^{page no 21 note 1} ‘As I thought. . .’; contrast 22. 2 on actions and events, ‘and not as I thought. . .’.

^{page no 21 note 2} Minutely discussed by Grosskinsky, A., Das Programm des Thukydides (Berlin, 1936).Google Scholar

^{page no 21 note 3} Antiphon began the circulation in writing of forensic speeches; the fragmentary Lysias xxxiv is the earliest (403) symbuleutic speech.

^{page no 22 note 1} What a speaker needs to say is often irrational. Cf.Winnington-Ingram, R. P., BICS xii (1965), 70–82.Google Scholar

^{page no 22 note 2} In vii. 49. 3 τò ξύμπαν είπεῖν serves this purpose; not so τò ξύμπαν . . . γνῶμεν in iv. 63. 2 or тò ξύμπαν in vi. 37. 2.

^{page no 22 note 3} iv. 85. 4 opposes ἒργῳ to γνώμη.

^{page no 23 note 1} Later historians used speeches, too; on the history of the genre see Walbank, F. W., Speeches in Greek Historians (Oxford, 1965)Google Scholar.

^{page no 23 note 2} Cf.Brunt, P. A., REG lxv (1952), 59–96.CrossRefGoogle Scholar

^{page no 23 note 3} Aristophanes Wasps 1186–1205 suggests that there was much narrative and reminiscence at Athenian parties.

^{page no 24 note 1} Cf.Grant, J. R., CQ N.S. XV (1965), 261-6CrossRefGoogle Scholar.

^{page no 26 note 1} I am in sympathy with the ξύμπασα γνώμη of Gomme, A. W., Essays in Greek History and Literature (Oxford, 1937), 156-89Google Scholar.

^{page no 27 note 1} Cf.Andrewes, A., PCPhS N.S. vi (1960), 1–9.Google Scholar