E.C. Marchant, Commentary on Thucydides Book 1

3. ἅς—when a rel. sentence stands for a substantive, no antecedent being expressed, Thuc. regularly attracts the rel., as in c. 37. 3 δικαστὰς ὧν βλάπτουσι. In 2.61 ἐγκαρτερεῖν ἃ ἔγνωτε is for ἐγκαρτερεῖν ἐκεῖνα (not ἐκείνοις) ἃ ἔ. But when a rel. sentence stands as an adjective, the rel. is not necessarily attracted; cf. e.g. c. 52. 2 αἰχμαλώτων ... οὗς ... εἶχον: 7.1. 2 etc. See n. on p. 87 l. 16.

καταδύσειαν—not sank, but ‘disabled.’ The usual way was to tow them away as prizes.

4. φονεύειν ... ζωγρεῖν — depending on ἐτράποντο as in ἐτράποντο τὰ πράγματα ἐνδιδόναι 2.65. Thus ἐτράποντο has here a double construction. φονεύω is an Ionic word, not found in prose outside Herod. and Thuc.

διεκπλέοντες—the edd. point out that this cannot be the technical διέκπλους—the object of which, indeed, was not to kill, but to disable an enemy's ship (see on c. 49）—but merely alludes to rowing in amongst the enemy's disabled ships.

6. οἱ ἐπὶ τῷ δεξιῷ κέρᾳ—the Ambraciots and Megarians (c. 48. 4) who were defeated and pursued to the mainland (c. 49. 5). It is not possible to see how the conquering Corinthians on the left could have fallen in with any of the flying right wing unless some of those on the right wing had not fled with the rest towards the mainland; and that some were left behind is suggested by ἐπὶ πολὺ ... ἐπεχουσῶν.