¹ I wish to record my thanks to Philip Hardie for much advice and encouragement.

² Sellar, W. Y., The Roman Poets of the Augustan Age: Virgil (Oxford, 1877), p. 199Google Scholar.

³ Hardie, P. R., Virgil's Aeneid: Cosmos and Imperium (Oxford, 1986), pp. 163–4Google Scholar.

⁴ Geo. 3.258: ‘versat in ossibus ignem’. Aen. 1.660: ‘ossibus implicet ignem’; Geo. 3.259 and Aen. 6.442: ‘durus amor’; Geo. 3.263: ‘nec moritura super crudeli funere virgo’. Aen. 4.308: ‘nec moritura tenet crudeli funere Dido’.

⁵ See Cairns, F., Virgil's Augustan Epic (Cambridge, 1989), p. 138CrossRefGoogle Scholar.

⁶ Brown, R., Lucretius on Love and Sex (New York, 1987), p. 142Google Scholar.

⁷ See, e.g., n. 15 below.

⁸ Pease, ad loc. See also McKeown, on Ovid, , Amores 1.2.1–4Google Scholar.

⁹ Op. cit., p. 204.

¹⁰ e.g. DRN 6.73. At 5.983 cura and quies are set in opposition. Cf. Aen. 4.379.

¹¹ Williams, R. D., Aeneid 1–6, ad loc. compares DRN 2.344ffGoogle Scholar.

¹² 4.531.

¹³ Noted by Ferguson, J.PVS 10 (1970), 57–63, p. 61Google Scholar.

¹⁴ 1061–2.

¹⁵ Quoted by Brown, op. cit., p. 73 n.50 in a discussion of lines 1061–2 but with no further comment.

¹⁶ Brown's translation.

¹⁷ Page, , Aeneid 1–6 (Macmillan, 1894)Google Scholar, ad loc. Austin's note is unconvincing. Surely out distinguishes between the time Dido is alone and the current time when she is in the company of Ascanius.

¹⁸ Op. cit., p. 72.

¹⁹ Is there word play on vultus/vulnus in 4.4, ‘infixi pectore vultus’ (vulnus appears two lines earlier) and 4.689, ‘infixum stridit sub pectore vulnus'?

²⁰ ‘alitur vitium vivitque tegendo’. The dependence on Lucretius is noted by R. F. Thomas, ad loc.

²¹ See Hardie, , op. cit., pp. 158ffGoogle Scholar.

²² Brown, , op. cit., p. 15Google Scholar.

²³ Brown, , op. cit., p. 54Google Scholar.

²⁴ Kenney, E. J., ‘Tityos and the Lover’, PCPHS (1970), 44–7Google Scholar. See further discussion below.

²⁵ Brown, , op. cit., p. 15Google Scholar.

²⁶ On Tibullus 1.3.75–6.

²⁷ Hardie, , op. cit., p. 178Google Scholar.

²⁸ Aen. 4.66.

²⁹ Boerma, R. E. H. Westendorp, ‘Vergil's Debt to Catullus’, Acta Classica 1 (1958), 51–63Google Scholar citing Theoc. 30.21.

³⁰ Boerma, , op. cit., p. 59Google Scholar, citing Cat. 66.23. Cf. also Cat. 35.15: ‘ignes edunt medullam’.

³¹ E. J. Kenney, op. cit.

³² Gale, M., ‘Myth in the De Rerum Natura of Lucretius’ (unpublished Ph.D. thesis, Cambridge, 1992), pp. 113ffGoogle Scholar. The example of the acquisition of fire by mortals, explained ‘scientifically’ but in language appropriate to the Prometheus myth at DRN 5.1091–1104, is discussed at p. 128.

³³ Kenney, E. J., ’Doctus Lucretius’, Mnemosyne 23 (1970), 366–92CrossRefGoogle Scholar. ‘More subtle is the device of borrowing characteristic imagery in order to turn it back on its originators and their too receptive readers.’ Gale, , op. cit. p. 138Google Scholar. (Kenney (op. cit. n. 24 above) points to a possible earlier Greek model at A.P. 12.160.)

³⁴ Servius on Verg. Eel. 6.42, ‘vulturem Hercules interemit, Prometheum tamen liberare, ne offenderet patrem, timuit.’

³⁵ Loc. cit.

³⁶ Op. cit., note on 1.3.76.

³⁷ Ferguson, J., ‘Catullus and Virgil’, PVS 11 (1971/1972), 25–42, p. 29Google Scholar.

³⁸ Lyne, R. O. A. M., Further Voices in Vergil's Aeneid (Oxford, 1987), ch. 3 (particularly p. 103)Google Scholar.

³⁹ Hesiod, , Works and Days 50ffGoogle Scholar.

⁴⁰ Pease, ad loc.

⁴¹ Hardie, , op. cit., p. 271Google Scholar.

⁴² In particular one might compare the words εἰ γ⋯ρ μ' ὑπ⋯ γ⋯ν ν⋯ρθεν θ' Αἴδου…εἰς ⋯π⋯ραντον Τ⋯ρταρον ἥκεν with ‘adigat me…ad umbras/pallentis umbras Erebo noctemque profundam'. Cf. PV 1050ff.

⁴³ An interesting passage which has some similarities to the Dido episode is Cicero, , Tusc. Disp. 2.23–5Google Scholar, where he translates lines from Aeschylus' Prometheus Lyomenos. There are the following verbal parallels: Cic. line 18: custodem…alo; Aen 4.2: vulnus alit; Cic. 21: a pectore; Aen. 4.2, 689: sub pectore; Cic. 22: pestes excipio; Aen. 4.90: peste teneri. Figo also occurs in both contexts. (Berg, W., ‘Daphnis and Prometheus’, TAPA (1965), 11–25Google Scholar, argues that the Prometheus trilogy influenced Vergil in the composition of Eclogue 5.)

⁴⁴ OLD s.v. 2b.

⁴⁵ It should be stressed at this point that it is the aspect of suffering which is prominent in the reference to Prometheus in Catullus' poem. The point is made by Bramble, J. C., PCPHS (1970), p. 32Google Scholar. ‘Catullus did not have to spend three lines dwelling on his [Prometheus'] punishment’. Later he asks why ‘Catullus devote[d] this amount of space to describing the agonies of Prometheus’.

⁴⁶ Aen. 6.432.

⁴⁷ Conte, G. B., The Rhetoric of Imitation (Cornell, 1986), p. 37Google Scholar.

⁴⁸ See, e.g., Camps, W., An Introduction to Virgil's Aeneid (Oxford, 1969), ch. ixGoogle Scholar.

⁴⁹ Lyne, , op. cit., p. 103Google Scholar.

⁵⁰ Lyne, , op. cit., p. 139Google Scholar. Gale, M., op. cit., p. 113Google Scholar, makes the attractive suggestion that Vergil learned this technique from Lucretius.

⁵¹ Aeschylus, , PV 85–7Google Scholar. The statement o f Kratos is ironic but with a n element o f truth.

⁵² Poschl, V., The Art of Vergil (trans. Seligson, G., 1970), p. 78Google Scholar (and p. 194 n. 34) argues that vatum is an objective genitive.

⁵³ A theme also in the Georgics. Thomas, at Geo. 4.488Google Scholar on Orpheus (incautus amans) remarks that ‘Orpheus' failure is emotional, a loss of control caused by amor’. Cf. 2.303 and Thomas ad loc.

⁵⁴ Verg, . Eel. 6.42Google Scholar.

⁵⁵ Cairns, F., Virgil's Augustan Epic (Cambridge, 1989), p. 143CrossRefGoogle Scholar.

⁵⁶ Aeschylus, , PV 447ff.Google Scholar

⁵⁷ Aen. 1.308.

⁵⁸ Significantly, Aeneas is engaged in construction when Mercury arrives with Jupiter's message. Aen. 4.260.