Litora Med., Rom., Gud. (corrected), ‘limina’ Pal., Gud. (originally). The same variety has met us 2. 321, though there the authority for ‘litora’ is very slight. Here it would seem to have the support of the great majority of MSS., the only other authorities distinctly quoted for ‘limina’ being one of Ribbeck's cursives, seven copies mentioned by Heins., and one by Burm., while we can hardly treat Pal. and Gud. as independent witnesses. ‘Litora’ however cannot be said to be a natural reading. The Tyrrhenian army may have been encamped near the shore (comp. v. 497), but the shore of the king is not the same as his seaside camp, and there is a disagreeable incongruity between ‘equites’ and ‘litora,’ not justified as in the case of “puppes signa ferre iubent” v. 498 by any rhetorical propriety. It seems best therefore to recall ‘limina,’ which was first displaced by Wagn. The ‘limina’ is here the door of the general's tent, but that does not make the expression less apposite, as the general notion is that of seeking the king in his abode. Comp. v. 145 above. Pal., Gud. (originally), Canon., and others have ‘Tyrrhena,’ which is tempting (comp. v. 526), but would perhaps be more plausible if ‘litora’ were read.
 “Bene ‘metu duplicant,’ nam inest semper in matribus votum” Serv. ‘Propius’ &c., ‘fear treads more closely on the heels of danger,’ probably including both the notion that as danger is nearer fear is greater (Heyne), and the conception of fear as coming nearer the danger by anticipating it (Wagn.). Cerda comp. Aristot. Rhet. 2. 5, τοῦτο γάρ ἐστι κίνδυνος, φοβεροῦ πλησιασμός.
 ‘The features of the war-god begin to loom larger.’ They realize war more as their kinsmen are departing to it. The conception is not quite the same as in 2. 369, with which it is there compared, as in the other passages quoted the mere sight of a physical object seems to be meant: here War is conceived of as a spectre which haunts the imagination.
 There are three possible readings, ‘inexpletum lacrimans’ Pal. (originally), ‘inexpletus lacrimans’ Pal. (corrected), Rom. (which has ‘inpletus,’ the original reading of Gud.), and ‘inexpletus lacrimis’ Med. All three are mentioned by Serv., who prefers the first; and this was the usual reading till Wagn., who introduced the second. ‘Inexpletus lacrimans’ is strongly supported by G. 4. 370, though there is a question there between “saxosus” and “saxosum,” and by A. 3. 70 “lenis crepitans,” 5. 764 “creber adspirans,” at the same time that it enables us to account easily for the two other readings. See also on G. 3. 28.
 Evander's yearning after the glories of his youth is modelled on two speeches of Nestor's, Il. 7. 132 foll., 11. 670 foll., though the Virgilian hero is much briefer than the Homeric.
 Qualis eram in loose apposition with ‘praeteritos annos.’ ‘Cum primam’ is generally taken as i. q. “cum primum,” itself the reading of some copies, apparently however of none of Ribbeck's, though it has been attributed to Rom. But it may be questioned whether it does not mean the front rank, which would be supported by Il. 11. 675, ὁ δ᾽ ἀμύνων ἡσι βόεσσιν Ἔβλητ᾽ ἐν πρώτοισιν ἐμῆς ἀχὸ χειρὸς ἄκοντι. Comp. 7. 531 note, 10. 125. ‘Praeneste’ fem. by synesis, Madv. § 412. So “gelida Praeneste” Juv. 3. 190. ‘Sub ipsa,’ under its very walls, the enemy probably being driven back there.
 As Serv. remarks, this burning of the spoils was a Roman practice supposed to have been introduced by Tarquinius Priscus, who, after a victory over the Sabines, burned their shields in honour of Vulcan. For instances in later history Larsch § 48 cites Livy 8. 30., 23. 46., 45. 33 &c. Comp. Plutarch, Marius 22. The spoils of the Latins are burnt 11. 193, as a sort of compensation to the Trojan dead. In Hom. arms seem only to be burnt in honour of the person who wore them.
 Feronia 7. 800.
 Erulus seems to have had three lives, not, like his prototype Geryon, three bodies, though it is difficult to distinguish the conceptions, at least if Aesch. Ag. 869 foll. is right in giving Geryon a separate life for each body. If we take ‘animas’ strictly, we must suppose ‘terna arma movenda’ to be a simple consequence of the three lives: having been killed, he could get up and fight again. Preller, Röm. Myth. p. 693, comp. the story of the centaur Mares, the first inhabitant of Ausonia, told by Aelian 9. 16. Serv. gives as a choice “‘movenda’ vel contra ipsum vel ab ipso,” and Peerlkamp and Ribbeck embrace the former alternative: but Forb. remarks justly that ‘totidem exuit armis’ is in favour of the latter, which is the ordinary view.
 Sternendus erat, he required to be laid low. Comp. Plaut, Bacch. 1. 1. 31, “Ah nimium ferus es. Mihi sum. Malacissandus es.” It is doubtful whether ‘leto.’ is dat., as Wagn. thinks, or abl. “Neque aversos dignatur sternere morti” 12. 464 is in favour of the one, “Sterneret ut subitaturbatam morte Camillam” 11. 796, “Sternere caede viros” 10. 119, of the other. Comp. G. 4. 432. ‘Tunc’ is here found in all Ribbeck's MSS. except one cursive. See on G. 2. 317 (2nd edition).
 It seems most natural, with Haupt and Ribbeck, to make this the apodosis of v. 560. Evander says that if he were as he once was he should go to combat by his son's side. Comp. Il. 7. 157, 8. ‘Usquam’ nearly i. q. “umquam:” comp. 5. 853 (note), “nusquam amittebat.” If there is any special force in the word here, it must mean neither at home nor on the battle-field.
 “Nate, tua” opens a line similarly 6. 689. “Finitimo” Med., Rom., Gud., ‘finitimos’ Pal. orginally, which Ribbeck adopts, taking it apparently in apposition with ‘funera.’ The old reading, ‘finitimus,’ which Heyne recalled, seems found only in one or two inferior MSS. Serv. interprets ‘finitimo:’ but his note is apparently confused with another which rather points to ‘finitimos,’ as he calls attention to Evander's feeling towards his neighbours as showing heroic unselfishness. But ‘finitimo’ is doubtless the true reading, being constructed not, as Burm. thought, with ‘ferro’ but with ‘capiti,’ which thus forms a periphrasis such as we see in Greek plays. ‘Usquam,’ Pal., Rom., Gud., was restored by Heins.; but Wagn. rightly recalled ‘umquam,’ the reading of Med. Gud. has also a variant ‘hostis.’
 Pierius' note on ‘viduasset’ may amuse the reader: “Servius ait, proprie ‘viduasset’ dictum a Vergilio, quia urbs est generis feminini: abusive vero et satis incongrue ab Horatio dictum ‘viduus pharetra Risit Apollo.’ Quia scilicet putat ipse ‘viduam’ quasi ‘viro iduam’ dici. Atqui sunt ex grammaticis quorum auctoritas minime contemnenda est qui ‘viduam’ a ‘ve’ et ‘duitate’ dictum velint, sicut ‘vesanus’ non sanus, atque ita non incongrue dixerit Horatius ‘viduus Apollo.’” Serv.'s supposed etymology is so far nearer the truth that “viduus” has the same root as “dividere” and “iduare.” ‘Viduo’ occurs Lucr. 5.840, where it is constructed with a gen., as here with an abl. ‘Urbem’ can hardly be Agylla, as Forb. thinks, as unless we read ‘finitimos,’ there is nothing in the context to favour Serv.'s supposition mentioned on v. 569 that Mezentius' treatment of his subjects would have concerned Evander. Mezentius was an ally of Turnus, and Turnus was an enemy of Evander, so that we may readily suppose that Mezentius had been a scourge to his neighbours of Pallanteum. Gossrau's solution, that Mezentius had persisted in his tyranny in defiance of Evander's counsels, seems quite gratuitous.
 Miserescere, in earlier Latin impersonal, is personal here, as in 2. 145.
 Comp. generally 4. 612. The gods and fate are made co ordinate, as in v. 512.
 “Urbem orant: taedet pelagi perferre laborem” 5. 617. ‘Patior’ Pal. (originally), Rom., Med., ‘patiar’ pal. (corrected), Gud. Evander speaks as one whose life is already a burden, as Serv. reminds us, comparing Ter. Phorm. 4. 1. 9 “senectus ipsa est morbus.” ‘Durare’ strictly to harden, transitively, hence harden one's self, hold out: hence, as we say, to endure, transitively. This last use is rare, and apparently not pre-Augustan. “Vix durare carinae Possunt inperiosius Aequor” Hor. 1 Od. 14. 7.
 It is difficult to decide between ‘nunc o nunc,’ Rom., and ‘nunc nunc o’ Med., Gud. Pal. has ‘nunc nunc o,’ ‘o’ having been originally omitted. On the whole Ribbeck seems right in preferring the reading of Rom., which makes ‘nunc,’ the paramount thought in Evander's mind, more emphatic, though Gossrau thinks differently. “Crudelem abrumpere vitam” 9. 497: comp. 4. 631.
 Sola et sera Med., Rom. (which seems to have been read by Serv., though on 9. 482 he quotes the other order), ‘sera et sola’ Pal. There seems no means of deciding between the two. Wagn. restored the former, Ribbeck recalls the latter. “Senectae sera meae requies” 9. 482. “Ea sola voluptas” 3. 660.
 ‘Conplexu’ Pal., Med. first reading, Gud., ‘conplexus’ Rom., Med. second reading. Euphony is perhaps in favour of the former. With the latter comp. 2. 490, “Amplexaeque tenent postis.” ‘Neu’ Pal. (originally), Med., Rom. was restored by Wagn. for ‘ne.’ ‘Nuntius’ here as elsewhere in Virg. may be either the messenger or the message: see on 4. 237.
 Comp. generally 4. 391, 2. The imperfeets are to be noticed, showing that the old man fails and is carried away while he is yet speaking.
[585-607] ‘The Trojans and Areadians march from the city towards Caere, and finally halt at a grove near the Tyrrhenian camp.’