‘At,’ the reading of some of the early editions, is supported by Serv. on 9. 656; but Wagn. justly observes that, coupled with ‘contra,’ it would create too strong an opposition.
 Audita is commonly rendered ‘heard of;’ in which sense “auditus” is frequently coupled with “visus,” even in the case of persons. Here however there would be no particular force in it, and it seems better, on the whole, to follow the suggestion of Serv., and suppose the reference to be to ‘clamore.’
 ‘O,’ as Wund. remarks, should have been followed by a vocative of the name of the goddess; for he is sure she is a goddess (‘O Dea certe’), though he knows not what goddess. Wund. comp. Demosth. de Cor. p. 232, Εἶτ᾽ ὦ—τί ἂν εἰπὼν δέ τις ὀρθῶς προσείποι;—ἔστιν ὅπου κ.τ.λ. To which may be added Aristoph., Clouds, 1378, ὦ τί σ᾽ εἴπω; Weidner refers to a passage in Ad Herenn. 4. 4, “tu istud ausus es dicere, homo omnium mortalium —quonam te digno moribus tuis appellem nomine?” which is given as an example of oratorical “dubitatio.” There is probably some sense of solemnity in ‘memorem.’ ‘Virgo’ is not to be pointed as a separate interrogative sentence (‘what shall I call thee? a virgin?’), as some have supposed, the word being applicable to a goddess as well as to a mortal maiden. ‘Haud—nec’ 7. 203 note.
 Hominem sonat: “humanum sonat” would be the common idiom. Persius however (3. 21) has “sonat vitium.” “Sapimus patruos” (Pers. 1. 11) is a similar expression. There is a slight similarity to this passage in Od. 6. 149 foll., and a somewhat stronger one in Apoll. R. 4. 1411 foll.
 Heyne appears to be right in dividing this line into two separate questions. Hand's notion (Tursell. 1. 315) that it is a case similar to those in which ‘certe’ follows “nescio an,” ‘whether or not—at all events,’ seems far-fetched. Looking to ‘una,’ it seems better to take ‘sanguinis’ as equivalent to “generis” (“sanguis meus,” 6. 835), not as an attributive genitive. Comp. however 6. 778, “Assaraci quam sanguinis Ilia mater Educet.” Perhaps it may be regarded here as a confusion of two modes of expression.
 Sis felix, ‘be propitious.’ Comp. E. 5. 65, “Sis bonus o felixque tuis.” Wund., following a hint of Heyne, thinks it may stand for χαῖρε, which is so common in Greek hymns; but the passage just cited is against this. ‘Quaecumque (es),’ a sort of vocative clause: comp. 8. 122, “Egredere o quicumque es.” For the thought comp. Od. 16. 183.
 Vastis et fluctibus is the reading of Pal. and other MSS. Rom. and Med. a m. pr. read ‘et vastis fluctibus,’ which is approved by Pierius, and restored by Heinsius and Heyne. It is undoubtedly true, as Wagn. says, that the former rhythm is that which we most frequently find in Virg.'s hexameters. The other however is by no means uncommon. It is therefore a question of ear in the particular passage, and the fuller close which, as Pierius says, is produced by ‘et vastis’ seems appropriate here.
 Comp. Od. 16. 181 foll.
[335-371] ‘Venus informs him that he is in the territory of Carthage, and tells the story of Dido's flight from Tyre to Africa.’