^{page 98 note 2} W. B. Sedgwick has already done so in Mnemosyne, Series 4, vol. iii (1950), from a different point of view.

^{page 98 note 3} 67. 27 (an emendation); 68. 81. In Callimachus too, 2W+3W is uncommon (R. Pfeiffer on frag. 230).

^{page 98 note 4} Platnauer compares Callimachus Loutra Pallados, 57 per cent.

^{page 98 note 5} 67. 35; 76. 25; 77. 1; 90. 3; 107. 5; 113. 3.

^{page 98 note 6} 68. 49; 76. 1; 84. 5; 101. 1. Cf. Munro, , Criticisms and Elucidations of Catullus ² (London, 1905), p. 154.Google Scholar

^{page 99 note 1} See p. 98, n. 1.

^{page 99 note 2} 68. 15, 39, 141a, 159; 69. 7; 86. 5.

^{page 99 note 3} 66. 27, 29; the first of these is an emendation, cf. p. 95, n. 3.

^{page 99 note 4} See Kroll, ad. loc.

^{page 99 note 5} See Vollmer, , Röm. Metrik. (Gercke Norden, Einleitung, i. 8, p. 12).Google Scholar

^{page 99 note 6} atque elided at 68. 82; 73. 6.

^{page 99 note 7} 66. 76; 76. 18; 87. 4.

^{page 99 note 8} Six in the pentameter (66. 48; 67. 44; 68. 158; 76. 10; 97. 2; 99. 8) and one in the hexameter (107. 1). Cf. classifications in Catulli Veronensis Liber edited by Friedrich, (Teubner, 1908), pp. 95–96Google Scholar, and by Cazza-niga (Torino, 1940), p. 131

^{page 100 note 1} See Hartenberger ‘de o finali’, who points out on p. 27 that the final -o of originally iambic words in Catullus is some times long and sometimes short, although the adverb modo elsewhere always ends with however, occurs at 10. 7.

^{page 100 note 2} ‘non (ita me di / ament) quicquam referre putavi.’

^{page 100 note 3} For an analysis of the words preceding 2D see Kroll on 64. 21.

^{page 100 note 4} 90.3.

^{page 100 note 5} Cf. Marx, F. (Abhdl. d. säcks. Ak. d. Wiss. Phil.-hist. Kl. xxxvii [1922]Google Scholar, I ff.), who stresses that in this respect Catullus resembles the Augustans rather than his predecessors and Lucretius. Cf. Wilkinson, , C.Q. xxxiv (1940), 33. n 6.Google Scholar

^{page 100 note 6} See Platnauer, M., op. cit., pp. 15–16Google Scholar, and Wilkinson, L. P., op. cit., pp. 37–40.Google Scholar

^{page 100 note 7} Lines 3, 41, 57, 61.

^{page 100 note 8} See p. 99 above.

^{page 100 note 9} In 11. 15, 25, 31.

^{page 101 note 1} The spondaic ending occurs in the Elegiac poems of Callimachus (R. Pfeiffer on frag. 303) and rarely in the Augustan Elegists (Platnauer, , op. cit., pp. 38–39Google Scholar). The trisyllabic spondaic ending also is Callimachean (Pfeiffer on frag. 756). Ovid has three certain examples (cf. Platnauer, p. 39). In none of these poets is the last word in a spondaic ending shorter than three syllables (cf. A. S. F. Gow on Theocritus i. 65).

^{page 101 note 2} At 11. 65, 87, 89, 109. The last three lines end with Greek proper names.

^{page 102 note 1} Cazzaniga (op. cit., pp. 139–46) has an analytic list of 356 of these.

^{page 102 note 2} Nobbe, , de metris Catulli (1820)Google Scholar. Cf. Lindsay, , Early Latin Verse, p. 309.Google Scholar

^{page 102 note 3} Cf. Svennung, J., Catullus Bildersprache, i (Lund, 1945), 25–30.Google Scholar

^{page 102 note 4} 68. 89–90.

^{page 102 note 5} 68. 90; 75.4; 77.4; 104.4.

^{page 102 note 6} 66. 48; 68. 158; 99. 8.

^{page 102 note 7} Munro, H. A. J., op. cit., pp. 151–2Google Scholar; Wheeler, A. L., Catullus and the Traditions of Ancient Poetry (Berkeley, California, 1934), p. 168.Google Scholar

^{page 102 note 8} Schuster, M., op. cit., p. 109Google Scholar; De Gubernatis, M. Len chantin, Il libra di Catullo ² (Florence, 1933, reprinted 1953)Google Scholar; Sedgwick quotes four other similar judgements, op. cit., p. 65, n. 2.