, after Virgil and Ovid, among Corippus, ' sources for linguistic imitation in the Laud. Just. (collected pp. 151f.)Google Scholar; Averil Cameron (op. cit. in note 1 above, pp. 7f.) emphasises their thematic affiliation; cf., for example, her note on Laud. Just. 1.288–9, on the similarity to Claudian of Corippus' picture of Roma, and ‘Corippus' Iohannis: Epic of Byzantine Africa’, ARCA 11, Papers of the Liverpool Latin Seminar Vol. IV, 1983, ed. Cairns, F. (Liverpool, 1984), p. 172 with n. 45Google Scholar, for parallels in Claudian (and elsewhere) for Corippus' adventus scene. For loci similes between Claudian and both poems of Corippus, see Amann, R., de Corippo priorum poetarum Latinorum imitatore (Diss., Oldenburg, 1885), pp. 33–7Google Scholar.

⁵⁷ See note 10 above for Agathias' reference to Paul's παιδεία. On the school of Egyptian poets in the fourth and fifth centuries, with whom Claudian is to be associated, see Alan Cameron, ‘Wandering Poets’: mythological poets like Nonnus are exceptions to the general preference of the school for contemporary subjects (pp. 470f.). Paul's debt to Nonnus needs no illustration; for a clear echo of Triph. 41 Ἴλιος ⋯κλινέεσσιν ⋯πεμβεβαυῖα θεμέθλοις, cf. Descr. 186 ἤδη μ⋯ν σθεναροῖσιν ⋯πεμβεβαυῖα θεμέθλοις; with Colluthus' repetition of εἴξατε at 171 εἴξατέ μοι πολέμοιο, συνήθεος εἴξατε νίκης, cf. Descr. 152 εἴξατέ μοι, Ῥώμης Καπετωλίδες, εἴξατε, ɸ⋯μαι; and with Musaeus 33 σαοɸροσύνῃ τε κα⋯ αἰδοῖ (ν. Ι. δ⋯), cf. Descr. 995 σαοɸροσύνη τε κα⋯ αἰδὼς in eadem sede (striking examples; Merian-Genast, J., de Paulo Silentiario byzantino, Diss., Leipzig, 1889, pp. 101–4Google Scholar, collects others).

⁵⁸ See note 4 above.

* I am grateful to Dr Michael Whitby for reading and commenting on this paper.