ὑπονοήσας—ὑπονοεῖν, like ὑποπτεύειν, sometimes= to view with suspicion, sometimes to conjecture.
ἀποχωρῆσαι . . . περιιδεῖν—the infin. with περιορᾶν occurs only in Herod. and Thuc.
λέγων ταῦτα ἂ καὶ αὐτῷ ἐδόκει—probably this means saying what he personally thought they ought to do This interpretation is strongly supported by οἱ δὲ ξυνεγίγνωσκον μὲν καὶ αὐτοὶ and ἐδόκει ποιητέα εἶναι below, with which καὶ αὐτῷ brings our clause into connexion. (The recent edd. who retain these words offer several suggestions, but all except Classen give alternatives and say that the meaning cannot be determined. Classen says ‘saying that which seemed to him from his own conjecture probable,’ thus agreeing with Haacke, Goller, and Arnold. But Bloomfield rightly says that the Syr. knew that the A. would retreat even before the battle—c. 67 Stall and Herwerden reject the words. I think that the edd. have all mistaken the meaning of ἐδόκει, which refers not to the knowledge of Hermocrates, but to his advice. Valla translates ‘Commemorans haec et alia quae ipsi videbantur,’ which, whatever he read, is certainly not what Thuc. meant.)
ἀποικοδομῆσαι—sc. χρεών.
τὰ στενόπορα—those among the hills west of Syr.
προφθάσαντας—Classen, Bohme, and Widmann prefer διαλαβόντας of B, and it is perhaps right. That διαλαμβάνω does not occur elsewhere in Thuc. does not count, as this part of the narrative contains several rare words and ἀγωνισμός, ἀντιτέχνησις, δυσανασχετῶ, and ἐπιφήμισμα, which are not found again in Thuc. προλαμβάνοντες is explained by the schol. φθάνοντες, ὥστε μὴ καταλαμβάνεσθαι at IV. 33. Possibly Thuc. wrote προδιαλαβόντας.
ξυνεγίγνωσκον—the prep. here has an adverbial force.
ἁσμένους . . . ἀναπεπαυμένους—so Sallust Jug. 53, 5 lacti quierant, Postgate's certain correction of laetique erant.
ἀναπεπαυμένους . . . ἑορτῆς οὔσης—cf. c. 51.1.
Ἡρακλεῖ . . . θυσία—the Syr set great store by the fact that the battle fell on a day sacred to Heracles, whose temple was on the hill close to the point at which the A. double wall touched it. Plut. Nic. 25 says that their μάντεις had reported that Heracles required that they should not begin the action. (For the speculations of Timaeus in after times see Plut. Nic 1.)
οὐ δοκεῖν—the verb of ‘saying’ has to be supplied from ξυνεγίγνωσκον by an idiom common in Gk. Oratio Obliqua. ἂν with ἐθελῆσαι.
πρὸς πόσιν τετράφθαι—the metaphorical meaning of τρέπεσθαι πρὸς is much commoner than the literal.
σφῶν—the gen. with πείθομαι is frequent in Herod. but is not found anywhere else in Attic prose. πείθεσθαι is here synonymous with ὺπακοῦσαι of l. 20; and the gen. is also helped by πάντα.
ἐξελθεῖν—depends on πείθεσθαι, which takes sometimes infin., sometimes ὥστε with infin.
ἐπὶ τούτοις—in conseqnence.
πέμπει—asyndeton after a demonstrative is fairly common.
ἐξ ὅσου = ἐς τοσοῦτον ἐξ ὅσου, the antecedent being omitted as in Plat. Phaedo 78B ὅθεν ἀπελίπομεν ἐπανέλθωμεν, and often with relative adverbs.
διάγγελοι—one would have thought that it was high time for even Nicias to be somewhat sceptical. Moreover he had played a similar trick on the Syr. early in 414 It is strange too that any man who had appeared in the δικαστήρια should not have understood the art of lying. But their sufferings prob. unnerved the A. for the time.
