The data for this study have been collected, classified, counted, and numeri- cally evaluated by the author personally. Their presentation in full allows verification and opportunity for further observation and experiment. The texts for the first two sections and for the humanists Politian, Sannazarius, and Faustus Andrelinus (the ‘puritans’) have been read for ‘homs’ at least twice, the rest once. Further examination of these latter might be expected to yield some unnoticed ‘homs’, perhaps to the order of an added 10%, but nearly all would be in the ‘omega’ category, mostly ‘()ls’, merely reinforcing the evidence already assembled. An article ‘Homoeoteleuton in non-dactylic Latin verse’ (Riv. di Fil. 120 (1992). 61-71) was submitted for publication before work began on the present study, of which it is wholly independent. Two important differences emerge. First, the incidence of ‘hom’ in classical dactylic verse with some exceptions (notably Horace, early and late Ovid, Germanicus, and Manilius) is far lower. A superficial observation indicates that the same holds good for Greek verse. Secondly, non-dactylic verse seems to take no special account of the ‘alpha’ category (noun plus attribute), so that its importance in dac- tylic verse came as a surprise. Ann Arbor, 1993 D. R. Shackleton Bailey CONTENTS INTRODUCTORY …………………………… . 1 CLASSICAL AND PRECLASSICAL ………………… . 11 Appendix Vergiliana ……………………………. . 11 Calpurnius … … … … … … … ………………. . 15 Catullus …………………………………… . 16 Cicero ……………………………………. . 18 Columella ………………………………….. . 20 Consolatio Liviae … … … … … … … … ………… . 20 Ennius ……………………………………. . 21 Germanicus … … … … … ……………………. . 24 Grattius …………………………………… . 25 Horace ……………………………………. . 26 Bias Latina … … … … … … …………………. . 32 Juvenal ……………………………………. . 34 Laus Pisonis … … … ………………………… . 36 Lucan …………………………………….. . 37 Lucretius …………………………………… . 41 Lygdamus ………………………………….. . 48 Manilius …………………………………… .