The preceding picture resembles one in Apoll. R. 4. 930 foll. (referred to on v. 241 above), where Thetis and the Nereids push the Argo through the Planctae. There it is apparently meant that the powers of the sea were visible: here it would be needless to suppose it to be meant, any more than in v. 241. Aeneas sees the extraordinary calm, and his anxiety, of which we are not told expressly, though we may infer it from the cares which preceded, vv. 700, 720, as from Venus' own, is followed by joy.
 Seeing the winds favourable, he orders the masts to be set up and the sails spread. Some copyists, not seeing the sense, wrote ‘remis’ for ‘velis,’ as if ‘bracchia’ meant the arms of the rowers, as in v. 136 above; and ‘remis’ is actually found in both Rom. and Med., though Pal. and the majority of the MSS. have ‘velis.’ ‘Bracchia’ however are the sail-yards, “veluti bracchia mali,” as Forc. says—a metaphor perhaps invented by Virg., and followed by Val. Fl. 1. 126, “Pallada velifero quaerentem bracchia malo,” of the building of the Argo. ‘Velis’ then will be the abl., the meaning being that the sails are stretched on the yards, which Virg. has chosen to express by saying that the yards are stretched with or in respect of the sails. Comp. 4. 506 note.
 The description is somewhat minute, perhaps in imitation of such passages as Il. 1. 433 foll. The important words are ‘una,’ ‘pariter,’ and ‘una,’ the rest being merely a description of sailing with a more or less shifting wind. ‘Pedes’ or πόδες were the ropes attached to the two lower corners of a square sail (Dict. A. ‘Ships’). The word is as old as Hom., occurring Od. 5. 260., 10. 32. These are fastened to the sides of the vessel, towards the stern, an operation briefly expressed by ‘fecere,’ which follows the analogy of “facere vela.” The wind keeps shifting, so the sails are spread (“solvere vela” 4. 574, opp. to ‘legere’) first left, then right, to catch it, and this is done ‘pariter’ (like ‘una’) by all the vessels at the same time. The omission of ‘nunc’ before ‘sinistros’ is to be noted. Forc. says it occurs sometimes, but gives no other instance of it.
 Cornua, the extremities of the ‘antennae’ (3. 549 note), are turned this way and that, ‘torquent detorquentque,’ as the sail is shifted. ‘Sua flamina’ like “ventis iturus non suis” Hor. Epod. 9. 30, showing that what is said of the shifting of the wind above is not intended to be more than may happen in the most favourable voyage.
 Ad hunc, after or according to him, a use of the preposition largely illustrated by Hand. Turs. 1, pp. 107 foll. The accusative generally expresses, what is here implied, the rule or law that is followed, as “ad voluntatem,” “ad arbitrium,” “ad nutum,” “ad numerum.”
 “Mediam metam” is a metaphor from the δίαυλος, where the race is round the goal, which accordingly marks that half the course is over. We may then comp. Ov. M. 3. 145, who says, speaking of midday, “Et sol ex aequo meta distabat utraque,” though the race he contemplates is a different one, from one point to another, each of which he calls ‘meta.’ But it is possible that Virg. may have an entirely different meaning, considering the arch of the sky as a ‘meta’ or cone, of which the topmost point is reached at midnight. This is evidently Serv.'s meaning when he says, “Perite locutus est: nam medium caelum meta est ἀναβιβάζοντος circuli, qui medius est inter ortum et occasum.” Such an interpretation is strongly confirmed by Cic. Div. 2. 6, who, speaking of an eclipse of the moon, says “quando illa e regione solis facta incurrat in umbram terrae, quae cst meta noctis,” words, as Forc. says, punctually commented on by Pliny 2. 10, “neque aliud esse noctem quam terrae umbram, figuram autem umbrae similem metae ac turbini inverso.” Heyne apparently confuses or combines the two explanations.
 With Jahn, Ladewig, and Ribbeck I have restored ‘laxabant,’ the reading of the earliest editions, and, as now appears, of all the best MSS., Med., Pal., Rom., Gud. &c., for ‘laxarant.’ The question between them is about as important as that between ‘conplebant’ and ‘conplerant’ v. 107 above: either might well stand, ‘laxarant’ being supported by ‘laxaverat’ v. 857, where the act is regarded as completed, ‘laxabant’ by ‘laxabant’ 9. 225, where it is regarded as continuing.
 The meaning seems to be that they slept on the benches beside their oars. ‘Dura’ is a touch of late civilization which we should scarcely have found in Hom.
 Heyne preferred ‘tristia somnia,’ a reading which seems to be very slenderly supported. The distinction attempted by Wagn., as if ‘somnia tristia’ meant ‘dreams, and those sad ones,’ ‘tristia somnia,’ ‘sad things, namely dreams,’ is surely overstrained. ‘Somnia’ we may say with Forc. is put for ‘somnos:’ in other words the poet talks of dreams when he means no more than sleep.
 ‘Insonti,’ as he did not yield to sleep deliberately, but was overcome by drowsiness against his will.
 Phorbas may be the same as the father of an Ilioneus killed by Peneleos Il. 14. 489 foll.: but all that we can say is that Virg. borrowed the name for one of Palinurus' comrades, who, from the speech he makes, may be reasonably supposed, as Gossrau observes, to have been acquainted with steering. ‘Fudit’ was read before Pier.
 Aequatae, not shifting, but taking the ship exactly in the stern (comp. v. 777), and filling the sails evenly. Comp. 4. 587 note. ‘Datur hora quieti’ is not explained by the commentators: yet it is susceptible of several meanings: (1) ‘the hour is given (you) for rest:’ (2) ‘the hour is sacred to rest:’ (3) ‘the hour is being given (by others) to rest,’ i. e. every one is asleep. On the whole the second seems preferable, though I know of no parallel expression in Virg. or elsewhere which might place it beyond doubt.
 Ponere caput 11. 830, Hor. 2 S. 8. 58. ‘Furare,’ as Heyne remarks, is used like the Greek κλέπτειν, though no more is meant than withdrawing, “subtrahere,” much as we in a different connexion might talk of stealing a nap. The construction with the dative is one of those facts which seem to point to a connexion between the dat. and the abl. See on E. 7. 47.
 Inire seems to contain the notion of entering upon, as in ‘inire magistratum.’ So “inire inperia” is said by Stat. Ach. 1. 280, not, as Forc. says, in the sense of ‘subire et iis parere,’ but with a special reference to a horse being only just submitted to the process of breaking in. Virg. probably avoided ‘obibo’ from his usual love of variety, wishing his readers to be reminded of the one compound by the other, while choosing a word which has a meaning of its own.
 “‘Vix attollens lumina’ aut a sideribus removens, aut certe numinis praesentia praegravatus, quod est melius,” Serv. Heyne agrees with this preference of the latter interpretation, but Wagn. and Forb. are surely right in adopting the former, which agrees with v. 853. Strictly speaking Palinurus would have to turn rather than raise his eyes in order to look at the pretended Phorbas; but the attitude of looking down is so natural to those engaged in work, that we easily understand how Virg. came to speak of looking up.
 Salis of the sea 1. 35 &c.
 Palinurus asks in effect, ‘Do you bid me, who know so well the real nature of this quiet sea, to act as if I did not know it?’ ‘Monstrum’ is apparently used of the sea to express its strange and noxious qualities, much as we should use ‘monster.’ We may comp. its use of the Trojan horse, 2. 245, of Polyphemus, 3. 658, of Cacus, 8. 198, as well as note on G. 1. 185.
 This and the next line present considerable difficulty, as the structure of v. 850 seems to show that ‘auris’ is the dative after ‘credam,’ while that of v. 851 pleads for coupling it with ‘caeli fraude sereni.’ Serv. appears to have read ‘caelo,’ the reading of some of Pierius' copies, and originally of Pal., and so Ribbeck; but though this would make it easy to take ‘auris’ as a dat., it would introduce clumsiness and obvious tautology. Donatus also read ‘caelo,’ reading too ‘fallacius’ for ‘fallacibus,’ and so making ‘quid—caelo’ parenthetical. A further change, also sanctioned by some MSS. (e. g. Gud. a m. p.), would be to read ‘caelo sereno;’ but ‘fraude’ would then be an awkward and superfluous adjunct of ‘deceptus.’ The proposal, revived by Bothe, to take ‘quid enim’ parenthetically, supplying ‘monstro’ to ‘credam,’ and leaving ‘auris’ to go with ‘fraude,’ had already been rejected with reason by Heyne as contrary to the sense of ‘quid enim.’ Accepting the ordinary pointing as the only natural one, we cannot separate ‘credam’ from ‘auris,’ as Jahn still wishes to do; while on the other hand to understand ‘et deceptus,’ ‘and that after having been deceived,’ with Heyne, Wagn., &c., seems scarcely natural. I would then regard it as one of the instances where Virg. has coupled by a copula two forms of expression not grammatically co-ordinate (see on 3. 329), ‘fallacibus auris’ being equivalent to ‘falsus auris,’ ‘deceptus caeli fraude’ to ‘fraudi caeli quae decepit me.’ As such it is rightly included by Wagn. in his Q. V. 34. 2, though with Heyne he gives to ‘et’ the sense of ‘et quidem.’ In these cases Virg. generally contents himself with coupling two words, such as an adverb and an adjective: here he goes farther, so that we might almost class it with instances of the confusion of two constructions, were it not that here the two constructions are completed before they are forced into co-ordination. ‘Auris’ was restored by Wagn. from Med. and Rom. for ‘austris,’ which is found in Pal. from a correction and in Gud., and is supported by Donatus and the Dresden Serv.