^{page 44 note 1} Professor Platt.

^{page 44 note 2} In Ψ 223 possible.

^{page 44 note 3} Koch, cited by Munro and by Platt.Google Scholar

^{page 45 note 1} The Clarendon Press edition (rccensuit D. B. Monro, 1901) is used throughoutGoogle Scholar

^{page 47 note 1} The point has been discussed (Classical Quarterly, April, 1908, pp. 94 sqq.).Google Scholar

^{page 48 note 1} It is essential that the causes assigned should be possible in Sanskrit too.

^{page 49 note 1} Of course, not historic present.

^{page 49 note 2} became disliked.

^{page 49 note 3} 3 cf. A 735, which is clearly past.

^{page 52 note 1} The figures include only those instances which occur in the introductions or resumptions of speeches.

^{page 52 note 2} The accent is neglected, as it was , etc.

^{page 53 note 1} Really a speech-augment (B 402); the natural ictus is . II 182 is Odyssean, but perhaps not quite so difficult as , because of the Π

^{page 53 note 2} The line shows five rarities, and the scansion is unexampled, even in the speeches.

^{page 55 note 1} The treatment of the imperfect is less conservative.

^{page 56 note 1} , etc., are here omitted, since ενis hardly a possible scansion. Type must be compared with , etc. An admissive augdactyl is the most convenient scansion of all; and type a little outnumbers type but not type +type .

^{page 56 note 2} In the Iliad without B² etc. (narrative) there are forty-six instances of type , eighteen instances of the corresponding aug mented aorists (, etc.).

^{page 58 note 1} If such a phrase applies to the genially atasthalous K, who is also responsible for the scansion

^{page 58 note 1} For brevity I use ‘admissive’ to mean beginning with a vowel.

^{page 58 note 2} By ‘Adonian ’ I mean any kind of; the special scansion will be called an Adoniac.