Talis se sata Nocte tulit, terrasque petivit.

Postquam acies videt Iliacas atque agmina Turni,
Alitis in parvae subitam collecta figuram,
Quae quondam in bustis aut culminibus desertis
Nocte sedens serum canit inportuna per umbras;
Hanc versa in faciem Turni se pestis ob ora
Fertque refertque sonans, clipeumque everberat alis.
Illi membra novus solvit formidine torpor,
Arrectaeque horrore comae, et vox faucibus haesit.
At, procul ut Dirae stridorem adgnovit et alas,
Infelix crinis scindit Iuturna solutos,
Unguibus ora soror foedans et pectora pugnis :
Quid nunc te tua, Turne, potest germana iuvare?
Aut quid iam durae superat mihi? qua tibi lucem

860.] "Virgo sata Nocte" 7. 331.

862. It is hardly necessary, with Serv., to press the words 'alitis parvae' so as to make them mean the noctua' or small owl rather than the 'bubo' or large one. The 'bubo was a very ill-omened bird: see Pliny 10. 16-18, who gives instances of the city having been purified on the sight of it. Comp. also the story in Josephus 19. 8 of the owl that appeared before the death of Herod Agrippa I. "Obscenas aves, Maestique cor bubonis, et raucae strigis Exsecta vivae viscera" Sen. Med. 732. For 'subitam ' Pal., Gud. originally, and another of Ribbeck's cursives have 'subito.' Collecta:' comp. Prop. 4. 8. 29, "In tenues humilem te colligis umbras" (Cerda). Pal. has 'coniecto:' Med. 'conversa.'

863, 864.] For the sense and rhythm comp. 4. 462, "Solaque culminibus ferali carmine bubo Saepe queri et longas in fletum ducere voces:" G. 1. 402, "Solis et occasum servans de culmine summo Nequiquam seros exercet noctua cantus," where 'seros' explains 'serum' here as = late into the evening. Umbram' for 'umbras' Minoraug. and Menag. pr. 'Inportuna' G. 1. 470 note.

865.]Versam' Med. a m. s. " In faciem conversa v. 623 above. Ob ora' Rom., with one of Ribbeck's cursives, confirmed by Arusianus, p. 250 L., and Serv. on A. 1. 233: 'ad ora' Pal., Gud., and another of Ribbeck's cursives: 'in ora' Med. a m. s., 'inodora' (a confusion between the two prepositions) a m. p.

866.] Comp. the story of Valerius Corvus and the Gaul in Livy 7. 26.

860

865

870

'Everberat,' beats till he is bewildered and overcome: such seems to be the force of this word, of which Forc. gives no instance before Virg. Quintilian, 2. 4. 18, mentions as a good subject for an exercise in criticisin "an sit credibile super caput Valerii pugnantis sedisse corvum, qui os oculosque hostis Galli rostro atque alis everberaret." "Everberatum mare tandem remis pertinacibus" Q. Curt. 4. 3. 18. Ovid, M. 14. 578, imitates Virg.

867.] Olli' Ribbeck, perhaps rightly, from Bigot. and one of his own cursives: Gud. has 'iolli,' with the o erased. 'Illi' Med., Pal., and Rom., and so Heyne and Wagn. 'Novus' as in G. 4. 357 (note), and A. 2. 228. V. 868 is repeated from 4. 280.

869.] 'Ac' Rom. for 'at.' Procul,' hard by, as E. 6. 16, A. 10. 835. Stridorem et alas' poetical for "stridentes alas :” comp. 11. 801, 802, "Nihil ipsa nec aurae Nec sonitus memor "aurae sonantis."

870.] Infelix' of one under the pressure of a sudden calamity, as 7. 376, v. 598 above. 'Crinis scindit solutos' = "solvit et scindit crinis." Scindit crinis' Rom.

871.] Soror' emphatic by position, like 'genetrix' v. 412 above.

872.] Iuvo' with double acc., as 10. 84, "aliquid Rutulos iuvisse."

873.] Durae' was restored by Heins. for miserae,' which is given as a variant in Gud. 'Durae,' because I could look so long on your struggle: the self-reproach is like that of Anna 4. 681, "sic te ut posita crudelis abessem," which is well comp. by Forb. Heyne refers to Statius, Theb. 12. 214, 215, "Et nunc me duram,

Arte morer? talin' possum me opponere monstro?
Iam iam linquo acies. Ne me terrete timentem,
Obscenae volucres; alarum verbera nosco
Letalemque sonum; nec fallunt iussa superba
Magnanimi Iovis. Haec pro virginitate reponit?
Quo vitam dedit aeternam? cur mortis adempta est
Condicio? possem tantos finire dolores

Nunc certe, et misero fratri comes ire per umbras.
Inmortalis ego? aut quicquam mihi dulce meorum
Te sine, frater, erit? O quae satis ima dehiscat
Terra mihi, Manisque deam demittat ad imos !
Tantum effata, caput glauco contexit amictu,
Multa gemens, et se fluvio dea condidit alto.
Aeneas instat contra telumque coruscat,

si quis tibi sensus, ad umbras, Me tardam
quereris Stygiis, fidissime, divis."

874.] Possim' Rom. 'Monstrum,' as often, of an omen: comp. dewvà méλwpa θεῶν, ΙΙ. 2. 321.

875, 876.] Iam iam' of action just beginning, like "iam iamque" 8. 708. Timentem' is not proleptic. Obscenus,' ill-omened: comp. G. 1. 470, A. 4. 455. "Obscenum omen est omen turpe" Varro, L. L.7.97. The pl. probably represents the bewilderment of Turnus' mind. Wagn.'s explanation, “una ex earum genere quae obscenae sunt," would equally suit the sing.

877.] Fallunt,' escape me. 'Solum' Med. a m. p. for 'sonum.'

878.] Magnanimi' ironical, as v. 144 above. Repono' = “rependo," as in Cic. ad Fam. 1. 9. 19, "ne tibi ego idem reponam, cum veneris." See Mayor on Juv. 1. 1 (2nd edition). Wagn. and Ribbeck rightly put a mark of interrogation at the end of the line.

875

880

885

&c.: comp. 4. 317, "Fuit aut tibi quicquam Dulce meum," where nearly the same words give a different shade of sense. Haud' for 'aut' Gud. corrected, with other less important copies.

883.] For the lengthening of the final syllable oferit' before a vowel, see Excursus to this book. For 'quae' Pal. and originally Gud. have 'quam.' 'Satis alta' Minoraug. and another of Ribbeck's cursives, with the Medicean and Porcian of Pierius: so Heins. and the edd. after him. till Ribbeck, who rightly restores satis ima' from Med., Rom., Pal. corrected, and Gud. See on 10. 675, where the same words occur. Here, as there, dehiscat' is given by a large majority of the best copies: Pal. originally, however, gives 'dehiscet' here, which is adopted by Ribbeck.

884.] Deam' emphatic by its position. 'Demittit' Pal. originally, 'demittat' Pal. corrected, Med., Rom., and Gud. 'Demittet' Ribbeck, from one of his cursives.

885.]"Glauco amictu' quasi propter undarum similitudinem" Serv. "Eum glauco velabat amictu Carbasus" 8. 33 of the river-god Tiberinus. See on 10.205.

,

886.] Fluvio' must apparently be taken, not of the Tiber, but of the river flowing out of Juturna's lake, though the epithet 'alto' seems exaggerated. Perhaps, as Mr. Munro has suggested, Virg. was thinking of his own line G. 4. 428, "Haec Proteus, et se iactu dedit aequor in altum."

887-952.] Aeneas and Turnus meet for their final encounter. Turnus, who is numbed and bewildered by the presence of the Fury, is at length struck down by a

Ingens, arboreum, et saevo sic pectore fatur :

890

Quae nunc deinde mora est? aut quid iam, Turne, retractas ?
Non cursu, saevis certandum est comminus armis.
Verte omnis tete in facies, et contrahe, quidquid
Sive animis sive arte vales; opta ardua pennis
Astra sequi, clausumque cava te condere terra.
Ille caput quassans: Non me tua fervida terrent
Dicta, ferox di me terrent et Iuppiter hostis.
Nec plura effatus, saxum circumspicit ingens,
Saxum antiquum, ingens campo quod forte iacebat,
Limes agro positus, litem ut discerneret arvis ;

wound from Aeneas' spear. He begs for
life: Aeneas is half moved by the entreaty,
until his eyes light on the belt of Pallas,
which was conspicuous on his enemy's
shoulder. The sight decides him to deal
the fatal blow.'

887, 888.] • Contra as opposed to Juturna, who has been helping her brother. 'Arboreum,' huge as a tree: comp. "centena arbore” of huge oars 10. 207. • Ingens' is joined by Serv. (whom Wagn. follows) with Aeneas; but it is better to join it with 'telum:' comp. 8. 622, of Aeneas' corselet, "Sanguineam, ingentem" &c. In such passages Virg. allows himself the use of a double epithet. Schrader conj. 'roboreum.'

889.] Nunc deinde is singular: what delay have you now to hope for?" "quae nunc mora est quae tibi deinde videtur eventura ?" 'Retracto' intrans., as in Livy 3. 52, Cic. Tusc. 1. 31. 76.

890.] Ribbeck thinks this verse should be placed after v. 893. On certandum est comminus' Serv. quaintly remarks, "Sed hoc ideo dicit, quia volnere tardus magis comminus pugnare desiderat."

891.] Παντοίης ἀρετῆς μιμνήσκεο· νῦν σε μάλα χρὴ Αἰχμητήν τ' ἔμεναι καὶ θαρσαλέον πολεμιστήν, ΙΙ. 22. 269.

893.] Astra sequi pennis' like "sidera voce sequentem 10. 193 note. Clausumve' Pal. and Menag. pr., and so Heyne and Ribbeck: clausumque,' which is more idiomatic, Med., Rom., Gud., with two other of Ribbeck's cursives. Que' is defended by Wagn. Q. V. 36. 10, who rightly adopts it: see on 10. 709, where there is a precisely similar case, and where, as here, Ribbeck follows Pal. in reading 've.'

894.] Caput quassans' Lucr. 2. 1164 of the old husbandman who sighs over the degeneracy of the race. 'Quassans caput" of Juno in anger 7. 292.

895

895.] "Ω πόποι, ή μάλα δή με θεοὶ θάναTóvde káλeσσav, says Hector, Il. 22. 297: but the language more nearly recalls Il. 17. 175, where Hector says, in answer to the reproaches of Glaucus, Obтoι èyà eßßiya μάχην, οὐδὲ κτύπον ἵππων· Ἀλλ ̓ αἰεί γε Διός κρείσσων νέος αιγιόχοιο, Ος τε καὶ ǎλкιμov áνdρa poßeî &c.; and 16. 844, where Patroclus says to Hector, 'Hồn vûv, Εκτορ, μεγάλ' εὔχεσ· σοὶ γὰρ ἔδωκε Νίκην Ζεὺς Κρονίδης καὶ ̓Απόλλων, οἵ με δάμασσαν 'Pnidíws. See also Il. 13. 811-12.

896-898.] This passage is modelled partly on II. 21. 403 foll. (of Athene in the battle of the gods), 'H d'àvaxaσoaμévn λίθον εἵλετο χειρὶ παχείῃ Κείμενον ἐν πεδίῳ, μέλανα, τρηχύν τε μέγαν τε, Τόν ῥ ̓ ἄνδρες πρότεροι θέσαν ἔμμεναι οὗρον αρούρης : partly on Il. 12. 445 foll., where Hector easily brandishes a stone which two men of a later age could hardly lift on to a waggon. Comp. Il. 5. 303 foll., where this language is slightly varied. Wagn. was inclined to object to the repetition of ingens,' which is however by no means pointless. In v. 897 Goth. pr. and some inferior copies have qui,' and so apparently Serv., who quotes from Sallust (Cat. 55), locus in carcere quod Tullianum appellatur." Heins. first restored 'quod,' which has both authority and grammar in its favour. Med. a m. p. gives q.,' a m. s. quod.' 'Litem ut discerneret arvis' is difficult, as discernere ' does not often bear the sense of decernere.' Forc. quotes Calp. Ecl. 2. 27, "Nec mora, discernunt digitis, prior incipit Idas:" and Nemes. Ecl. 1 (or Calp. 8). 52, “ruricolum discernere lites." Arvis' might be taken either as dative ('for' concerning') or as abl. (in'). It is quite possible that Virg., after his fashion, meant to suggest two phrases, "decernere litem" and "discernere arva."

Vix illud lecti bis sex cervice subirent,
Qualia nunc hominum producit corpora tellus;
Ille manu raptum trepida torquebat in hostem,
Altior insurgens et cursu concitus heros.
Sed neque currentem se nec cognoscit euntem,
Tollentemve manus saxumve inmane moventem;
Genua labant, gelidus concrevit frigore sanguis.
Tum lapis ipse viri, vacuum per inane volutus,
Nec spatium evasit totum, neque pertulit ictum.
Ac velut in somnis, oculos ubi languida pressit
Nocte quies, nequiquam avidos extendere cursus

899.] Illum' for illud' Med., with one of Ribbeck's cursives: so the MSS. of Augustine de Civ. Dei 15. 9. Tòv d'ou кe δύ ̓ ἀνέρε δήμου ἀρίστω Ρηϊδίως ἐπ ̓ ἅμαξαν an' ovdeos oxxíoσeiav Oloι vûv BρoToi eio, II. 12. 447 foll. In Apollonius R. 3. 1365 four youths could hardly lift the stone a finger's breadth from the ground.

900.] Virg. amplifies Homer's olo vov &c., by bringing in the notion of the earth's motherhood, so copiously illustrated by Lucr. 5. 820 foll. The idea developed there is that the earth's productive force, like a woman's, wears out with continued child-bearing, and that her later offspring is therefore weaker and punier than the earlier: "Sed quia finem aliquam pariendi debet habere, Destitit, ut mulier spatio defessa vetusto v. 826-7 (where see Munro). Comp. also the language of 2. 1150 foll., "Iamque adeo fracta est aetas, effetaque tellus Vix animalia parva creat, quae cuncta creavit." The language here recalls Lucr. 2. 589, "Tellus habet in se corpora prima."

901, 902.] Torquebat: "bene imperfecto usus est tempore, quia non est perfectum quod voluit" Serv. Ille-heros :' so II. 5. 308, Auτàр öy' hpws Eσrn yv ἐριπών. With altior insurgens' comp. "arduus insurgens" 11. 755, “altior exsurgens" ib. 697. Cursu concitus,' running at full speed.

903.] Τὸν δ ̓ ἄτη φρένας εἷλε, λύθεν δ ̓ ὑπὸ φαίδιμα γυῖα, Στῇ δὲ ταφών &c. of Patroclus when disarmed by Apollo 11. 16. 805. But Virg. is more minute and delicate in his description. Currentem-euntem,' whether he runs or inoves. Peerlkamp's tasteless conjecture 'eundem' would add nothing to 'Se cognoscere' Lucr. 6. 1214. 904.] The readings vary much in this line. Tollentemque' Med. with Gud. originally, and another of Ribbeck's curVOL. III.

the sense.

900

905

sives: 'tollentemve' Pal. and Rom., with Gud. corrected. Then Med., Rom., and originally Gud. give 'manus:' Pal. ‘manu,' and so two of Ribbeck's cursives. Manu Heyne, Wagn., Forb., and Gossr.: 'manus' Ribbeck, more rightly: for Virg. is obviously wishing to describe every movement of Turnus: currentem,' 'euntem,' tollentem manus,' 'saxum moventem ;' and 'tollentem manu saxum et moventem' would only describe one movement, and that by a somewhat meaningless repetition of words. Finally Gud. and one of Ribbeck's cursives give 'saxumque' for

saxumve,' and so Heyne, Wagn., &c., against the decided balance of authority: 'saxumve,' rightly, Ribbeck. For 'moventem' Pal. originally had ‘movebat.'

Rom.

905.] Genua labant' 5. 432. has lavant.' Concrevit' perf. 906, 907.] Ipse,' even the stone fails to do its work. 'Lapis viri' : "lapis a viro proiectus" (Forbiger). 'Inane' (the Lucretian and Ciceronian word for 'void') used of the air, as in v. 354 above. "Non per inane meat vacuum" Lucr. 2. 151. Serv. wished wrongly to join totum' with 'ictum.'

Pertulit ictum' like "vires non pertulit" 10. 786. Nec pertulit' Rom. and one of Ribbeck's cursives, and so Heyne.

908.] The hint for this simile is given Ι. 22. 199 foll., Ὡς δ ̓ ἐν ὀνείρῳ οὐ δύναται φεύγοντα διώκειν, Οὔτ ̓ ἄρ ̓ ὁ τὸν δύναται ὑποφεύγειν, οὔθ ̓ ὁ διώκειν, Ως ὁ τὸν οὐ δύνατο μάρψαι ποσίν, οὐδ ̓ ὃς ἀλύξαι. The rhythm and language recall Lucr. 4. 453 foll., "Denique cum suavi devinxit membra sopore Somnus, et in summa corpus iacet omne quiete, Tum vigilare tamen nobis et membra movere Nostra videmur” &c. "In somnis" is also a favourite phrase of Lucr.: see Munro on 3.431. 'Pressit:' 6. 521," Pressit iacentem Dulcis et alta quies." 909.] Extendere cursus' is somewhat H h

Velle videmur, et in mediis conatibus aegri
Succidimus; non lingua valet, non corpore notae
Sufficiunt vires, nec vox aut verba sequuntur:
Sic Turno, quacumque viam virtute petivit,
Successum dea dira negat. Tum pectore sensus
Vertuntur varii; Rutulos adspectat et urbem,
Cunctaturque metu, telumque instare tremescit;
Nec, quo se eripiat, nec, qua vi tendat in hostem,
Nec currus usquam videt aurigamve sororem.
Cunctanti telum Aeneas fatale coruscat,
Sortitus fortunam oculis, et corpore toto
Eminus intorquet. Murali concita numquam

Tormento sic saxa fremunt, nec fulmine tanti
Dissultant crepitus. Volat atri turbinis instar

911.] Volet Med. a m. p. for 'valet.' 'Corpore' local abl. for dat.: comp. "si virgineum sufuderit ore ruborem ” G. 1. 430; "mucrone sese induat" 10. 681.

912.] Vox,' voice; 'verba,' distinct words: "voces verbaque" Lucr. 4. 533. In v. 318 above 'voces' and 'verba' seem synonymous.

913, 914.] Quamcumque' Pal. originally. Tum pectore' &c. Heyne thinks"vertit, versat, varia consilia, sensus, animo Turnus." But this does not do justice either to 'sensus' or to 'vertuntur:' the meaning is rather his feelings shift in distraction.' 'Adspectare' of a longing gaze, as in G. 3. 228., A. 5. 615.

916.] Comp. Il. 22. 293 foll. (of Hector), Στῆ δὲ κατηφήσας, οὐδ ̓ ἀλλ ̓ ἔχε μείλινον ἔγχος, Δηίφοβον δ ̓ ἐκάλει λευκάσπιδα, μακρὸν ἀΰσας, Ηιτεέ μιν δόρυ μακρόν· ὁ δ ̓ OŬTi oi éryúlev ev. Letum' Pal. and the MSS. of Rufinianus, p. 258 R.: so Ribbeck: 'telum' Med., Rom., Gud., and so Heyne and Wagn. Telum' is better in itself, and is confirmed by the parody of Ausonius, Cent. Nupt. 92. Instare:' a prose writer would probably have used a participle for this infinitive.

917.] For quo Gud. originally has 'quos,' corrected 'qua.'

918.] Aurigamve' Med., Pal., and one

910

915

920

of Ribbeck's cursives: so Heyne and Ribbeck, probably rightly: 'aurigamque' Rom. and Gud., followed by Wagn. Juturna corresponds in some measure to Deiphobus in'Il. Ì. c.

6

919.] Cunctanti' dat., as if "minatur" stood for coruscat:' or it may = 'in cunctantem.'

920.] Homer (II. 22. 321) simply says Εἰσορόων χρόα καλόν, ὅπη εἴξειε μάλιστα. Fortunam' is explained by Heyne (following Serv.) as="locum quem fortuna dabat:" a simpler way would be to take it in its ordinary sense, sortitus fortunam oculis' meaning 'having hit upon success with his eyes,' i. e. by looking about for it. Fortuna' of the success of a weapon 10. 422: "(da telo) Fortunam atque viam duri per pectus Halaesi."

921.] Murali,' for destroying walls: so "falces murales," Caes. B. G. 3. 14.

922.] Fremunt' of the noise of the stones against the wall. Tanto' Pal. for 'tanti.'' Fulmen' is "the stroke or bolt or fire of thunder" (Munro on Lucr. 6. 160 foll.), not merely the noise : ' fulmine, in or with the bolt. Virg. may have been thinking of Luer. 6. 329, where the swiftness of the bolt is compared to missiles: "validis quae de tormentis missa feruntur."

923.] Dissultant' of the bursting sound: rumpuntur' would have been the more ordinary expression: comp. Soph. Teuc. 2 (fr. 517 Nauck), Bрorth 8' ἐῤῥάγη δι' ἀστραπῆς: and Il. 16. 78, Εκτορος (φωνή) . . . Τρωσὶ κελεύοντος περιάγνυται.