^{page 291 note 1} In Harvard Studies in Classical Philology, xviii. (1907), pp. iff., I have endeavoured to demonstrate that the principles propounded by Heliodorus and Hephaestion suffice for the analysis of the Aeolic lyrics in Greek Comedy. In the present discussion I shall attempt to prove that this theory of analysis is in accord with the probable origin of Aeolic verse, and I shall draw my illustrations mainly from Sophocles.

^{page 291 note 2} Apel, Böckh, Rossbach-Westphal.

^{page 292 note 1} The facts have been ascertained with great patience and are clearly stated. See Hermann Olden triberg,Die Hymnen des Rigveda, vol. i., Metrische respecund text geschichtliche Prolegomena (1888), pp. iff. From a following table (p. 14) it appears that of the sixteen possible forms of the first metre (two units in four places)fifteen occur. The four that predominate are, in the order of preference( is not found). See also E. Vernon Arnold, Vedic Metre in its Historical Development (1905), pp. 149 ff. On the metres of the Mahabharata, see E. Washburn Hopkins, The Great Epic of India, Its Character and Origin (1901), pp. 1916. The final metre of the cloka is iambic (p.239). For the Vedic acatalectic and catalectic tri meter of twelve and eleven syllables, ending respec tively see Oldenberg, op cit., pp. 42 ff., Arnold, op. cit., pp. 175 ff., and (for the Mahabharata and Ramayana) Hopkins, op. cit., pp. 273 ff. It is one of the many services rendered to metrical science by Rudolf Westphal that he noted many years ago the Avestan and Vedic dimeters and trimeters and pointed out their significant bearing on Greek metres. See Zur vergleickenden Metrik der indogermanischen Volkbr, Kuhn's Zeitschrift, ix. (i860), pp. 437 ff.

^{page 294 note 1} See Harvard Studies in Classical Philology, xviii. the polyschematist dimeter here assumes six different (1907), p. 26, where I have analysed the 45 verses tetrasyllable forms, 1, 5, 6, 8, 10, 11 above, of the parabasis of the Nubes. The first metre of

^{page 294 note 2} 1149

^{page 296 note 1} Sapphic hendecasyllable, Heph. 43, II f. c.

^{page 296 note 2} Phalaecean, Heph. 32, 21 f. c.

^{page 296 note 3} ‘Alcaic dodecasy liable,’ Heph. 33, 12 f. c.

^{page 296 note 4} Asclepiadean, Heph. 33, 5 f. c.

^{page 296 note 5} Pindaric hendecasyllable, Heph. 44, I2f. c.

^{page 296 note 6} These are true trimeters, and the analysis of the first pentameter into two dimeters connected by a single metre (von Wilamowitz, Choriambische Dimeter in Sitzimgsberichte d. preussischen Akademie, xxxviii 1902, p. 889) is inadmissible. The initial metre of the trimeter in this first pentameter is polyschematist Its prevailing form in the seven strophes is (once, in the sixth strophe, with the second long resolved , but the Variant occurs twice (in the third and fifth strophes).

^{page 297 note 1} “das verhaltniss, in das die attischen dichter die aeolischen und ionischen verse uberhaupt gesetzt haben, ist vielleicht das schwierigste problem derattischen metrik” (von Wilamowitz, Herakles²ii, p147)

^{page 298 note 1} Cf. in the strophe (787 f.)

^{page 300 note 1} “das abschliessende glied ist um eine sylbe langer als der glykoneus; diese verwendung des langeren gliedes ist nicht so haufig wie die des kiirzeren (des pherekrateus) zu diesem zwecke, aber auch gewohn lich” (von Wilamowitz, ad loc.).

^{page 300 note 2} “abschliessende reihe, nicht verkiirzt, sondern erweitertliberdenglykoneus, wiedasindem aeolischen nicht auf die wiederholung desselben metrons beru- henden, versbau gewbhnlich ist" (von Wilamowitz ad he). A period that is mainly in Ionic metre follows

^{page 300 note 3} TelesilIeum.

^{page 300 note 4} One form of the colon Reizianum. See Lindsay's classification of these cola in his edition of the Captivi (1900), p. 100

^{page 300 note 5} See Westphal, Die Gliederung der Aeschyleischen Tragodie, p. 20 ff., for an arrangement of this ode in nine metrically equivalent periods.

^{page 301 note 1} The famous Sapphic strophe consists of three catalectic polyschematist trimeters and the Adonius as here the Adonius follows two tetrameters.

^{page 303 note 1} Acephalization, in the first metre, of 7 or 12 (p. 293J and choriambic catalexis (p. 294).

^{page 303 note 2} See von Wilamowitz's discussion of the ‘primitive dimeter’ in his Choriambische Dimeter, pp. 886 f.

^{page 303 note 3} For examples of rovh (indicated by a dot) in verses already cited in this discussion see pp. 297 f. (‘iambic’ and ‘trochaic’ cola), 299, 302.

^{page 303 note 4} Cf. Alcaeus, frg. 15 (), and the fonddimeter ness of the lyric poets for the greater Asclepiadean.

^{page 304 note 1} Theory after theory as to the periodic relation of the parts of the strophe to one another has been advanced only to be abandoned. See, for example, Westphal's recantation of his doctrine of eurhythmy in his Griechische Metrik²(1868), pp. xvii ff. Metric is inadequate to deaf finally with this problem; the music to which these strophes were set would alone suffice, by plainly revealing equations and variations in the melody, to determine the poet's complete, detailed intention as to metrical structure. See the emphatic statement of Hugo Gleditsch in his Metrik³ (Handbuck der klassischen Altertumsiuissenschaft, ii. 3),1901, p. 103. Most hazardous of all is the attempt to conform all Greek lyrics uniformly to the simple model exhibited in Alcman's Partheneum, “Stollen, Gegenstollen, Abgesang” (a a b). That the Greek dramatic poets, in particular, who display at times parodos and stasimon as a whole, should never have the structure of the single strophe is in itself highly improbable, and the attempt to reduce all the strophes of tragedy to this simple type often compels resort to extremely arbitrary and fantastic devices. Otto Schroder, however, has reverted to this abandoned theory and would force it rigidly on all Greek lyrics. See his Cantica of Aeschylus, Sophocles and Aristophanes (with Euripides promised) and in particular his discussion of “ Binnenresponsion ” in Vorarbeiten zur Griechischen Versgeschichie (1908), pp. 136 ff. (=Neue Jahrbiicher f. d. klassische Altertum, xv., 1905, pp. 93 ff.), and the summary tables of analyses of the structure of the Odes of Pindar and Bacchylides in Vorarbeiten, pp. 105 f and 119f

^{page 304 note 2} If indeed the use of this clause can be established.See p. 299

^{page 304 note 3} “Dass das kein Fuss in dem Sinne gewesen ist, wie Iambus, Trochaus und alle die, welche durch ihre Wiederholung Verse bilden, konnen auch seine Verehrer nicht leugnen, deren es zur Zeit wieder giebt” (Choriambische Dimeter, p. 888).

^{page 306 note 1} Hermann wondered that Burney and Gaisford had not forthwith seen that he was right and the Greekgrammarians wrong, and takes them roundly to task (Elementa, pp. 222, 225 f.), but even Hermann did not deny that there was such a foot as the antispast (Ibid. p. 76).

^{page 306 note 2} These chance to be the forms of the two tetrameters by which Hermann attempted to establish his position that such verses are choriambic with antecedent ‘basis’ (Elementa, pp. 223ff.). He proposes four questions to Antispasticians, to which he thinks there is no answer: (i) Why may not the first half of each antispastic metre (syzygy) begin with the variable dissyllable (o o) ? (ii) How is the variable dissyllable to be accounted for at the beginning of even the first metre? (iii) Why are the metres that follow the first always purely antispastic and not dispondaic or composed of the first () or fourth () epitrite ? (iv) Why may not the last ‘iambic’ metre also be antispastic like the others? “Ad has omnes quaestiones nihil est quod respondere isti possint”! But all these questions are satisfactorily answered, I think, if one accepts the theory of the origin of Aeolic verse presented above.

^{page 307 note 1} “ Er is dazu unbrauchbar, da der Natur nach statt jeder der beiden Ktirzen eine Lange eintreten kann ” (Choriambische Dimeter, p. 888). Cf. Hermann: “Porro quidest, cur syzygiae, quae sequuntur primam, purum semper antispastum, neque etiam vel dis pondeum, vel primum quartumve epitritum recipiant qui pedes, quia per numeri legem et in principio et in fine antispasti syllaba anceps locum habet, non apparet quare exclusi sint” (Elementa, p. 223).

^{page 308 note 1} 1 give von Wilamowitz's text but not his cola In his commentary (Herakles² ii., p. 145), he declares in that synaphea is nowhere probable in this period since, if 663, 664 (, as he divides the period) were joined, the division of the resulting tetrameter would have to be and adds “man miisste so abteilen, weil in glykonean fur die abteilung der zusammenstoss der betontensylbenentscheidendist.” The example is not a fortunate illustration of his doctrine, since nobody can deny that may occur in the first metre of the ‘choriambic’ dimeter, but it is important to note the ground of his objection.

^{page 308 note 2} Quoted by von Wilamowitz in his Choriambische Dimeter, p. 871, in illustration of the forms of the ‘choriambic’ (polyschematist) dimeter, and analyzed in just the form here given.

^{page 308 note 3} Cf. Schroder's analysis in his Aeschyli Cantica, Dip.1. The antispast at the beginning of the third colon is, of course, unobjectionable, but he makes dodrans diiamb and dodrans + choriamb. It is in structive to note how Schroder disposes of the Asclepiadeans in Sophocles. He recognizes the tri meter (but assumes division into two cola), tetrameter Ph. 175 and pentameter ( O.C. I76f., Siof.). To the trimeter in Ant. 944 he gives the value of six theses, but seven to 945f., 949, 951 As an alternative he proposes minor Ionic scansion of 944–951 (Canlica, p. 18) with frequent use of the forms and Minor Ionics (which belong, it should be noted, to Ionian rhythm) offer at times a ready resource, if one would escape the antispast, but, as will be seen (p. 309), one only leaps by the use of this device from one difficulty into another. Von Wilamowitz {Choriambische Dip meter, p. 894) analyzes the catalectic Asclepiadean trimeters in Arist. Eq. 559 f. (a purely Aeolic ode) as ‘steigende Ioniker’ But, taking the variable forms of the first metre in Asclepiadeans ) into account, what defence can be made of Ionic trimeters in which the first metre is It is obvious, of course that Ionic scansion of the acatalectic Asclepiadean trimeter is impossible. Schroder analyzes the Ascle piadeans in O. C. 694–706 (p. 305 above), which chance to be all catalectic except one, as Ionics, reading alev instead of els aliv (cod. L) in 703 f. Yet the second verse in the lyric, has exactly the same metrical constitution as O.C. I76f., 510, which he scans differently (see above). He regards TV 851 f. an acephalous Asclepiadean pentameter, as composed of major Ionics (in the drama!), the series closing with the molossus

^{page 309 note 1} Dodrants or sesquimeters, to use Schroder's phraseology. See Philologus (1903), p. 162, note and Vorarbeiten, pp. 107 ff. Von Wilamowitz uses as models of these hexasyllabic phrases Horace's line Maecenas atavis edite regibus, dividing it in halves, Under this theory of analysis the greater Asclepiadean (a tetrameter) becomes with a single forlorn choriamb marooned in the middle of the verse.

^{page 309 note 2} “sed quoniam ea compositio sic facta est, ut dua arses [theses] sibi proximae sint, longe vehementiorem ille et asperiorem incessum habet, quam Creticus” (Elementa, p. 76).

^{page 309 note 3} See the quotation from von Wilamowitz on p. 308, note (“zusammenstoss der betonten sylben”).

^{page 309 note 4} ‘ Ictus,’in the sense of stress, plays a great ro1e in modern books on Greek and Roman metre, but see Goodell's consideration of the evidence in his Chapters on Greek Metric (1901), pp. 155 ff.