This whole description is closely modelled on Od. 5. 43 foll., part of which coincides with Il. 24. 339 foll.
 See on v. 223. Hom. has ἅμα πνοίῃς ἀνέμοιο, which seems to mean ‘together with the wind,’ not, as some have taken it, and as Taubm. and apparently Gossrau take Virg.'s words, with speed like the wind's.
 For the nature of the caduceus see Dict. Biog. Hermes. ‘Hac—resignat’ is a parenthetical sentence, as Jahn has seen, answering the purpose of a relative clause, to express the ordinary functions of the wand. ‘Evocat’ then will mean “evocare solet,” while ‘agit’ v. 245 on the contrary refers simply to what Mercury is doing during his present journey to Carthage. Hom. merely dwells on the power of the wand to produce or dispel sleep, τῇ τ᾽ ἀνδρῶν ὄμματα θέλγει Ὧν ἐθέλει, τοὺς δ᾽ αὖτε καὶ ὑπνώοντας ἐγείρει. Virg., while including this notion in ‘dat somnos adimitque,’ extends it to breaking or producing the sleep of death, Hermes being the ψυχοπομπός.
 Hermes has his wand when he conducts the souls of the suitors to the shades at the beginning of Od. 24. ‘Tartara tristia’ Med., Pal., &c.: ‘tristia Tartara’ was the order before Burm. For ‘mittit’ Pal. originally had ‘ducit.’
 I follow Henry in accepting Turnebus' explanation (Advers. 24. 26) of ‘lumina morte resignat’ by a reference to the Roman custom of closing a friend's eyes at the moment of death and afterwards opening them again when the body was laid on the pile seven days afterwards. ‘Signare’ (Stat. Theb. 3. 129) and ‘sigillare’ (Varro ap. Non. 2. 785) were used of closing the eyes of the dead, so that ‘resignare’ would naturally express the reverse process. It does not appear that this action is clsewhere attributed to Mercury; but it would be a natural part of the Roman conception of a ψυχοπομπός, the object being, as Henry suggests, that the dead might see their way to the lower world. We may add this then to the other instances in which Virg. has mixed the customs of his own times with those of the heroic age. ‘Morte’ then will mean during, i. e. after, death, and the words will = “lumina mortuorum resignat.”
 In Hom. Hermes comes down from Olympus upon Pieria, and thence throws himself on the sea. Virg. knows nothing of the local Olympus, but, wishing as usual to follow Hom., makes Mercury take Mount Atlas as a halting-place between the sky and Carthage. The belief that Atlas stood as a pillar between heaven and earth doubtless seemed a reason for his introduction here, though perhaps it rather confuses the image. Voss (Mythologische Briefe, vol. 2, letter 27, referred to by Henry) suggests a more elaborate explanation, according to which there were three openings in the heavenly Olympus, one in the vertex, immediately over the earthly mountain, the other two at the sides, eastern and western, the latter being the route taken by Mercury as the nearest to Carthage. I do not know whether it is an objection to this view that when in 1. 225 Jupiter looks down upon Libya, he is expressly said to be standing not at the west gate of Olympus but at the vertex: but in any case it seems rash to suppose that the ancient poets generally were agreed in their notions of Olympian topography.
 Duri is rightly explained by Serv. “laboriosi,” recalling us in fact to the sufferings of the old Titan, which Aesch., it will be remembered, regards as parallel to those of Prometheus (Prom. 347 foll.).
 It has been questioned whether there are pines in Africa. In any case ‘pinifer’ is a natural epithet of a mountain (E. 10. 14 &c.). The identification in detail of the mountain and the Titan perhaps seems a little ungraceful. Sides, head, and shoulders are natural enough; but the chin and the beard strike a modern reader as grotesque. Henry however rightly reminds us that Virg. is not personifying the mountain, but describing one who having been a demigod had become a mountain by transformation, so there is some excuse for pursuing the resemblance minutely.
 ‘Poising himself on even wings’ is generally understood to be an expression for ‘pausing in his flight,’ as the wings would then appear more level than when the flier was at full speed. So in 5. 657., 9. 14, “paribus se sustulit alis” is explained as marking the beginning of flight. But it may be questioned whether the words mean more than ‘equal,’ ‘well-balanced wings,’ so as to be applicable to any moment in Mercury's flight.
 Hom. goes on, σεύατ᾽ ἔπειτ᾽ ἐπὶ κῦμα. Virg. leaves this to be supplied, partly from the description of the bird, ‘humilis volat aequora iuxta,’ partly from v. 256. Virg. does not specify the bird, but Hom. has λάρῳ ὄρνιθι ἐοικώς.
 This and the two following lines were condemned by Heyne and Bryant, and are rejected by most of the later editors. The external grounds for suspecting them are, that in one of the Berne MSS. they are written in later ink, that some MSS. place v. 258 before v. 257, and that most copies give either ‘ac Libyae’ (Med., &c.) or ‘ad Libyae’ (Pal. a m. s., Gud. a m. pr., &c.). See note at the end of this book. The chief internal ground is the awkwardness of expression in v. 257, as it would be equally objectionable to remove the point after ‘volabat,’ so as to connect ‘volabat litus Libyae,’ to make ‘litus ventosque secabat’ a zeugma for “litus radebat ventosque secabat,” and to adopt the reading ‘ad Libyae,’ which the instances quoted by Weichert do not sufficiently support. But I believe the difficulty will vanish if we understand ‘litus ventosque secabat,’ ‘he was dividing the shore from the winds,’ i. e. he was flying close to the shore, so as to be, as it were, between the winds and the land—a repetition in more specific and defined language of ‘terras inter caelumque volabat.’ So v. 695, “Quae luctantem animam nexosque resolveret artus” = “quae animam ab artubus resolveret.” The jingle ‘volabat— secabat,’ whether graceful or not, is not unexampled in Virg. Nor can any argument unfavourable to the genuineness of the lines be founded on the words ‘Materno veniens ab avo.’ The poet may seem somewhat late in remembering Mercury's relationship to Atlas; but it is not unlike the indirect manner in which he sometimes introduces what he has to say. The fact of the relationship he was likely enough to mention, as he has done afterwards, 8. 138 foll., where again Atlas is spoken of as the upholder of the heavens. This note must not conclude without a mention of Bentley's proposal to read ‘legebat’ for ‘volabat,’ a substitution which, if we might deal with Virg.'s verses as a tutor with a pupil's exercise, might be accepted as an improvement, but has no probability on critical grounds.
[259-278] ‘Mercury alights, and finds Aeneas in gay attire superintending the buildings of Carthage. He remonstrates with him, deliver's Jupiter's message, and vanishes.’
 Aeneas is taking part in the erection of buildings public and private, which we have had described more at length 1. 423 foll. ‘Novantem’ apparently with reference to the huts which these more splendid edifices were to displace.
 It does not seem necessary to follow Wagn. in beginning a new sentence with ‘atque,’ as if that particle were meant to call attention to an unexpected novelty. It has rather the usual continuative force, implying that Aeneas' Tyrian dress was quite in keeping with the work he had undertaken as Dido's architect. ‘Stellatus iaspide,’ shining with jaspers as with stars, either on the hilt or on the scabbard. Comp. Juv. 5. 42 foll., where the present passage is alluded to. The sing. seems to have a plural force, as in Juv. 5. 38, “inaequalis beryllo phialas.” Wagn. rightly remarks that ‘iaspide’ is quadrisyllable.
 Ardebat with ‘murice.’ The ‘laena’ seems here to be a mark of luxury, as in Pers. 1. 30; but it appears from Juv. 5. 131., 7. 73 that it was also worn by the poor. See Dict. A. s. v. The stress then is here probably to be laid on the colour as indicating the costliness.
 Comp. 11. 72 foll., where v. 264 is repeated. ‘Quae munera,’ as the context shows, can only refer to the ‘laena;’ but Virg. was probably led to use the plural by thinking at the same time of the sword, which was doubtless Dido's present also.
 Invadit, attacks him, like “adgreditur,” v. 92, but stronger. Forb. comp. Tac. A. 6. 4, “Agrippa consules anni prioris invasit,” where a speech in the oratio obliqua follows. Wagn. thinks ‘altae’ inappropriate in the mouth of Mercury, but it has its force here, as well as in v. 97 (note)—the same which is expressed in the next line by ‘pulchram urbem.’ ‘You are occupying yourself in contributing to the grandeur of a city which is not only not yours, but sure to be one day your bitterest enemy.’
 It signifies little whether we put a note of interrogation after ‘exstruis,’ as most editors have done, or, as Wund. prefers, one of exclamation. In either case ‘oblite’ had better be connected with the preceding sentence, being in effect equivalent to ‘oblitus’ (comp. 2. 283 note), instead of constituting a kind of interjectional sentence by itself. The first reading of Pal. was ‘ignare,’ perhaps, as Ribbeck hints, from 3. 382.
 It is questioned whether ‘torquet’ here refers to physical movement or to government. “Torquet qui sidera mundi” 9. 93 is for the first, “Cuncta tuo qui bella pater sub numine torques” 12. 180 for the other. If we must decide a question which probably did not present itself as sharply to Virg.'s mind as to ours, we shall perhaps do wisely in saying with Wagn. on 9. 93 that the physical sense is the prominent one here in ‘torquet caelum,’ the moral in ‘torquet terras.’ For ‘et terras’ Pal. and Gud. have ‘ac terram.’
 Serv. accounts for the change from the language of v. 235 by saying that Aeneas would not have recognized Carthage as hostile, though Jupiter knew it really to be so. ‘Terere otia’ like “terere tempus” &c. Cerda, remarking on ‘teris—terris,’ thinks that Virg. intended to allude to the etymology of “terra” from “terere.” This is of course absurd; but the jingle can hardly have been unintentional, either here or in such passages as v. 238, “parere parabat,” 10. 191, 192 “cănit—cānentem,” ib. 417, 418 “cănens—cānentia.” See on 2. 494.
 See on 6. 405.
 This line is omitted by Med., Pal., and several other MSS., while in others it bears the mark of having been added afterwards. Pomponius Sabinus has a note on it, which unfortunately is imperfect, merely containing the words “Nec super secundum Apronianum.” On the one hand it may easily have slipped out, especially if the transcriber was writing from his recollection of Jupiter's speech; on the other hand it may as easily have crept in from that speech. There it was almost necessary, as there is no pronoun in v. 232 to fix the reference pointedly to Aeneas: here the emphatic position of ‘te’ in v. 260 renders the addition needless, though it is still graceful. On the whole it seems best to print the line in brackets.
 Spes Iuli not the hopes of the kingdom entertained by Iulus, but the hope of manhood afforded by Iulus. See on 1. 556. Comp. 6. 364, “per spes surgentis Iuli.” ‘Surgere’ is used of a race springing up E. 4. 9. ‘Ascanium’ and ‘Iuli’ form a good instance of a mere poetical repetition, both being simply appellative names, so that according to the ordinary rules of language they ought, as employed here, to stand for two persons.
 Wagn., Q. V. 40, finds an inconsistency between this line and v. 236, where he understands ‘prolem Ausoniam’ of Silvius (see 6. 760 foll.). If it is an inconsistency, it is one not confined to this passage, as it is clear from 1. 267 foll. that Virg. at times regarded Ascanius as the founder of the Alban, and hence of the Roman dynasty. See further on 6. 764.
 Mercury vanishes like Apollo 9. 656 foll., where this and the next line are repeated with the change of a single word. ‘Mortalis visus’ is Virgil's indirect way of telling us that Mercury appeared in a visible though divine form. At the second apparition (v. 556 foll.), which however differs from this as taking place in sleep, the fact is told us directly. Serv. remarks that Mercury had repeated all his message, and accordingly understands ‘medio sermone’ to mean ‘when the dialogue was half over,’ in other words, without giving Aeneas time to reply. Donatus has the same interpretation, which is certainly an ingenious one. Gossrau accepts it. But we should not be justified in fixing this special sense of ‘sermo’ unless we could prove that it was never used where only one person was speaking, which 12. 940 shows not to be the case. ‘Medio sermone’ then will have its ordinary sense, which must be explained by supposing either that Mercury became actually invisible before his speech was ended, or that his speech seemed to end abruptly from the suddenness with which he closed it and vanished. In either case the expression is rhetorical and must not be closely pressed, as all that the poet means is to give the effect of a sudden and transient apparition. So below v. 388 there is no sign of imperfection in Dido's actual speech, but “medium sermonem abrumpit” is intended to mark the abruptness and violence of her manner in closing it.
[279-295] ‘Aeneas is thunderstruck and perplexed. At last he gives orders for instant departure, trusting to find an opportunity of breaking the news to Dido.’