Dorville wished to read ‘exarsit et’ for ‘exarserat:’ but the two ablatives are constructed in different ways, as in G. 3. 439 &c., though it is not easy to choose among possible constructions. Perhaps ‘atro felle’ is best taken as attributive, ‘furiis’ being causal or modal. Comp. Il. 1. 103, μένεος δὲ μέγα φρένες ἀμφιμέλαιναι Πίμπλαντ᾽. ‘Exarserat’ in past time answers to the instantaneous perf. in present: comp. 2. 257.
 Aetherii was introduced by Burm. and retained by Heyne: but in the principal MSS. where it occurs (Med. a m. p., Gud., and another of Ribbeck's cursives) ‘et’ is omitted, which shows the origin of the corruption. Wagn. also observes that Olympus alone is called “aetherius,” other mountains ‘aerii.’ Either epithet is an exaggeration as applied to the Aventine. ‘Cursu petit’ 2. 399 &c.
 Turbatus is applied to different emotions, and here to fear. For its combination with ‘oculis’ comp. Eur. Iph. A. 1127, σύγχυσιν ἔχοντες καὶ ταραγμὸν ὀμμάτων, where grief seems to be meant, and for the eyes, as affected by fear, Soph. Aj. 139, “μέγαν ὄκνον ἔχω καὶ πεφόβημαι Πτηνῆς ὡς ὄμμα πελείας”, and perhaps Aesch. Pers. 168, ἀμφὶ δ᾽ ὀφθαλμοῖς φόβος. It would be possible to construct ‘oculis’ with ‘videre:’ but the abl. would be weak. Serv. mentions another reading ‘oculi,’ which is found in a few inferior MSS. and adopted by Gossrau: but Evander does not elsewhere speak of himself as present at the scene. Gud. originally has ‘oculos,’ which is approved by Heins. and Heyne. “Ocior Euro,” 12. 733, Hor. 2 Od. 16. 24.
 Pedibus—alas. It does not appear what is the original source of this metaphor, which has since become so common. The image of wings in Greek is used rather to express raising from the ground than carrying along (comp. Il. 19. 386, of Achilles, where perhaps the two notions are combined); and so where ἀναπτερόω is said of fear (Eur. Supp. 89) it expresses fluttering, not speed.
 “Dentibus infrendens” 3. 664. Rom. has ‘frendens,’ an aberration which should be taken into account in estimating the probabilities of the reading in such passages as 4. 54. “Fervidus ira” 9. 736.
 “Lassa resedit” 2. 739. The compound seems to express sitting down after doing anything, as here after exertion.
 The rock inclined to the river, which was on its left: Hercules pushed it from the other side (‘dexter’), and made it fall into the river. ‘Iugo’ with ‘prona.’ ‘Ut’ like “ut forte,” as it happened to incline.
 In adversum nitens like “in medium niti” Lucr. 1.1053. ‘Adversum’ might mean the side opposite to that which inclined towards the river, i. e. the right side; but it seems better to understand with reference to Hercules, who pushes full against it.
 Comp. generally 2. 483 foll.
 Si qua may be taken as “si quando,” and referred to the class of usages noticed on 1. 181, E. 1. 54. But it may equally well express the mysterious nature of the agency. Comp. 2. 479 “qua vi maria alta tumescant.”
 Imitated from Il. 20. 61 foll. (of the earthquake caused by Poseidon), Ἔδδεισεν δ᾽ ὑπένερθεν ἄναξ ἐνέρων Ἀϊδωνεύς: Δείσας δ᾽ ἐκ θρόνου ἆλτο, καὶ ἴαχε, μή οἱ ὕπερθεν Γαῖαν ἀναρ᾽ῥήξειε Ποσειδάων ἐνοσίχθων, Οἰκία δὲ θνητοῖσι καὶ ἀθανάτοισι φανήῃ Σμερδαλἔ, εὐρώεντα, τά τε στυγέουσι θεοί περ.
 External authority is in favour of the omission of ‘que,’ Med., Pal., Gud. giving ‘trepident.’ Ribbeck adopts it: but Wagn. seems right in retaining ‘que,’ as the asyndeton does not suit a dependent sentence like this, though it is natural in an ordinary comParison expressed in the indicative. To make ‘trepident’ the apodosis is not to be thought of. ‘Trepidentque’ is supported by Rom., ‘trepidantque,’ the ind. being evidently a mere error, just as v. 24 Rom., Pal., Gud., and the second reading of Med. have ‘reserat.’
 Ribbeck seems right in omitting ‘in’ before ‘luce’ with Pal. and the original readings of Med. and Gud. The meaning then will be that Cacus is surprised by, not in the light.
 Rudentem properly used of the bellowing of an animal, here of “semihominis Caci” v. 194. ‘Insueta’ may mean that his roars were now for the first time those of terror (comp. v. 222): but it seems rather to mean strange sounds such as are not wont to come from one in human form, though they may have been his usual utterances. So ἀήθης is used of things unpleasant.
 Telis may be used generally, or may refer to his arrows.
 For ‘molaribus’ comp. Il. 12. 161, κόρυθες βαλλόμεναι μυλάκεσσι. The word, which occurs again Ov. M. 3. 59, seems to be poetical, the prose expression being “molaris lapis.” ‘Molaribus instat’ like “instant verbere” G. 3. 106.
 Iecit Rom., Med., ‘iniecit’ Pal., Gud. The former, which Wagn. restores, seems better in combination with ‘per ignem:’ but ‘iniecit’ would mean “iniecit antro,” and the substitution of the simple verb for the compound may have originally been accidental: see on v. 230 above. Elsewhere when ‘iniicere’ is used in Virg. it is followed by a dat. (9. 553) or ‘in’ with acc. (2. 408).
 ‘In nodum conplexus,’ twining his arms and legs round him. Ov. M. 9. 58 (of wrestlers) “vix solvi duros a corpore nexus.” It was thus, as Heyne observes, that Hercules killed the Nemean lion and Antaeus. Prop. and Ov. make him use his club.
 Elidere is the proper word for strangling: see Bentley on Hor. 3 Od. 27. 59; where however “laedere collum,” not “elidere collum,” which Bentley coniectures is the proper reading: ‘laedere’ being used in its primary sense of crushing. ‘His strangled eyes’ if course means ‘his eyes starting out of his head from strangulation.’ ‘Angere’ in its strict sense G. 3. 497. As Serv. remarks, it is used less properly with ‘oculos’ than with ‘guttur.’ The notion of ‘siccum sanguine guttur’ seems to be that it was the stoppage of blood rather than breath which caused death. ‘Siccae sanguine fauces,’ in different sense, 9. 64. Sen. Ep. 59. 12 uses “siccum volnus” of a wound from which the blood cannot escape.