“Nemorum Latonia custos” 9. 405. “Cultor nemorum” G. 1. 14, where however ‘cultor’ is not simply an inhabitant, but a cultivator. Phaedr. 2. 4. 3 has “sus nemoricultrix,” like “cerva silvicultrix” Catull. 61 (63). 72.
 Ipse seems to mean ‘I, who have the most right, as making a vow of what is my own.’ So Serv. “Bene ‘ipse pater,’ quoniam auctoramenti potestatem nisi patres non habent.” Serv. takes ‘tua’ with ‘supplex,’ like ἱκέτης σός; but this could hardly stand here unless ‘prima’ were constructed similarly. ‘Tua tela,’ because the weapon is dedicated to Diana, “donum Triviae” v. 566. These very words contain the dedication, which is made in Virg.'s characteristically indirect manner. Metabus himself ceases to be a warrior, and becomes a wild man, and it is not unnatural that at this moment he should speak of a war-spear as if it might be a hunting-javelin. ‘Tua tela’ will then be like “tua quercus” 10. 423 on the one hand, while on the other it may be comp. with “nostris armis” v. 536. ‘Prima,’ the first weapons she holds are thine.
 Tuam, thy servant. Comp. Hor. 3 Od. 4. 21, “Vester, Camenae, vester in arduos Tollor Sabinos,” and contrast “meus” Pers. 5. 88, my own man, or, as we more commonly say, my own master. With ‘dubiis committitur auris’ comp. “incertis committere ventis” Lucr. 5.782.
 Inmittit sends into the air, G. 2. 364. ‘Sonuere undae’ is generally taken, resounded with the noise of the spear, an exaggeration which would be sufficiently after Virg.'s manner. But it would be more poetical, though perhaps more modern, to understand the words of the roaring of the swollen flood, which would intensify the danger of the critical moment. At the same time a sort of parallel may be intended between the noise of the water and the hurtling of the spear, ‘iaculo stridente.’ Serv. says “‘sonuere undae:’ hic distinguendum,” from which it appears that some wished to point after ‘amnem.’ ‘Rapidum’ again enforces the notion of danger, and perhaps is intended to parallel the motion of the spear.
 Doubts why Camilla is called ‘infelix’ have been raised since the time of Serv.; but it evidently refers not to her future destiny, which would be flat, but to the moment during which she is sent on her perilous venture.
 Donum Triviae refers to the spear as well as to Camilla: see on v. 558. Cerda understands it as the gift not to but of Diana, who had granted his prayer, and presented him with his daughter's life (comp. 2. 31 note); but this is less likely. For ‘vellit’ Pal. and Gud. have ‘tollit,’ the latter with ‘vellit’ as a variant, conceivably, as Ribbeck suggests, from G. 4. 273.
 For ‘ullae’ Axt and Peerlkamp conj. ‘villae:’ a slight enough change, but the other is more Virgilian. Forb.'s objection that the ‘villa’ was not known to the heroic ages might be got over, as Virg. is not consistent in that respect. For ‘tectis, moenibus accepere’ see on 7. 210.
 “‘Manus dedisset,’ consensisset,” Serv. Heyne says “manus dat proprie victus, tum omnino qui cedit, qui placatur, h. l. qui quod humanitas aliorum vel hospitalitas offert accipit.” ‘Feritate’ Madv. § 255, who gives other instances of the abl. of “the efficient cause in the agent himself by which a thing is done.” Here we may either make ‘non dedisset’ a positive notion, or say that ‘feritate’ is a restraining cause, like “prae feritate.” Rom. has ‘nec’ for ‘neque.’ ‘Neque dedisset’ 9. 704 note.
 Pastorum with ‘montibus’ (comp. G. 3. 476, “desertaque regna Pastorum et longe saltus lateque vacantis”), not, as Serv., Wagn. &c. think, with ‘aevum,’ which would be rather a questionable expression for “aevum pastorale.” At the same time the position of ‘pastorum’ shows that it is meant rhetorically to colour the whole verse. ‘Et,’ which Brunck and Wakef. questioned, means ‘and so.’ ‘Exigere aevum’ 10. 53.
 ‘In dumis’ = “inter dumos,” much as in G. 4. 130. “In silvis inter deserta ferarum Lustra” 3. 646. ‘Horrentia’ because of the wood. ‘Hinc,’ which Burm. introduced for ‘hic,’ seemingly without authority, was the original reading of Pal.
 Nutribat for ‘nutriebat,’ like “lenibat,” “polibat” &c., Madv. § 115 b. With ‘teneris inmulgens ubera labris’ comp. Liv. Andr. inc. 9, “Quem ego nefrendem alui, lacteam inmulgens opem,” doubtless from the Aegisthus, being a translation of Aesch. Cho. 897, μαστόν, πρὸς ᾧ σὺ πολλὰ δὴ βρίζων ἅμα Οὔλοισιν ἐξήμελξας εὐτραφὲς γάλα.
 Institerat vestigia is really a cogn. acc., “insistere pro insistendo facere,” as Forb. explains it. But this construction is so rare in Latin as compared with Greek, that we can hardly suppose that Virg. intended more than a variety of the construction with the acc. of the object, which we have had in 6. 563, G. 3. 164, ‘vestigia’ being identified with the ground on which she planted her feet. Comp. Lucr. 1.406, “Cum semel institerunt vestigia certa viai.” ‘Pedum plantis’ will be abl. of instr., though Heyne is doubtless right in supposing that Virg. meant an allusion to the construction with the dat. without any acc., ‘plantis institerat.’ ‘Pedum plantis’ occur together 8. 458; here however ‘pedum’ might go with ‘vestigia,’ as in Lucr. 3.4, “Fixa pedum pono pressis vestigia signis.” ‘Ut primis’ = “ut primum” (4. 259 &c.), as Wagn. remarks.
 Oneravit, the common reading before Wagn., is found in one of Ribbeck's cursives, and confirmed by Serv., who says “onus enim quicquid teneris inponitur manibus,” without noticing the existence of a variant. It can hardly be doubted that the word came, as Wagn. supposes, from a recollection of 10. 868. Possibly Serv. may have written his note from memory, as otherwise he could scarcely have failed to mention ‘armavit,’ though he might not have approved of it. Forb. may be right in pressing the pl. ‘palmas,’ as if she carried in two hands what a grown person would have carried in one, though ‘tela puerilia’ v. 578 is a little against this.
 Crinali auro: see on 4. 138. ‘Crinale’ is found alone Ov. M. 5. 53, “(ornabat) maDidos myrrha curvum crinale capillos,” where the epithet suggests a “fibula,” or perhaps a “circulus” (10. 138), rather than, as Forc. thinks, an “acus.” In what sense the tiger's skin supplied the place of the ‘crinale aurum’ is difficult to see. Perhaps the head of the beast formed a cap, as in the case of the lion's skin 7. 667. Cerda refers to Val. F. 6. 704 foll., where when a personage clad in tiger's skin is slain, it is said “perquam optima fictione” (as he thinks), “subitos ex ore cruores Saucia tigris hiat vitamque effundit herilem.” But Virg. may have written loosely, intending no more than that a tiger's hide was Camilla's only ornament. For a beast's hide worn by a hunter as a chlamys comp. v. 679 below, 1. 323. Meantime one inferior MS. has ‘vittae’ for ‘pallae,’ doubtless from 7. 352. The ‘palla’ was long: see on 1. 404.
 Tyrrhena per oppida: Heyne remarks that Campania, which lay on the other side of the river Amasenus, was formerly inhabited by the Etruscans. Gossrau refers to Müller's Etrusker, Einl. p. 4. ‘Multae illam matres optavere nurum,’ a variety for saying ‘she had many suitors,’ 7. 54. The parents, we may remember, in ancient times generally made the match. For the expression Cerda comp. Catull. 60 (62). 42, “Multi illum pueri, multae optavere puellae.”
 Diana: see on v. 536 above.
 Tali, opposed to service in the woods. ‘Conata’ virtually = “et conata fuisset.” ‘Conata lacessere:’ she actually provoked the Trojans: but as her military efforts were abortive, Virg. has chosen a word denoting endeavour, rather than such a word as “ausa.” Serv., after explaining the force of the participle ‘conata,’ goes on to say, “Hac autem ratione ostendit quare non suscensa a Troianis,” where the last words are marked by the editors as corrupt. All will be right if for ‘suscensa a’ we read ‘suscenseat,’ the meaning being that Diana mentions Camilla's having provoked the combat to show why she does not resent the conduct of the Trojans in fighting with her favourite.
 The commentators are at some pains to reconcile ‘cara foret’ with “cara mihi ante alias” above v. 537. But there is no real difficulty. Diana means that if Camilla had remained with her, she would have been able to treat her with fondness, which now she cannot do. Wakef. read ‘cura’ from a single inferior MS. ‘Vellem’ &c. virtually contains the protasis of which this line is the apodosis.
 Pierius reads ‘urguemur,’ which however does not appear in any MS. or early edition. For ‘acerbis’ Gud. has a variant ‘iniquis,’ which is the reading of some inferior copies, probably a recollection of 2. 257., 3. 17, or 10. 380. For a similar variety see on 1. 668. ‘Acerbus’ of premature death, as 6. 429, v. 28 above.
 All Ribbeck's MSS. but one cursive read ‘Italusque.’ In Med. however ‘q.’ (sic) is written over an erasure, and Serv. appears to have read ‘ve,’ his explanation being “sive eam Troianus interemerit sive aliquis de Aeneae auxiliis.” ‘Que’ might be defended, the penalty being looked upon as one which menaced all alike, ‘pariter,’ so long as the action was contingent: but ‘ve’ is much more natural. The question is one of a class on which the authority even of the best MSS. is inconclusive. See on 10. 108. ‘Italus’ must refer, as Serv. takes it, to Aeneas' Italian auxiliaries. He mentions however a notion which some had that Arruns was a partisan of Turnus, who was jealous of the military success of one of the weaker sex. ‘Det sanguine poenas’ 2. 366.
 Forc. quotes another instance of ‘inspoliatus’ from Quint. 7. 1. ‘Spoliare’ generally takes an acc. of the person: but Forc. gives two or three instances where it has one of the thing. With the line generally comp. 4. 392, “Marmoreo referunt thalamo stratisque reponunt.” ‘Patriae’ a sort of local dative, such as is used to express motion to. It is doubtless a vestige of the locative, though Virg. probably did not think it so. See on 6. 84. This seems better than with Forb. to take ‘reponere’ as = “reddere.”
 Levis with ‘auras.’ It is difficult to decide between ‘delapsa’ (Med.) and ‘demissa,’ supported by the rest of Ribbeck's MSS., two cursives having ‘dimissa.’ The former is perhaps better in itself, but the latter is sufficiently good, whether we take it as passive, sent down by Diana, or middle, sending herself down. Neither can be accounted for with certainty by supposing a transcriber to have thought of other passages, as though the words occur elsewhere in Virg. with similar applications, the verbal resemblance between any of the passages and the present is not great. On the whole, unless external authority is to go for nothing, it seems safer to read ‘demissa.’ W. Ribbeck quotes “at illa levis caeli delapsa per auras” from Epit. Il. 95, which, if itself beyond suspicion, doubtless shows that the epitomator found ‘delapsa’ in his copy of Virg.
 Insonuit either from the rapidity of her flight, as Heyne thinks, or from the motion of her bow and arrows. The wellknown passage Il. 1. 46 is in favour of the latter, especially as ‘nigro circumdata turbine corpus’ seems to be from ὁ δ᾽ ἤϊε νυκτὶ ἐοικώς. The dark storm-cloud is doubtless meant, as Serv. says, to be in keeping with the errand of vengeance. So Juno 10. 634 (note) comes down “agens hiemem nimbo succincta per auras.” “Turbine nigro” G. 1. 320: comp. A. 1. 511., 10. 603., 12. 923.
[597-647] ‘The Rutulian and Trojan cavalry meet. After various advances and retreats, they engage in earnest.’