^{page 19 note 1} ‘The Etymology of Greek , Language xxiii (1947), 207 ff.Google Scholar

^{page 20 note 1} It is possible, however, as Professor Dover has pointed out to me, that it was formed earlier, but that no trace of the earlier use survives.

^{page 21 note 1} Also possibly in Aristophanes, Thes. 1197, if we accept the reading (so van Leeuwen). See my article in Classical Greek' in C.Q. N.S. xii (1962), 235 ff.Google Scholar In it I rejected Wharton's proposal to derive from which is exactly the reverse of my own. My objection based on the dialectal difference ( in Lesbian and Ionic: in Attic and Doric) is not really affected by Thes. 1197, since there the speaker is a barbarian; but in any case the argument from silence by itself has little force when we are dealing with a word which occurs rarely anywhere. The remaining objections which I raised will not apply to my opposite derivation.

^{page 21 note 2} Cf. Postgate, J. P., Camb. Phil. Soc. Trans. 1886–93, pp. 55–58.Google Scholar

^{page 21 note 3} Also probably in Hesychius, who cites , though glossing it with : the true gloss, appears attached to the neighbouring

^{page 21 note 4} Professor Dover suggests an alternative explanation of Seiva, that it arose in the nominative with lengthened into d Sciva on rhythmical grounds. This would of course require prior extension of from accusative to nominative.

^{page 22 note 1} A Dictionary of Slang and Unconventional English, s.v.

^{page 22 note 2} Wildhagen, and Héraucourt, , German- English Dictionary, s.v.Google Scholar

^{page 22 note 3} Cf. Schol. ad Lucian, Vit. Auct. 19 (quoted by Starkie, ad Ar. Vesp. 524): But desire for concealment is not the only motive for its use.

^{page 22 note 4} With the masculine and feminine article so far as I have seen, always refers to a person. This is possibly a relic of its origin, for would have referred to a man just as did

^{page 22 note 5} Schwyzer, , Griech. Gramm. i. 612, wrongly quotes as a late form.Google Scholar

^{page 24 note 1} It will be seen that I would in any case much restrict the interjectional use. Some have gone so far as to regard all uses of as ‘outside the construction’ (so Starkie, ad Vesp. 524: cf. Rogers, ad Pax 268), but such a view cannot be maintained.

^{page 24 note 2} Not ‘do you eat such a fish?’, as L.S.J.

^{page 24 note 3} Thus in the division of speakers by Lloyd-Jones (O.C.T.).

^{page 25 note 11} Not so indefinite as ‘somebody’. L.S.J. are in error in translating Arist. Pol. 1262^a3 as ‘mine or some other's’: rather it is ‘he is my son, or this or diat man's‘, and the passage continues There is a contrast with in Dem. 4. 19 Demosthenes does not use in the interjectional sense of marking a new thought, but the usage recurs in later prose, in Lucian (cf. L.S.J.).

^{page 25 note 12} Kühner-Blass (i. 615) refer to use of the word by Plato, but I have been unable to trace this.