^{page 1 note 1} It is a conflation of papers read to the Oxford branch of the Classical Association in June 1958 and to die Cambridge Philological Society in February 1960. I am grateful for useful advice on both occasions: more particularly, to M. I. Finley and P. A. Brunt. I have not attempted to deal with the large literature of the subject except where it is directly relevant to my own argument. Herter, H.'s careful essay ‘Pylos und Melos’ in Rh. Mus. XCVII (1954), 316–43Google Scholar refers to everything recent; cf. also Wasserman, F. M.'s survey in Cl. Weekly, L (1956/1957), 93–4Google Scholar.

^{page 2 note 1} In III, 91, 1 ἐς Μῆλον was enough (though in 91, 2, as in v, 84, 2, their position as islanders is referred to in explanation).

^{page 2 note 2} Treu, M. in Historia, II (1953/1954), 253–73Google Scholar and in (1954/5), 58–9 argued that Melos was tributary from 425 onwards but rebelled in 416. Eberhardt, W. in Historia, VIII (1959), 284–314Google Scholar deals adequately with this, and with the general hostility between Athens and Melos.

^{page 4 note 1} Further, the argument about Syracuse given to Nikias in VI, 11, 3, ἐκείνως δ’ οὐκ εἰκὸς ἀρχὴν ἐπὶ ἀρχὴν στρατεῦσαι, was not based on subsequent experience.

^{page 4 note 2} Momigliano, , Mem. Acc. Sc. di Torino, II, 67 (1930), 11Google Scholar; De Sanctis, , Rend. dei Lincei, VI, 6 (1930), 299 ffGoogle Scholar. ( = Studi di storia della storiografia greca, pp. 73 ff.), does not really get to grips with Momigliano's point: Isokrates 8. 78, 105 does not amount to a narrative alternative to Xenophon's, and in any case a publicist with a case to argue about the Social War is not to be trusted that far. de Romilly, Mme, Th. et l'impérialisme athénien, pp. 232–3Google Scholar argues that the subjects of Athens are included in Xenophon, 's πολλοὶ δὲ καὶ ἄλλοι τῶν ‛Ελλήνων (Hell, II, 2, 19)Google Scholar, but even so they are not given the lead. It is relevant that Xenophon must have read the Dialogue.

^{page 4 note 3} Beloch, , Gr. Gesch. II ², 2, 14Google Scholar, and Momigliano suppose that the Dialogue was written immediately after the event. For reasons given in the text I find this unlikely.

^{page 7 note 1} Similarly, the one direct reference to the problem of empire in the Funeral Speech, II, 41, 3 οὔτε τῷ ὑπηκόῳ κατάμεμΨιν ὡς οὐχ ὑπ’ ἀξίων ἄρχεται, might be more highly appreciated by the rulers than by the ruled.

^{page 8 note 1} Meyer, , Forschungen, II, 391–4Google Scholar; Schwartz, , Geschichtswerk des Th., pp. 147–85Google Scholar de Romilly, pp. 131–2; Gomme, , J.H.S. LXXI (1951), 75–81Google Scholar (in his n. 15 and de Romilly, p. 124, the controversy about Kimon is irrelevant: Thucydides, who was not an empire-builder, was Perikles' adversary at the time in question); Wade-Gery, , J.H.S. LII (1932), 215, cf. p. 221Google Scholar ( = Essays in Greek History, pp. 252–3, cf. p. 260).

^{page 8 note 2} This is, I think, implied in the strong praise given to Thucydides in Plato, , Meno 94 dGoogle Scholar, and Aristotle, , Ἀθ. Π 28, 5Google Scholar.

^{page 9 note 1} Propounded by both Meyer and Schwartz (see n. 1, p. 8).