¹ The nostos (Od. 9–12Google Scholar), but especially the song of Demodocus about the sack of Troy (Od. 8.499–520), which causes Odysseus to weep (Od. 8.521–2). For a fuller discussion of the exact influence of Homer in this passage see Knauer, G. N., Die Aeneis und Homer, Hypomnemata 7 (Göttingen, 1964), pp. 169–72.Google Scholar

² So Servius, Heyne, Forbiger, Conington, Austin on Aen. 2.7.

³ The later literary tradition, apart from stressing connections between Thessaly and Aegina, does not significantly depart from the Homeric picture of the Myrmidons (see Schmidt, J., ‘Myrmidones’, RE 31, 1108–11).Google Scholar

⁴ cf. Il. 9.484 (also Apollod. 3.175 and Find. fr. 173). In historical times the Dolopes are quite firmly established in Thessaly, but their role in myth seems to have been rather limited (see Miller, J., ‘Dolopes’, RE 9, 1289–90).Google Scholar

⁵ Aen. 2.7 (see above) and Aen. 2.785: ‘non ego Myrmidonum sedes Dolopumve superbas/aspiciam’. Myrmidons also appear at Aen. 2.252 and Aen. 11.403.

⁶ cf. ‘hie Dolopum manus, hie saevus tendebat Achilles;/classibus hie locus, hie acie certare solebant’ (Aen. 2.29–30). But the mere fact that they are mentioned in the same verse does not necessarily mean that the Dolopes were encamped with Achilles. The opposition in the second line between classibus and acie, emphasized by the repeated hic, suggests that the same anaphora in the first line may also function to contrast the different (or even opposing) points on the shore held by the Dolopes and Achilles. Cf. also Austin's comment on Aen. 2.29.

⁷ ‘undique collecti invadunt, acerrimus Aiax/et gemini Atridae Dolopumque exercitus omnis’ (Aen. 2.414–15).

⁸ After Scyria pubes (Aen. 2.477) Vergil reverts to Danai (Aen. 2.495) and Danaum (Aen. 2.505), a name which since Homer has been applied to Greeks in general, see Miller, J., ‘Danaoi’, RE 8, 2093–4.Google Scholar

⁹ Wordplays on proper names occur in Homer (see Dimock, G., ‘The Name of Odysseus’, The Hudson Review 9 [1956], 52–70CrossRefGoogle Scholar) and are fairly common in Greek tragedy, e.g. Aesch. Ag. 686–90Google Scholar (Helen), Sept. 576–9Google Scholar (Polyneices); Soph. Aj. 430–1Google Scholar (Ajax), OT 397, 924–6Google Scholar, possibly 1038 (Oedipus); Eur. Bacch. 508Google Scholar (Pentheus). Cf. also McCartney, E. S., ‘Puns and Plays on Proper Names’, CJ 14 (1918–1919), 343–58Google Scholar; Stanford, W. B., Ambiguity in Greek Literature (Oxford, 1939)Google Scholar; and the excellent comment of E. R. Dodds on Eur. Bacch. 367 about the feeling of the ancients that a person's name revealed something essential about him.

For discussions of wordplays in early Latin poetry see Skutsch, O., The Annals of Quinlus Ennius (Oxford, 1985)Google Scholar on Ann. 127, 222, 231, 290, 369, 540 SkGoogle Scholar; Friedländer, P., ‘The Pattern of Sound and Atomistic Theory in Lucretius’, AJP 62 (1941), 16–34Google Scholar; and Snyder, J. M., Puns and Poetry in Lucretius’ ‘De rerum natura’ (Amsterdam, 1980)Google Scholar.

Vergil furnishes numerous instances of etymological or aetiological wordplay, e.g. dis AGrippa secundis/Arduus AGmen AGens (Aen. 8.682–3), LATe LATio (Aen. 8.14), LATiumque… LATuisset (Aen. 8.322–3). Further examples in Marouzeau, J., Quelques aspects de la formation du latin littéraire (Paris, 1949), pp. 71–9Google Scholar; Knight, W. F. Jackson, Roman Vergil (London, 1944), pp. 197–206Google Scholar; Hanssen, J. S. Th., ‘Virgilian Notes’, SO 26 (1948), 113–25CrossRefGoogle Scholar; Kraggerud, E., ‘Einige Namen in der Aeneis’, SO 36 (1960), 30–9CrossRefGoogle Scholar; Morland, H., ‘Nisus, Euryalus und andere Namen in der Aeneis’, SO 33 (1957), 87–109CrossRefGoogle Scholar; Mørland, H., ‘Zu den Namen in der Aeneis’, SO 36 (1960), 21–9CrossRefGoogle Scholar; Morland, H., ‘Zu einigen Stellen in der Aeneis’, SO 48 (1973), 7–23.CrossRefGoogle Scholar The most recent treatment of wordplay in Latin writers is that of Ahl, F., Melaformations: Soundplay and Wordplay in Ovid and Other Classical Poets (Ithaca, 1985), esp. pp. 62–3, 281, 303–4Google Scholar on etymological wordplays with names.

¹⁰ cf. ὅν ποτ' ⋯ς ⋯κροπóλιν δ⋯λον ἤγαγε δîος Οδυσσε⋯ς (Od. 8.494).

¹¹ The differences in vowel quantity between Myrmidŏnum and dōnum would not have been an impediment to the wordplay, since Latin literature furnishes many examples of ‘wordplay across vowel quantity’ (see Ahl, , op. cit. [n. 9], pp. 56–7).Google Scholar

¹² The traditional view that ruit implies ‘upward motion’ here (see Austin, ad loc.Google Scholar and Henry, J., Aeneidea ii. 24–5Google Scholar) has been challenged by Mack, S. in CQ 30 (1980), 153–8CrossRefGoogle Scholar and most recently by Knox, P. E. in CQ 39 (1989), 265.CrossRefGoogle Scholar The image of night's fall (ruit, Aen. 2.250) anticipates the image of Troy crashing down (ruit alto a culmine Troia, Aen. 2.290) from its lofty height (see Mack, , above, p. 158Google Scholar).

¹³ Miller, ‘Dolopes’ he., col. 1289 (citing Fick, , Griechische Personennamen, p. 387Google Scholar) suggests that the name may derive from δ⋯λος.

¹⁴ Although Servius objects to this obscene sequence of sounds, many scholars doubt whether cacenphata would have been consistently avoided (cf. Forbiger, Conington, Austin ad loc., and Hanssen, J. S. Th., ‘Remarks on Euphony-Cacophony, and the Language of Virgil’. SO 22 [1942], 105–6).CrossRefGoogle Scholar Yet surely Vergil would not have been indifferent to sound effects which rhetoricians found offensive. When he did use them in Aen. 2.27, he may well have done so ‘to stigmatize the detestable invading force of the Greeks’ (see Highet, G., The Speeches in Vergil's Aeneid [Princeton, 1972], pp. 130–1).Google Scholar Of the six collocations of -CA CA- in Vergil four occur in contexts where cacenphaton might not be out of place (the others are Geo. 2.13 and Aen. 3.203). Three of these four involve Greeks (DoriCA CAstra, Aen. 2.27 = 6.88; AchaiCA CAstra, Aen. 2.462), while one occurs in Venus' speech outlining her plot to ensnare Dido (Aen. 1.673–4).

¹⁵ Servius, followed by others (e.g. Henry, , Aeneidea ii. 44Google Scholar), explains Aen. 2.31 as follows: ‘non quod ipsa dedit, sed quod ei oblatum est’. But there is obviously intentional ambiguity here, for the Horse was fashioned with Athena's help (cf. Od. 8.493, Il. 15.71), as divina Palladis arte (Aen. 2.15) clearly indicates (pace Henry, , AeneideaGoogle Scholar ii.31–8). The same ambiguity also attends Sinon's words when in Aen. 2.189 he warns of violating the DONA Minervae. Shortly thereafter the suggestion of ‘divine gifts’ finds its full expression in the description of the night (Aen. 2.268–9):

tempus erat quo prima quies mortalibus aegris

incipit et DONO DIVUM gratissima serpit.

¹⁶ Danaum insidias is normally taken to mean ‘the treacherous device of the Greeks’ (so OLD), but the phrase is ambiguous, for its etymologically more literal sense is ‘the place where the Greeks sit in ambush’, so that Capys ends up saying more than he realizes. Henry (Aeneidea ii.43–4) suggests that Danaum insidias may derive from ξεστ⋯ν λ⋯χον' Αργε⋯ων (Eur. Troad. 534).Google Scholar

¹⁷ cf. Soph. Aj. 665Google Scholar, Eur. Med. 618Google Scholar, and Otto, A., Die Sprichwörter und sprichwörtlichen Redensarten der Römer (Leipzig, 1890), pp. 120–1.Google Scholar

¹⁸ With the appearance of Sinon the theme of treachery (DANAUM INSIDIAS, Aen. 2.65 and cf. 309) becomes focused on one person who is the archetype for all (crimine ab uno disce omnes, Aen. 2.65) as a consummate practitioner of perjury and deception (INSID11S periurique ARTE Sinonis, Aen. 2.195), a craft seen as quintessentially Greek (cf. Aen. 2.106: ARTISque PELASGAE: also Aen. 2.152: ARTE PELASGA, where it is equated with DOLIS). Austin (on Aen. 2.152) notes how ‘the repetition brings out the parallelism between Sinon's craft in the first part of his story and that of this new and final chapter’. The only instance of Pelasgus attested in earlier Latin poetry is the Ennian sub Mane Pelasgo (Ann. 14Sk), which may well have inspired the Vergilian phrase. For the notion of ruin brought about by a combination of Greek guile and divine disfavour cf. Aesch. Pers. 361–2.Google Scholar

¹⁹ Iliad: Δαναο⋯, 141 times; 'Αργεîοι, 188; 'Αχαιο⋯, 602. Odyssey: Δαναο⋯, 13 times; 'Αργεîοι, 34; 'Αχαιο⋯, 124.

²⁰ Once in Lucr. 1.86 and in Var. At. 1.4 (M).

²¹ cf. Cap. 819Google Scholar, Mer. 226Google Scholar, etc., also CIL 1.1531.7, Naev. 40M, Caecil. com. 176R, Pac. trag. 207R. For other examples of wordplay involving a change of vowel see Ahl, , op. cit. (n. 9), pp. 57–9Google Scholar and passim.

²² cf. Leonidas of Tarentum (A.P. 6.305.1–2): δρα…| θ⋯κατο δε⋯ơοζοςθ⋯κατο δε⋯ơοζος Δωρι⋯ος κεΦαλ⋯, and Archimelus (in Athen. 5.209b): κ⋯ρπον π⋯ονα δωροΦορ<inline2>;ν, | Σικελ⋯ας σκαπτοûχος ⋯ Δωρικ⋯ς. Could examples like these have suggested Dorica castra (Aen. 2.27)?

²³ I am indebted to the Editors for drawing my attention to this possibility.

²⁴ Alex. 269, 710, 887, 1269, 1381.Google Scholar Of these, 269 and 1269 occur in a Trojan context, but they have no immediate relevance for Aen. 2.

²⁵ Fr. 41: Πλευρ⋯ τε κα⋯ θώρηκα δι⋯ρικεν ίν⋯ου ἄχρις.

Fr. 42: Τ⋯ ῥ⋯ οἱ δ⋯νος ⋯πασεν Εκτωρ.

These fragments have usually been assigned to the poem about Hyacinthus, in which E. rejects the version in which the flower springs from the blood of Ajax (cf. Skutsch, F., ‘Euphorion’, RE 11, 1181).Google Scholar

²⁶ cf. fr. 44, 54, 56, 59, 60, 62, 66, 72, 73 Powell.

²⁷ See Servius on Aen. 2.32, 79, 201, 341; also cf. Skutsch, F., RE 11, 1188Google Scholar; Heinze, R., Virgils epische Technik (Leipzig, 1915), p. 18Google Scholar; Austin on Aen. 2.201.