1. For bibliography see Briggs, W. W., Aufstieg und Niedergang der römischen Welt ii31.2 (1981), pp. 1268ff., esp. 1327–31, 1334–7Google Scholar; also N. B. Crowther, ibid., ii.30.3 (1983), pp. 1622ff., esp. 1633ff.

2. Callimachus, fr. 1.21–24 Pfeiffer; cf. Clausen, W., GRBS 5 (1964), 181–96Google Scholar = Approaches to Catullus, ed. Quinn, K. (Cambridge, 1972), pp. 269–84Google Scholar; Commager, S., The Odes of Horace: a Critical Study (Bloomington and London, 1962), ch.1Google Scholar.

3. The passage draws especially on Lucr. 5.416–508; for the birth of living creatures, see further 5.783–836.

4. For other epyllia (Calvus' Io, Valerius Cato's Dictynna, etc.) see Lyne, R. O. A. M., CQ 28 (1978), 167–187, esp.170, 173–4CrossRefGoogle Scholar; see further Gutzwiller, K., Studies in the Hellenistic Epyllion (Konigstein, 1981)Google Scholar.

5. E. g. lines 41–42, 61 (Atalanta not named); 62–63 (Phaethon's death not explained); 76 (‘Dulichias’ used to refer to Ulysses); see also n. 16. Cf. Catullus 64.228, 290, 324, 346. In all these passages the reader needs to know the story already in order to get the point.

6. I refer to the myth of Pasiphae and the bull. Cf. the incestuous passion of Myrrha/Zmyrna. Scylla's nurse in the Ciris suspects her charge of the same vicious longings (237–40).

7. Cf. Catullus 64.132–201; Lyne, on Ciris 163–90, 404–58Google Scholar.

8. Cf. Moschus, , Europa 37–62Google Scholar; Callimachus, Hecale fr. 260 (= Lloyd-Jones, and Parsons, , Supplementum Hellenisticum 288), 16ff.Google Scholar; Catullus 64. 50–266; Ciris 294–309.

9. Cf. Callimachus, fr. 612: Nisbet-Hubbard, on Hor. Odes 1.7.23Google Scholar.

10. Nerea, Hylan, Pasiphaen, Phaethontiadas; for metre, see the mannered hiatus of line 44.

11. Lines 47, 52, 77; also the sympathy implicit in line 45, and the use of ‘empathetic’ adjectives: 77, ‘timidos’; 81, ‘infelix’. Cf. Catullus 64. 60–62 (esp. ‘eheu’), 69–71 (esp. the vocatives and ‘a misera’), etc.; Virg., Geo. 4.465–6, 491, 504–5, 525–7Google Scholar.

12. See now Knox, P. E., Ovid's Metamorphoses and the Tradition of Augustan Poetry (PCPhS Suppl. 11, 1986)Google Scholar, though the suggestion that the sixth Eclogue itself (as opposed to earlier catalogue poetry and Hellenistic work on metamorphoses) provided the primary inspiration for Ovid's poem does not convince.

13. Cf. Hor., Odes 2.9.18–24Google Scholar. One passage which might be taken to indicate non-elegiac composition by Gallus is Ecl. 10.50f., ‘ibo et Chalcidico quae sunt mihi condita versu / carmina’ (cf. R. Coleman's commentary ad loc). But more than one view is possible here: imitation of Euphorion might have been included in Gallus' Amores, as Callimachus is imitated by Propertius; love-poetry and learned poetry are not incompatible, as Catullus 68b shows. If ‘amores’ in lines 6, 53, and 54 of Eclogue 10 puns on the title of Gallus' elegies, then the likelihood that these were the location of his Euphorionic imitation is increased.

Or Gallus might have embarked on more ambitious and learned poetry but abandoned it, returning to love-elegy (cf. Catullus 35 on Cornificius). See further Crowther (n.l), pp. 1631 ff., and in CQ 20 (1970), 325fGoogle Scholar.

14. On such ‘framing’ effects see also Hutchinson, G. O., Hellenistic Poetry (Oxford, 1988), pp. 284–6Google Scholar; he, however, attaches less importance to the literary background (and to Callimachean poetics in particular) than I should.

15. Cf. Hubbard's, M. rewarding paper in PCPhS n.s. 21 (1975), 53–62Google Scholar.

16. See Serv. auct. ad loc: ‘hunc fluvium Hyacinthi causa Apollo dicitur amasse’: Eur., Helen 1469–75Google Scholar; Ov., Met. 10. 162–219Google Scholar. For such ‘implicit myths’ see Lyne, , Further Voices in Virgil's Aeneid (Oxford, 1987), pp. 139–40Google Scholar. The concept is obviously a dangerous one, but some examples, including this one, seem certain, especially in self-consciously and ostentatiously learned poetry like this.

17. Cf. the suggestive remarks of Ross, D. O. Jr, Backgrounds to Augustan Poetry (Cambridge, 1975), p. 111Google Scholar (though there is much else in this context that I would not accept). See also Griffin, J., G&R 26 (1979), 70–73Google Scholar = Latin Poets and Roman Life (London, 1986), pp. 175–8Google Scholar.

18. Catullus on the ‘senes severiores’: 5.2. Attacks Memmius, 28.9ff.; Piso, 28 and 47; Caesar, 29 and 93 (cf. Suet., D. lul. 73Google Scholar; Tac., Ann. 4.34)Google Scholar; Mamurra, 21 with Fordyce's note, 94, 114–15, etc. Criticizes bombastic epic: 36, 95, 95b; and bad writing in general: 14, 22. On his own poetry and those of his circle: 1, 35, 50, 68a. 33–36, 95. Cf. Ticidas fr. 2 Morel; Cinna fr.14. See further Wiseman, T. P., Clio's Cosmetics (Leicester, 1979), chh. 9–11Google Scholar.

19. Wilkinson, L. P., The Georgia of Virgil (Cambridge, 1969), pp. 20–24Google Scholar, citing esp. Serv., Ecl. 6.13Google Scholar, Probus, , Vita 10–12Google Scholar, and a fragmentary Herculaneum papyrus, P. Herc. 1082; see esp. Gigante, M., Cron. Erc. 3 (1973), 86–87Google Scholar. The evidence is hardly very substantial (and the Catalepton poems should not be used), but there is a plausible enough case.

20. The relationship between Lucretius and Catullus is still debated: on chronological grounds it is more plausible that Catullus imitates Lucretius, but his imitations are largely confined to poem 64: see C. Bailey's commentary on Lucretius iii, pp. 1753–4. Lucretius himself uses Callimachean (and neoteric?) language (see Kenney, E. J., Mnemos. 23 (1970), 366–92CrossRefGoogle Scholar; Brown, R. D., ICS 7 (1982), 77–97) but his attitude to erotic poetry such as the novi poetae composed is antagonisticGoogle Scholar: see Brown, R. D., Lucretius on Love and Sex (Leiden, 1988), pp. 132–5Google Scholar.

21. For a contrary view see Spoerri, W., MH 27 (1970), 144–63Google Scholar.

22. For epyllion-style poetry in the Georgics, see 4.315–553 (the story of Aristaeus and Orpheus); for aitia in the later works of Virgil, see e.g. Geo. 2.380–96, 3.113–17, 4.15, 149–53; Aen. 5.59–60, 596–603, 6.69–70, etc.; for pastoral elements in the Aeneid, see Anderson, W. S., TAPA 99 (1968), 1–17Google Scholar; for the pervasive influence of Lucretius, see now Hardie, P., Virgil's Aeneid: Cosmos and Imperium (Oxford, 1986), ch. 5Google Scholar.

23. Like ‘surgamus’ in 75 (Jenkyns, R., JRS 75 (1985), 72)Google Scholar, the emphatic references to ‘umbra’ should be read on a metaphorical or symbolic level. Compare such uses as in Cic., Tusc. 2.27Google Scholar, ‘vita umbratilis’; Quintilian 10.5.16f., and other passages cited by Mayor on Juvenal 7.105, 173.

24. So also with one's predecessors: poets adopting a model redefine that model in terms suited to their own poetic aims and contextual needs: e.g. Horace on Lucilius, Juvenal on Horace, Propertius on Catullus; above all Ovid on a long procession of alleged earlier ‘lovepoets’ (Tristia 2, esp. 361–468).

An earlier version of this paper was read to a meeting of the Bristol branch of the Classical Association in November 1987. I am also grateful to Don Fowler and Peter Parsons for their shrewd criticisms.