A description, as Henry rightly takes it, not of the temple but of the adytum, which, as at Delphi, was a cavern in the rock. ‘Euboicae rupis,’ the rock or hill of Cumae: see on v. 9. ‘Latus rupis excisum in antrum’ is a variety, as Heyne observes, for “antrum excisum in latere rupis.”
 Aditus and ‘ostia’ seem rightly explained by Henry as a sort of Virgilian hendiadys, ‘aditus per centum lata ostia.’ But it is not easy to understand what these entrances were. On the whole the consistency of the description seems to require that we should understand them to be the entrances of the adytum, opening into the temple (comp. 3. 92, where the ‘adytum’ is opened similarly at the giving of the response): but a hundred doors communicating from one side of the temple to a cavern beyond form a picture which is not readily grasped. Meanwhile the general tenor of the narrative is well illustrated by a graphic description of a worshipper at Delphi approaching the ‘adytum’ in the Oxford ‘Arnold Prize Essay’ for 1859, by my friend Mr. Bowen of Balliol College. I quote it in an Appendix to this book, as it is too long for a note.
 Ruunt expresses the general practice: through these doors the responses of the Sibyl are habitually communicated.
 Limen, sc. ‘antri;’ whether identical with any of these doors we are not told. The Sibyl goes into the cave (comp. v. 77); Aeneas and the Trojans remain outside. ‘Poscere fata’ is explained by what follows, v. 52. The sacrifices had been performed, but prayer was still necessary to obtain the response, and this was the time for prayer, the god having already manifested himself. The words seem to mean ‘to ask Apollo for oracles,’ ‘fata’ being used as in 1. 382 &c. Comp. G. 3. 456, “meliora deos sedet omina poscens,” and possibly A. 3. 456, where however see note. Elsewhere, as in 7. 272 &c., the fates themselves are said ‘poscere.’ “Tempus poscere” 9. 12. For the construction see on G. 1. 213.
 Ante fores like ‘ad limen.’ ‘Unus’ = ‘idem,’ with which it is not unfrequently joined: see Forc. The sense is not that her countenance and colour keep changing, but that they are different from what they were before.
 Comptae: Heyne remarks that her hair would be already unbound, as the sacrifice had been made (see on 3. 370), so that Virg. must here mean that the hair stood on end or was tossed about. But we need not press the poet so closely. Unbound or dishevelled hair was usual when a priest or prophet approached the gods: and Virg. has chosen to represent the hair of the Sibyl as becoming disordered at this particular point of the story.
‘Rabie’ with ‘tument.’ As the
forms of the gods and of the dead were
supposed to be larger than those of ordinary
humanity (see on 2. 773), so the Sibyl
seems to increase in stature under the
divine afflatus. In less poetical language
we should say that she rises to her full
height, and every limb is stretched with
excitement. The picture is virtually the
same as that of Wordsworth's Laodamis,
expecting an answer to her prayer:
“Her countenance brightens, and her eye
Her bosom heaves and swells, her stature grows.
 Sonare of a person speaking loudly, 12. 529. With the expression generally comp. 1. 328, “haud tibi voltus Mortalis, nec vox hominem sonat.” ‘Quando’ is causal rather than temporal, so that Heyne's suggestion, adopted by Jahn, to place a period at ‘sonans,’ and connect ‘adflata est’ with what follows, would be no improvement. With the position of ‘quando’ comp. 10. 366, “aspera quis natura loci dimittere quando Suasit equos.” Serv. explains ‘adflata’ “nondum deo plena, sed adflata vicinitate numinis;” but ‘adflare’ and ‘adflatus’ (subst.) are terms regularly used of divine inspiration (see Forc.), like ἐπιπνεῖν and its cognates.
 With ‘propiore’ comp. the use of ‘praesens,’ ‘adesse,’ of divine favour, and the cognate ‘propitius.’ ‘Cessas in vota’ is a variety for the more ordinary use of ‘cessare’ with the abl., as in Cic. Sen. 5, “neque unquam in suo studio atque opere cessavit.” Forb. comp. “audere in proelia” 2. 347. We should expect the construction ‘in’ with acc. after a verb signifying tendency to promote an object rather than the reverse: but the explanation doubtless is that the absence of such a tendency is considered to affect the object in question no less than its presence. The phrase is imitated by Sen., Medea 406, “Nunquam meus cessabit in poenas furor.” “Vota precesque” 11. 158.
 Enim gives the reason why he should pray fervently, and ‘ante’ refers to fervent prayer as implied in its opposite ‘cessas.’ ‘Dehiscent’ is used of the flying open of the doors, in accommodation to ‘ora.’
 The earlier commentators, following Serv., were satisfied with making ‘attonitae’ = “facientis attonitos.” Later editors, who see that both on poetical and grammatical grounds it is to be understood strictly, the house being conceived of and endowed with human feelings, are still divided as to the force which should be given to it, Heyne and Wagn. referring it to the effect of the sudden opening of the doors, Henry to the spell-bound silence which prevents the opening, while Forb. after Süpfl understands it generally of the condition of the cave as possessed by the god. On the whole Henry's interpretation seems to give the most consistent and poetical picture. He compares a similar application of the word in Lucan 2. 21, “Sic funere primo Attonitae tacuere domus, cum corpora nondum Conclamata iacent.” The Sibyl, in describing the feelings of the ‘domus,’ is in effect describing her own. The effect of the inspiration is to bewilder and confound her, so that she cannot at first master herself sufficiently to speak; and so now after a hurried injunction to Aeneas she relapses into her stormy silence.
[56-76] ‘Aeneas invokes Apollo as the patron of Troy and his own guide in his wanderings, praying of him, of the gods who have hitherto opposed Troy, and of the Sibyl herself, that he may at last be allowed to settle in Latium, and promising a temple to Apollo and Diana. He begs the Sibyl not to write but speak her oracle.’
 Dardana in prose would be constructed with ‘Paridis’ rather than with ‘tela:’ but it is in any case emphatic, as its position shows. Achilles, the greatest enemy of Troy, had been destroyed by Apollo, and not only this, but destroyed through the instrumentality of a Trojan. The joint agency of Apollo and Paris in the death of Achilles was part of the Homeric tradition, Il. 22. 359, other stories making Paris the sole agent (Dict. M. ‘Achilles’). In Ov. M. 12. 580 foll. Apollo, at the instance of Neptune, appears to Paris, encourages him to shoot at Achilles rather than at meaner foes, and guides his aim. ‘Direxti:’ see on 5. 786. With ‘tela manusque,’ which may be called a species of hendiadys, the notion being a single one, the hand fixing the arrow or the arrow fixed by the hand, comp. Aesch. Ag. 111, ξὺν δορὶ καὶ χερὶ πράκτορι.
 Tot is probably to be explained from the context, and especially from v. 62. ‘So many seas as I have entered, it is time that I should rest.’ ‘Intravi’ implies that the seas were previously unknown to him, “hospita aequora,” as they are called 3. 373. ‘Duce te’ need not mean that Apollo showed the way, but merely that he prompted them to sail till they should reach Italy. Comp. “me duce” 10. 92. ‘Repostas’ 3. 364 note. ‘Penitus’ is only an extension of the same notion, so that the two words = “longe remotas.”
 On a comparison of 3. 496., 5. 629, it may be doubted whether ‘fugientis’ is gen. sing., or, as Wagn. suggests, acc. pl. Perhaps it is more like Virg. to separate the noun from its epithet. ‘fugientes’ is said to be the reading of eight MSS. examined by Burm. ‘Prendimus’ may be either present or perf., but the former seems rhetorically preferable. The word is meant to be graphic, expressing a physical grasp of a thing which had nearly slipped away. Comp. 12. 775, “teloque sequi, quem prendere cursu Non poterat.” Wagn. (ed. mi.) seems right in exchanging the period usually placed after ‘oras’ for a semicolon, so as to make v. 62 a kind of apodosis. See on v. 59.
 Hac separated from ‘tenus,’ as in 5. 603. ‘Troiana fortuna’ is said bitterly, ‘Troy's usual fortune.’ Gossrau comp. Hor. 3 Od. 3. 61, “Troiae renascens alite lugubri Fortuna tristi clade iterabitur.” ‘Fuerit,’ the perf. subj. used as a past opt. or imperative. ‘Let ill-fortune have followed us up to this point, but let her do so no longer.’ The use is not quite the same as that of ‘fuit’ 2. 325, as here the force of the past is partially given by ‘hactenus.’
 Wagn. would write ‘Pergamiae:’ see on 3. 133. A few MSS. have ‘parcite,’ which Wakef. adopts.
 “Dique deaeque omnes” G. 1. 21. ‘Obstare’ is here used of that which creates dislike, without any reference to active opposition. So Sil. 17. 551 (quoted by Forb.), “tantumne obstat mea gloria divis?” an obvious imitation of Virg., Pers. 5. 163, “an siccis dedecus obstem Cognatis?” “Ilium et ingens Gloria Teucrorum” 2. 325.
 Praescius with gen. is found also in Val. Flaccus and Tac. (see Forc.), on the analogy of ‘conscius,’ ‘inscius,’ ‘nescius,’ &c. ‘Da:’ for the sense see 3. 85 (note), for the construction 5. 689. Some editions make the parenthesis to end with ‘posco,’ which Heyne rightly rejects.
 It is extremely difficult to say whether ‘fatis’ is the dat., as Burm. thinks, or the abl., as Peerlkamp and Forb. contend. Either expression would be Virgilian (comp. 7. 120, “fatis mihi debita tellus,” with 11. 759 “fatis debitus Arruns”), and either would yield an appropriate sense, as the fates may be represented either as satisfying the requirements of others, or as having their own requirements satisfied (comp. the passages where the fates are said ‘poscere,’ 4. 614 &c.). Where the fates are identified with an individual, as here by the possessive pronoun ‘meis,’ they assume as it were a subordinate position (comp. 7. 293, “fatis contraria nostris Fata Phrygum”), and so may be regarded not as causing events, but as demanding their fulfilment from some other power. The question then is whether the Sibyl is here regarded as the person through whom a demand is made on destiny, or on whom the destinies of private persons make their demand. On the whole I think it must be left open, as there seems nothing in the context, in the nature of the case, or in parallel passages to incline the scale either way, though Val. F. 5. 508 (quoted by Forb.), “Non aliena peto terrisve indebita nostris,” looks as if that author understood ‘fatis’ as dative. ‘Considere’ 4. 349, where as here the names of Italy and the Trojans are contrasted by way of emphasis. Rom. has ‘consistere.’
 “Ut solet, miscet historiam. Nam hoc templum in Palatio ab Augusto factum est: sed quia Augustus cohaeret Iulio, qui ab Aenea ducebat originem, vult ergo Augustum parentum vota solvisse.” Serv. The temple was built in honour of Apollo (Suet. Oct. 29), but it appears from the description in Prop. 3. 23. 15 that the statue of the god stood between statues of Latona and Diana. ‘Templum’ was restored by Heins. from Med. and Rom. for ‘templa’ (Pal., Gud. &c.). Henry prefers the latter, but in the parallel instances he quotes the plural is put for the sing. for the metre, which could not be pleaded here: and the change seems due to some copyist who supposed two temples to be intended. “Templum de marmore” 4. 457, G. 3. 13.
 Instituam is connected with ‘templum’ and ‘dies’ by a kind of zeugma, not unlike “moresque viris et moenia ponet” 1. 264. ‘Instituere aras’ occurs Val. F. 3. 426. Rom. has ‘constituam,’ which would suit ‘templum,’ but not ‘dies.’ The ‘festi dies’ are the ludi Apollinares instituted B. C. 212 (Liv. 25. 12).
 It might appear at first sight as if Aeneas were promising the Sibyl a temple: but the reference is doubtless to the honours paid by the Romans to the Sibylline books, which were first placed in the Capitol, and afterwards deposited by Augustus under the base of the statue of his Palatine Apollo. The latter is of course especially alluded to. In Ov. M. 14. 128, to which Heyne refers, Aeneas promises the Sibyl a temple in so many words; but she expressly declines the offer, as not being a goddess. ‘Penetralia’ may possibly point to the secrecy of the place where the books were laid up: but it is often used rather vaguely, and in Sil. 13. 62 it seems to stand for a movable shrine, if not for the statue of a deity. ‘Manere’ of a thing in the future 7. 319 &c.
 Hic, i. e. ‘regnis nostris.’ ‘Tuas sortes arcanaque fata’ refers of course to the Sibylline books, which were entrusted to the charge of ‘lecti viri,’ at first two, then ten, afterwards fifteen or more. ‘Sortes’ of oracles 4. 346.
 Dicta meae genti: the oracles had not as yet been uttered but are conceived of as uttered at the time to which Aeneas looks forward, so that it is in fact an invitation to the Sibyl to utter them. Not unlike is “pugnata in ordine bella” 8. 629. ‘Ponam’ is used much as in 1. 264, of setting up permanently.
 Alma is specially applied to goddesses, 1. 618., 10. 215, 220 &c., a sort of equivalent to the Greek πότνια, and so is applied as a complimentary appellation to the Sibyl here and v. 117. ‘Tantum,’ as Forb. remarks, is frequently used in adjurations, as in 8. 78. The request here made formed part of the advice of Helenus, 3. 456. “Foliis mandat” 3. 444.
 Comp. 3. 448 foll.
 3. 457. “Pausam facit ore loquendi” is quoted from Lucilius by Non. v. ‘pausa.’ Enn. A. fr. inc. 108 has “pausam facere fremendi,” where Vahlen not improbably reads “facere ore,” Columna “fecere.” ‘Ore’ with ‘loquendi,’ as in 1. 614 &c.; it might however go with ‘finem dedit.’ Some critics have thought the hemistich spurious: but there is nothing un-Virgilian about it, and it is apparently found in all the MSS.
[77-97] ‘The Sibyl still struggles with the god: at last the doors fly open, and she finds voice. She tells him that his perils on land will be as great as those on sea; that another Iliad is opening; but that he must not despair, as deliverance will dawn from an unlooked-for quarter.’