than has been generally recognized (cf. White, _op. cit., passim_). In my opinion this sort of thing was even more common at an earlier period, and in this way it was possible for the comus to have a quasi-agon from which the later histrionic agon could easily develop. Of course, the chorus leaders could not appear in individualized rôles, as the actors did in the Aristophanic agon, for characters had not yet been introduced into comedy; but they could engage in a contest of perfectly general, depersonalized billingsgate or, at a later period, speak as the poet’s mouthpiece for the pros or cons of any question. Thus, they would not represent individual men, with an individual’s name and characterization, but _any_ men. Their sentiments would have been equally appropriate in the mouths of any of the other choreutae. The agon and parabasis must necessarily have been flanked on either side by a processional and a recessional. In their simplest form, these need not have involved more than silent marching in and out again; but probably the flute accompaniment was always present, and singing would soon be added. Even when words and singing were employed, there was no necessity of these being newly composed for each occasion or even original at all. It will be remembered that in Aristophanes’ earliest and latest plays he did not write special exodia but borrowed from earlier poets any popular airs that suited his purpose.[99] Moreover, Aristophanes’ exodi lack the balanced structure which is characteristic of all divisions which descend directly from the primitive comus; but in this instance that fact has no significance, for the reason that by the end of a comedy (or comus) the two half-choruses would always be reconciled and go marching off together. Nevertheless, the intrusion of the histrionic element, the comparative rarity of the earliest dramatic meter (the trochaic tetrameter), and the absence of a canonical structure make it plain that the recessional of the primitive comus never developed into a regular division—in other words, that the exodus of Aristophanic comedy was the product of a later period. On the other hand, the Aristophanic parodus resembles the agon and the parabasis in making a large use of the tetrameter (_op. cit._, p. 185). Moreover, it contains distinct survivals of epirrhematic composition (_ibid._, pp. 159 and 366), so that, in spite of its histrionic elements and the absence of a canonical form, the parodus ought to be considered as having been exclusively choral by origin and as having developed out of the simple processional before the comus became histrionic.