^{page 3 note 1} If we are to believe the thes. ling. Lat., there is another in anth. Lat. Ries. 439 1 ‘quid saeuis, Cyparē? domiti modo terga iuuenci’, for Mr Reisch says ‘Cyparus nom. uir. gr.’ But Cypare, as usual, is feminine: the poem is addressed to a woman who has seduced a boy and finds him a backward lover; addressed to a man it would be absurd.

^{page 4 note 1} I accept, for the sake of argument, the assumption that Dodonawas the place which the author named.

^{page 4 note 2} The truth is more fully stated in relation to a particular case at Die neuen Responsionsfreiheitan I p. 3.

^{page 5 note 1} Only Mr Rothstein was likely to adduce IV 344.

^{page 5 note 2} One, not three; for in 220 the sense as well as the form is plural, and in 669, ‘ut ipsa ponderi deprimatur ad spinam cutes’, cutes is nom. sing., a form which Mr Gudeman neglects to register.

^{page 6 note 1} The observation which forbids it, though made so long ago as 1861 by Lucian Mueller de r. m. p. 300, is not yet universally known; and just as Schneidewin had conjectured ‘aut aperi faciem aut tu tunicata laua’ at Mart. III 3 4, so did Friedlaender, afterwards conjecture ‘os hominis, mulsum et me rogat Hippocrates’ ib. IX 94 2Google Scholar, and DrPostgate, ‘miramur, facili ut temperat arte manus’ at Prop. II 1 10Google Scholar. All three conjectures have a second vice, in that they end the first half of the pentameter with words which are not allowed to stand there.

^{page 6 note 2} Lucian Mueller de r. m. ed. 2 p. 390 explains the lengthening as due to the following mute and liquid; and, if Caesarea made sense, it would not be incredible that Martial, who ventures on the lengthening ‘Romanā stringis’ even èν αρσει, V 69 3, should here venture even on the lengthening ‘Caesareā praestitit.’

^{page 7 note 1} But Ouid. trist. III i 78 should be added to the list in thes. ling. Lat. V p. 1652 20 and deleted at 1655 7–9, where its place may be filled by anth. Lat. Ries. 423 3.

^{page 8 note 1} Why not faciem coram? Perhaps to avoid rhyming a declinable with an indeclinable word(see Lachmann on Prop. I 5 20); perhaps merely to obtain the regular equipoise of substantive and adjective; or perhaps this poet thought, though Ovid did not, that a preposition and its case cohere so closely when in contact as to interfere with the division ofthe verse, Their separation is often much wider, as in Tib. I 6 30 ‘contra quis ferat arma deos?’

^{page 8 note 2} At Ouid. art. Ill 216; but he quite mistook the sense and wrote nec for et.

^{page 12 note 1} 17 in Virgil, 12 in Ovid's metamorphoses, 5 in other poets; 22 with a double consonant or a mute and liquid following, 12 without.