^{page 99 note 1} Hauler, E. in Wiener Studien, 1917, vol. 39, 122–134Google Scholar. He proposes the line Potius quam dicta confingit <mimis suis>. Another of his lines begins with Nusquam pugnant. But is it likely that Laberius would open a senarius with two sugspondee words? Plautus allows it only once, and then it is in an iambic octonarius in a canticum (Trin. 285).

^{page 99 note 2} Professor Lindsay was kind enough to look through the present article, and I have incorporated in it some of his criticisms and suggestions.

^{page 100 note 1} Hingst does not notice that ∪,–∪–, e.g. in otio does not seem to be accepted as a proper equivalent in the best writers. It is not admitted by Phaedrus. The only instance which he quotes (p. 31) is from the so-called Caecilius Balbus.

^{page 100 note 2} Lindsay would find here a contrast between the prepositions in and ab, and compares Lord Neave's line on the Permissive Bill, ‘To permit me to prevent you From having your glass of beer.’ This seems to me now the right explanation, and in the parallel passage Pers. 478 I should be inclined to accept the reading ‘ne quis mi (for mihi) in iure abiurassit,’ so as to secure a similar contrast in this trochaic tetrameter.

^{page 101 note 1} The end of the second metron seems to have been a peculiarly sensitive part of the senarius. This is shown among other indications by the admission of a syllaba anceps at this point, e.g. Truc. 425. It is clear that some sort of pause could be made here if it were deemed necessary.

^{page 101 note 2} Hingst has not noticed this instance.

^{page 104 note 1} These choriamb-words obey the same rule as the molossus-word in this part of the verse. They cannot be followed by an anapaest- or spondee-word unless these have a heavy break before them, or an enclitic following them, e.g. Bacch. 246 Mnesilochus? uiuit ualet. Pers. 706 multimodis scriptumst tuom. Hence when the MSS. give dimidium iussit dari in Aul. 291 modern editors prefer to alter to iussit dimidium with Aulus Gellius, so as to secure the correct rhythm and also the alliteration. In Rud. 1264 the MSS., on the contrary, unite in giving cenam continuo coqui and not continuo cenam. The only passages which do not obey this rule are Stich. 439 and a group of instances (Bacch. 144; Ibid. 348; Curc. 271; Most. 58; Rud. 27; Trin. 583), where in each case the choriamb-word is a compound of uenio. Leo would restore forms in -uenat, -uenant. But even if we do not adopt this course, none of these instances causes difficulty. In all of them accent and ictus in the last metron can be made to agree owing either to enclisis, elision, or close connexion. Why, then, in Stich. 439 (a passage which Hingst omits) do editors still tolerate the line ad Sangarinum cenam coqui when it is quite easy to read Sagarinum, which form indeed is definitely required in line 644?

^{page 104 note 2} That the emphasis for which I am contending does not depend solely on heavy words, such as molossi and choriambs, but is also due to the position, is shewn by Stich. 361–2:

credo hercle adueniens nomen mutabit mihi facietque extemplo Crucisalum me ex Chrysalo,

where a similar emphasis for a pun is obtained by a fourth paeon-word.