See on G. 2. 528.
 Anchises stands on the stern, which was the sacred part of the vessel, containing representations of its tutelary gods. So 10. 171, “aurato fulgebat Apolline puppis.” These are not to be confounded with the παράσημα, or figureheads, which were placed on the prow. For ‘celsa’ Ribbeck reads ‘prima,’ which is found as a variant in Gud., and is doubtless the first reading of Pal., where the word is ‘ima.’ In 8. 680, where the words recur, Priscian quotes ‘prima.’ There is something to be said for the change, as MSS. are fond of assimilating to each other passages already similar in part (see 1. 668), but the external evidence for it seems hardly sufficient.
 This comprehensive enumeration seems intended to include all the gods. Those who question the propriety of ‘et terrae’ may accept Serv.'s explanation, “ad quam iturus sum.” ‘Tempestatumque potentes’ G. 1. 27 note.
 Ferte in the sense of ‘date,’ perhaps with an allusion to the use of ‘ferre’ of a wafting or carrying wind. ‘Vento’ apparently as in 1. 307., 2. 25., 4. 46, an instrumental or modal ablative. Altogether the expression is a harsh one, and could hardly be justified except by a reference to Virg.'s practice of alluding to one form of words while he uses another.
 It is a question whether ‘Minervae’ belongs to ‘arce’ or to ‘templum.’ If it were established that the place was called “Arx Minervae” as well as “Castrum,” the former would be the more natural construction; otherwise probability would seem in favour of the latter.
 The action of the east wind on the water is said to have hollowed out the harbour. For ‘Euroo,’ a rare adjective, seemingly occurring only in Priscian, Periegesis, v. 871, the old reading was ‘Eoo:’ but the great majority of MSS. support the word in the text.
 ‘Adspergine’ was the reading before Heins.; but ‘adspargine,’ the older form, has the authority of the grammarian Velius Longus, as well as of Med., Pal., and Gud.
 Latet is not inconsistent with ‘patescit’ above. The harbour is retired and in fact concealed between the rocks on each side of it: but as the ships approach a way is seen between the barriers. Aeneas is giving a general account of the haven, not describing its features as they broke upon him gradually. The supposed inconsistency however may have given rise to an unmetrical reading ‘late patet,’ which Serv. attributes to Donatus. Wagn. thinks Donatus' reading was ‘patet,’ ‘late patet’ being added as an explanation. Forb. thinks ‘latet’ refers to a later point in the approach than ‘patescit,’ the rocks impeding the view as the voyagers drew nearer; but this seems less natural. ‘Bracchia’ and ‘muro’ are two metaphors to express the same thing, the rocks which form the two sides of the haven. Comp. 2. 481, “ore fenestram.” ‘Bracchia’ however has the further propriety of being used for a line of wall in fortification: see Forc. Ov. M. 11. 230 has the same metaphor, “Est sinus Haemoniae curvos falcatus in arcus: Bracchia procurrunt,” probably imitating this passage. There is a similar picture in Od. 10. 89 foll. The general features are not unlike those of the harbour in A. 1. 162 foll. as the commentators remark.
 Turriti is apparently to be understood metaphorically, crowned as with towers. ‘Refugit:’ the eminence on which the temple is placed slopes downwards, so that, as the ships approach, the building appears to recede. Germ. comp. Prop. 5. 6. 15, “Est Phoebi fugiens Athamana ad litora portus.”
 Primum omen: the first object which meets us, regarded consequently as an omen. Comp. generally 1. 442 (note), where a horse is similarly interpreted as symbolical, and for ‘primum’ 7. 118 (note).
 For Anchises' power of interpreting omens see on 2. 687. ‘Hospita:’ see on v. 377. ‘Portare’ as of a messenger. ‘Yours is a message of war.’ Comp. Ter. Heaut. 4. 1. 12, “Nescio quid peccati portat haec purgatio.”
 Olim is used generally. ‘Yet the time comes when these same beasts are trained to put their shoulders to the car.’
 Iugo seems to be an instrumental or modal abl. Horses are yoked together, and thus made jointly amenable to the bit. The concord thus produced is a symbol of peace, besides conducing to peaceful arts, such as ploughing.
 The reading before Heins. was ‘spes est pacis.’ On ‘numina sancta’ Pomponius Sabinus observes, “Ita et Probus legit et Apronianus,” a note which has been explained as alluding either to a possible variety, ‘nomina’ for ‘numina,’ or to an actual one, ‘sacra’ for ‘sancta,’ which is found in a few MSS. But the remark may perhaps refer to the early part of the line.
 ‘Armisonus’ is a rare word, perhaps only used by Claudian, Rape of Proserpine, 3. 67, where it is an epithet of a cave. Here the reference to Pallas' martial character is in keeping with the previous lines.
 See on v. 405. ‘Phrygio’ may either mean ‘embroidered’ (see on v. 483 above), so as to correspond to ‘purpureo’ v. 405, or merely designate the custom as a Trojan one, in the spirit of vv. 408, 409. A reading ‘capute’ leads Wagn. to suggest that Virg. may have written ‘caput,’ as in 5. 309 &c. Pal. and Gud. a m. pr. have ‘aram.’
 Praeceptis, a sort of instrumental ablative. So “legibus et institutis” Cic. de Sen. 11. See Madv. § 255. ‘Dederat quae maxuma,’ which he had given as the greatest, as we should say, on which he had insisted most, referring to vv. 433 foll.
 Argivae is not, as Heyne thinks, an ornamental epithet, but points out the reason why Juno is to be propitiated, as the patroness of the enemies of Troy. ‘Adolemus’ E. 8. 65 note. ‘Honores’ 1. 49.
[548-569] ‘Setting sail again, we pass by Tarentum, and come within sight of Aetna. We avoid Charybdis, but are tossed by the waves, till at last at evening we land in the Cyclops' country.’