Virg., as Heyne observes, characterizes this secret postern gate in four ways,—first simply as ‘limen,’ then bringing in the notion of secrecy in ‘caecae fores,’ then in ‘pervius usus’ &c. explaining the object of this second entry, and lastly in ‘postes relicti a tergo’ giving the situation of the door, at the back of the building. ‘Caecae fores’ would be expressed in Greek by Ψευδόθυρον. ‘Pervius usus’ in apposition with ‘fores’ &c., as we might call a gate a thoroughfare.
 Tectorum inter se seems to mean merely that by entering this door you might pass from room to room, as you might by entering the front door, only from a different direction. ‘Relicti’ seems best connected with ‘a tergo,’ as a sort of periphrasis for “postica.” Henry's ‘abandoned in the confusion of combat’ is hardly so good, and his proposal to connect ‘a tergo’ with the whole of what precedes is met by the observation just made, that Virg. brings out the several characteristics of the door piecemeal, while giving what are apparently four synonymes for it.
 Incomitata and ‘trahebat’ are noted by Wund. as contrary to the representation of Homer, who describes Andromache not as carrying Astyanax herself, but as attended by her nurse. Virg. of course may be wrong; but he evidently means the privacy of the postern to account for Andromache's being able to visit the king and queen without pomp or attendance of any sort. ‘Saepius solebat’ E. 1. 21.
 Soceros Priam and Hecuba, both of whom are included under the masculine denomination, as ‘patres’ v. 579 stands for ‘parentes.’ ‘Socrus’ seems originally to have been masculine as well as feminine: see Forcell. ‘Trahebat,’ as in v. 321, as the child would not be able to keep pace with her. As Gossrau remarks, the contrast of the former security of Andromache and her child with the agony of the present struggle is pathetic.
 Evado of mounting a height, 4. 685: see Forcell. Henry rightly observes, that it means strictly to pass through the intermediate space and come out on the other side. Aeneas means that he enters the palace through this postern, and scales the roof.
 Iactabant inrita, “spargebant quasi nil profutura,” Serv. The meaning may be not merely that their darts were unavailing, but that they felt them to be so, and accordingly launched them weakly; but this would perhaps be a refinement.
 In Il. 21. 526 foll. Priam mounts a tower, and sees the havoc made by Achilles. Seneca (Troad. 1072 foll.) combines Virg. and Hom., speaking of a tower where Priam was wont to stand and marshal the battle. ‘Turrim’ is the MSS. reading, supported, so Gell. 13. 19 tells us, by Valerius Probus. Charisius however (p. 25 P) quotes the line with ‘turrem,’ as an instance of Virg.'s usage. ‘In praecipiti stare’ is a phrase found in Juv.147. Here it might mean ‘so high as almost to topple over,’ which is the ordinary interpretation; but as this would create a tautology with what follows, Gossrau and Henry seem right in supposing it to signify that the tower stood not in the middle of the palace, but at the extreme edge of one of its sides, so that it would fall not on but over the roof, as is the case v. 465. “Summis tectis,” not the roof of the palace, but the roof of the tower, ‘tectis’ being a modal ablative, like “arcem attollere tectis” 3. 134 note.
 For ‘Achaia’ I have restored ‘Achaica,’ which is the reading of Med. and Pal., while fragm. Vat. has ‘Achaia.’ The κακέμφατον of which the commentators complain (after Serv. on v. 27) can hardly have been felt by Virg., or he would not have written ‘Dorica castra’ in the passage just referred to; while the form ‘Achaica’ is supported by 5. 623, where there is scarcely any difference of reading.
 Tabulata is doubtless the flooring of the ‘turris,’ as in 12. 672, “flammis inter tabulata volutus Ad caelum undabat vertex turrimque tenebat.” Caesar B. G. 6. 29 speaks of “turris tabulatorum quattuor,” of four stories. ‘Summa’ is probably not to be pressed, as Henry rightly objects that Aeneas and his friends would be likely not to get into the tower and try to dislodge the top of it, but to stand on the roof of the palace and endeavour to overthrow the entire tower. Virg. does not tell us how many ‘tabulata’ there were: he merely says that they applied their leverage to the flooring, as affording a point in which implements might be inserted; and he may very well call this flooring ‘summa’ merely as being above or on the roof of the palace. ‘Labantis iuncturas’ is not a very strict expression, as the joining would not totter itself, though it would make the wall totter. ‘Altis’ is generally taken ‘high;’ but it may equally well mean ‘deep,’ the tower being overthrown from the bottom. ‘Sedibus’ then will be the foundation.
 The change of tense in ‘inpulimus’ of course shows the rapidity of the action. With this use of ‘inpellere’ Wund. comp. 4. 22, “animumque labantem Inpulit,” Forb. Lucan 6. 35, “Exstruitur quod non aries inpellere saevus, Quod non ulla queat violenti machina belli.” ‘Ruinam trahit’ v. 631 and elsewhere. So perhaps “ducet ruinam” Hor. 2 Od. 17. 9, “trahere” and “ducere” giving the notion of height, as elsewhere of length. The early commentators remark on the acceleration of the movement of the verse.
 ‘And the shower of missiles from besiegers and besieged is as heavy as ever.’
[469-485] ‘Pyrrhus stands at the gate, like a snake that has renewed its youth, surrounded by his comrades. He makes a breach in the door, and the interior of the palace is disclosed.’