A new conjecture eam <suaui> deuinctam and A. Palmer’s conjecture eam <leni> deuinctam (preferably the latter one) are recommended in Cat. 64.122 on the grounds of plausibility of haplographic omission of both inserted words: for the first one, suaui deuinctam > deuinctam (cf. Lucr. 4.453–454, Varro Logistor. fr. 28 Bolisani); for the second one, leui deui- > deui-, given the affinity of n and u in Latin minuscule script (cf. Enn. Ann. 2 Skutsch, Lucr. 4.1009, Hor. Carm. 3.1.22–23, Ou. Met. 8.823, Ou. Fast. 4.653, Ou. Trist. 4.3.22). The conjectures preferred by modern editors and researches of Catullus’ text (viz., K. Lachmann’s <uenerit>, aut ut, preferred by most modern editors, Pomponius Laetus’ eam <placido> deuinctam, preferred by D.R. Shackleton Bailey and G. Trimble, and J.M. Trappes-Lomax’s aut ut securo deuinctam) do not meet the requirement of providing a plausible explanation of the process of corruption. Both conjectures recommended also correspond in sense to the epithet dulci that we find in the corresponding place of a similar expression in Ciris 206 (a passage obviously dependent on Catullus) and that E. Baehrens argued to have correspondences in a number of passages connected with Ariadne’s dream (Philostr. Imag. 1.15.1, Nonn. Dion. 47.320, Prop. 1.3.7); however, inserting the word dulci itself in Cat. 64.122 (as done in MS Grenoble, Bibliothèque Municipale 549 and in ed. Romana 1472) is impossible, as pointed by G. Trimble, since Catullus uses the word dulcis just two lines earlier, in Cat. 64.120.