“Acer Atinas” 12. 661, which shows that ‘acer’ here does not go with ‘fugit.’ Atinas is only mentioned in these two places, though Dryden wished to substitute him for Asilas in 12. 550. But for the difference in quantity, it would be natural to connect him with the town Atina, mentioned 7. 630.
 Nearly repeated from 8. 596. Rom. and originally fragm. Vat. have ‘quadripedo,’ which may possibly be right, though Forc. cites no higher authority for “quadrupedus” than Ammianus (14. 2).
 Comp. 4. 586, 589, the first of which will illustrate ‘e speculis,’ the second ‘percussae pectora.’ Virg. not unfrequently imitates himself in this way: e. g. vv. 871, 872 above supply a hint to 12. 463, 464, while v. 872 was partly suggested by 10. 797, the next line, as has just been remarked, similarly suggesting v. 873. Taken separately, each of these resemblances would be a mere coincidence; but when two come together, we feel that there must be something conscious in the self-reproduction. Med., fragm. Vat., and one of Ribbeck's cursives originally omitted ‘e,’ perhaps, as Ribbeck thinks, from a notion about the quantity of ‘pulvis,’ as Serv. remarks that the last syll., though short here, is long 1. 478. See Excursus to Book 12. Some copies, including another of Ribbeck's cursives, give ‘e muris.’
 Cursu, at full speed, 2. 321. Heyne comp. with this description Il. 13. 122 foll.; but the resemblance is not very marked. We might also comp. generally the latter part of Il. 21. See too above, 9. 722 foll.
 The enemies, following on their heels, blend with them (‘mixto agmine’), and fall upon them (‘super’), killing them on the threshold of the gates, and even after they have got into their dwellings. Forb. erroneously takes ‘inimica turba’ of the Latins, who tumble on and destroy each other, understanding ‘confixi’ of their being pierced by the weapons of their friends, and making ‘super’ mean ‘beside the danger from the enemy.’
 One of Ribbeck's cursives has ‘fugiunt.’
 Intra Med., Rom., which may be right. Wagn.'s argument for ‘inter,’ that Virg. meant not that they had got inside their houses, but that they were within the gate and in the neighbourhood of their houses, is futile, as Virg. evidently intends the former, not the latter. There is a gradual climax: they are killed on the threshold of the gate, killed within the walls, killed even within the houses. ‘Inter’ however would stand equally well for either. ‘Tuta domorum:’ see on 1. 422, and Madv. there referred to, Munro on Lucr. 1.315. Here it is difficult to say whether the notion of quality, ‘the security of home,’ or the partitive notion, i. q. “domorum penetralia,” is the prominent one.
 “Oriturque miserrima caedes” 2. 411, where, as here, it is the slaughter of countrymen by countrymen. With the structure of these lines comp. 12. 410, “It tristis ad aethera clamor Bellantum iuvenum et duro sub Marte cadentum.”
 The commentators have made considerable difficulty about this line, but the sense is perfectly plain. The Latins already within the walls close the gates, and finding that their pursued countrymen attempt to force their way in, repel them as they would the enemy: the pursued rush on, and fight as they would if they were an invading force; and so there is carnage on both sides.
 “Ante oculos et ora parentum” 2. 531. The shutting of the gates has been mentioned (v. 883) as the act of a part of those within, who would naturally be divided in their counsels; and the relatives of those shut out would lament the act, even if they did not resist it. For ‘exclusi’ divided into ‘pars—pars’ comp. 12. 277 foll.
 As the rout hurries on, some miss the causeway that leads to the gate, and are carried down the bank of the moat on each side of it. ‘Ruina’ is the rout pouring headlong, not, as Heyne thinks, the breaking down of the sides of the moat.
 For ‘arietat’ see Forc. One instance of it is quoted from Att. Brutus fr. 1 in its natural sense, the butting of a ram; the rest are more or less metaphorical. In a passage like the present it may be doubted whether the metaphor is from the animal or the battering-ram. Seneca appears to be rather fond of the word. ‘Duros obiice postis,’ i. q. “dura obiice postis,” as “pictas abiete puppis” 5. 663 i. q. “picta abiete puppis,” a construction which is especially common in Pers. It is even possible that the words in Book 5 may have suggested those here, ‘inmissis frenis’ having recalled to the poet's mind his own “inmissis habenis,” cited just above. See on v. 877. ‘Postis’ the door, 2. 480.
 Ipsae matres 5. 767. ‘De muris’ with ‘iaciunt.’ ‘Summo certamine’ as in 5. 197, not, as Serv. explains it, “in extremo discrimine,” in spite of v. 476, which he comp. This whole description is not quite harmonized with the preceding: both illustrate the crisis of the Latin fortunes, but one gives a melancholy picture, warriors wasting their strength on each other in the instinct of self-preservation, the other a more cheerful and hopeful one, women roused to deeds of manly daring. It may be said to be one of the few instances in which traces of imperfection are found in these later books.
 Monstrat, i. e. “monstrat iacere:” so “conferre manum pudor iraque monstrat” 9. 44, comp. by Wagn., who rightly rejects Heyne's two alternatives “monstrat Latinas matronas, earum animan et virtutem declarat” and “monstrat Camillam, insignem reddit.” ‘Ut videre Camillam’ is no part of the parenthesis, as Gossrau, Forb., and Ribbeck make it, but goes with ‘iaciunt:’ the sight of Camilla's valour has such an effect on them that they try to imitate her in their way. Serv. well remarks “‘ut videre Camillam:’ scilicet quae pro aliena patria cecidisse videtur. Sane ut videre, non relatam ex pugna, sed ut exemplum virtutis eius viderunt.” heyne, who discusses this latter question, remarks that nothing is said about the carrying away of her body, which had doubtless been done by Diana, v. 593. For ‘verus’ Rom. has ‘versus,’ which seems a mere blunder, though Pierius tries to explain it. It is just conceivable that it may have arisen from a misapprehension of Serv.'s note “‘Monstrat amor verus:’ qui apparet in adversis.”
 Comp. 7. 505 foll., 524 foll.
 Praecipites like ‘trepidae.’ ‘Pro moenibus,’ not for “in moenibus,” as Forb., but like “pro patria mori.” ‘Ardent,’ restored by Heins., is found in Med. (first reading), Pal., fragm. Vat., Rom. and Gud. originally. ‘Audent’ is the second reading of Med., and is found in Gud. corrected and two other of Ribbeck's cursives. Either would be sufficiently good, but external authority and the occurrence of ‘audent’ at the end of v. 884 are reasons for preferring ‘ardent.’ ‘Primae’ may possibly mean in the front rank: comp. 10. 125 and its context. It matters little whether we take it with ‘ardent’ or (by attraction) with ‘mori.’
[896-915] ‘Acca takes the news to Turnus, who breaks up his ambush and hastens to the city. Immediately afterwards Aeneas comes up, passes the defile safely, and marches towards the city himself. Night however prevents an engagement.’