^{page 168 note 1} I should like to express my gratitude to Professor D. S. Robertson and Mr. H. T. Deas for their most valuable help and advice; and to the librarians and photographers of the libraries containing Aristophanes manuscripts for their kind assistance.

^{page 168 note 2} I follow the sigla used by J. W. White in his article ‘The Manuscripts of Aristo phanes’ (Class. Phil. 1906, pp. 1 ff.) which have now become standard.

^{page 168 note 3} The identification was first made by A. von Velsen in 1871 (Übsr den cod. Urbinas der Lysistrata, p. 53.)

^{page 169 note 1} This MS. was formerly designatec G.127, and before that III.D.8.

^{page 169 note 2} E is placed in the 15th cent, by Zuretti, C. O., Analecta Aristophanea, p. 35Google Scholar, anc Puntoni, V., ‘Codices graeci Mutinenses’ (Stud. it. di fil. class, iv (1896)), p. 466.Google Scholar T. W Allen seems to have hesitated between the two opinions (see R. T. Elliott's ed. of Ach.p. xi).

^{page 169 note 3} Cf. the facsimile, ‘Codex Ravennas 137.4.A. phototypice editus: praefatus est J. van Leeuwen’ (Leyden, 1904), also K. Zacher, ‘Die Handschriften und Classen der Aristophanesscholien’ (Jahrb. für class. Phil. suppl. xvi (1888)), pp. 529 ff.

^{page 170 note 1} Attention was concentrated on R's scholia at first because of its age and its superiority in the text. Zacher's Handschr. und Class., however, demonstrated R's inferiority to V as a source for the scholia, which was further emphasized by the criticism deservedly evoked by W. G. Rutherford's attempt in his Scholia Aristophanica to edit R's scholia by themselves (see Zacher, in Philologus suppl. vii, pp. 498ff.Google Scholar; A. Römer, Stud, zu Aristoph…). Rutherford's edition had, however, the merit of providing an almost completely accurate transcription of the scholia in R, the work of H. Graeven.

^{page 170 note 2} I have omitted the which RVE interpolates after as it seems certain hat a new note begins here; this is in fact he first note on the line in R, and ΓΘ have a new lemma for it.

^{page 170 note 3} Cf. schol. Pax 1071, Av. 962.

^{page 170 note 4}

^{page 170 note 5} There is also a scholion at 117 found only in RM Suid.

^{page 170 note 6} So M Suid.

^{page 170 note 7} So Suid.

^{page 171 note 1} So M Suid.

^{page 171 note 2} Some of our examples, and schol. 508 quoted below, suggest that there is a common hyparchetype behind R and Suidas and some readings in M. This is very probable; but consideration of the question had better be postponed until we have examined M and Suid. in detail.

^{page 172 note 1} Except that Coulon ignored Θ and wrongly included M among the descendants of Φ.

^{page 172 note 2} Cf. also schol. 9 (p. 170 note 2).

^{page 173 note 1} Vide Dübn. in adnot. (The scholion in question is omitted by Aid.)

^{page 173 note 2} The difficult and controversial question of how many scribes shared the writing of the whole MS. need not be considered here, as it has no critical significance. Full information is given by Zacher in Philologus, xli (1882), pp. 11 ff. (summarized, with some corrections, in Handschr. und Clas s., pp. 505 ff.), and by T. W. Allen in the introduction to the facsimile (Facsimile of the Codex Venetus Marcianus 474, with a preface by J. W. White and an introduction by T. W. Allen, London and Boston, 1902), pp. 9 ff.

^{page 173 note 3} E.g. a gloss on 251 is placed after schol. 236.

^{page 174 note 1} Irregularities of this kind occur through out the MS., and are particularly in evidence where numerals of reference are used. See Zacher, , Handschr. und Class., pp. 513 ff.Google Scholar

^{page 174 note 2} Id., pp. 517 ff.

^{page 174 note 3} On Eq. 1077, 1151, 1169, and 1173.

^{page 174 note 4} E.g. schol. 436 ad fin. 445 (1. 24 Dübn.) 785 (I. 32 Dübn.)

^{page 175 note 1} Occasionally E forgets to insert the name of a speaker for which he has left a space, e.g. before in Eq. 1110

^{page 176 note 1} This reading is found also in M; but as we shall see when we come to examine it in detail, M appears to be descended from an edition compiled from more than one source; and in any case the point is unimportant. (The other MSS. read )

^{page 176 note 2} Other metrical notes found in E and omitted by V occur at 322 (E-Γ²), 380 (E.Γ²M), 1035 (EΓ³), 1037 (EΓΓ³Θ), and 1041 (EΓΓ³Θ). The last two are plainly Heliodorean, and are rightly included by J. W. White in his text of the Heliodorean scholia in The Verse of Greek Comedy. The others are doubtful, though all refer to passages on which no certainly Helio dorean scholion survives; that on 380 appears to be based on the Triclinian scholion given by Aid. at 335.

^{page 176 note 3} E places this scholion at 312; but though the normally denotes a change of metre, ‘the distich spoken by a coryphaeus that frequently follows a strophe he (sc. Heliodorus) includes with the song and does not give it the The following line has the even if it is the same metron’ (White, V.G.C., §850). Cf. e.g. Schol. Nub. 476.

^{page 176 note 4} No metrical scholion on Eg. not found in V, Θ, or Aid. is as yet accessible in print; White, like his predecessors, is dependent on Dübner's reports of the MSS. in this play.

^{page 176 note 5} E was first brought to notice in 1892 by Zuretti (Analecta Aristophanea, pp. 35 ff.), who attempted to show that it was Musurus's sole source. Zacher tested this theory over PL 1–200 and could not agree with it (see Bursian, 1892, p. 30).

^{page 176 note 6} E was used by E. Cary in his article, ‘The Manuscript Tradition of the Acharnenses’ (Harvard Studies in Class. Phil, xviii (1907), pp. 157 ff.) and by R. T. Elliott in his edition of Ach. (Oxford, 1914); it is also mentioned by W. Rennie and W. J. M. Starkie in their editions of the play (both London, 1909), although they do not give a full apparatus. Its readings in Av. are reported by J. W. White and E. Cary in their article ‘Collations of the Manuscripts of Aristophanes’ Aves' (Harvard Studies, xxix (1918), pp. 77 ff.); and it was used by White for his edition of schol. Av. (London and Boston, 1914).

^{page 178 note 1} was also conjectured by Musurus.

^{page 179 note 1} is found only in Γ²MSuid.

^{page 179 note 2} This note appears at first sight to be Heliodorean, but miscounts the lines, ignores the iambics 1035–6, and is inconsistent with the note at 1035 and the undoubtedly Heliodorean scholion at 1037 (see p. 176, n. 2), both of which are found in Γ³ itself. Perhaps the note should stand at 1051, where it would be correct and where no metrical scholion survives in the MSS.

^{page 180 note 1} The same scribe also wrote Nub. 1171 fin., Ran., and revised the work of his colleague who wrote PL and Nub. 1–1170.

^{page 180 note 2} Θ is by far the clearest of our MSS. in referring scholia to the text, which he does by numerals of reference, beginning from a on each page. He begins by using lemmata as well, but discards them after 374.

^{page 180 note 3} As we shall see, the correcting hands in Γ do not belong within the φ tradition

^{page 183 note 1} This (and ) are the nearest approximations found in the MSS. to the true reading restored by W. G. Clark.

^{page 183 note 2} Coulon (following Velsen-Zacher) is wrong in saying that Γ² has a variant the belongs to of schol. 755 (1. 16 Dübn.) which comes exactly opposite 1. 767 of the text.

^{page 184 note 1} We saw that in two such points (at 1001 and 1284) Θ² alone has the true reading.

^{page 185 note 1} Θ²s at 1172 is merely an explanation of the elision.