With the manuscript reading puri (1026) accepted or emended (multi, pupi, pusi, parui, poti), the dreaming bed-wetters have been identified as cleanly people, many people, innocents or children, boys, drinkers or drunks.Footnote ¹ Supposing that they are dinner guests, conuiuae, who have well and truly slaked their thirst, in sharp contrast with the desperately dreaming sitiens of lines 1024–5, I offer a new conjecture.

Brown has assembled several passages from Plautus, Lucilius, Horace and Martial illustrating a ‘minor comic and satiric theme’ of drunken bed-wetting and thinks that Merrill's conjecture poti, ‘those who have drunk’, would ‘tie in’ Lucretius’ lines to that theme; he cites among others Hor. Sat. 1.3.90, where a posited friend is not to be considered less of a friend if he has drunkenly urinated on his host's couch, comminxit lectum potus. As Brown admits, however, the replacement of poti by puri is hard to explain. Merrill himselfFootnote ² imagined two stages of palaeographic corruption, poti > puti > puri. I propose:

‘people who are full / who have had their fill of drinking’. The adjective, which may hardly stand substantively, will depend on an unexpressed subject, homines, just as the preceding participle sitiens probably depends on an unexpressed homo. plenus is used often of convivial dining (so, for example, Lucr. 3.938 plenus … uitae conuiua, 3.960 satur ac plenus … discede) and specifically of wine drinking: see, for example, Varro, Sat. Men. 154 plenus uini et ueneris; Ter. Hec. 823 uini plenum; Cic. Verr. 5.94, Red. sen. 13.3 uini, somni, stupri plenus. Of particular relevance to the proposed pleni is Ovid's juxtaposing of sitiens and plenus in his complaint that the thirsty reader drinks first from others’ cups, from Ovid's only when he has had his fill (illa bibit sitiens lector, mea pocula plenus, Pont 3.4.55); if Lucretius did in fact write pleni, that line would surely deserve a place in Zingerle's list of Lucretian ‘echoes’ in Ovid!Footnote ³ The dreamers then are probably conuiuae, and well-off ones, to judge by the reference to expensive Babylonian covers (1029, where––by the way––we might consider reading cum <et> Babylonica magnifico splendore rigantur ‘when even [sc. in the most costly cases of bed-wetting] Babylonian covers in their wondrous splendour are drenched’.Footnote ⁴ Rhetorically, the dreamers’ wealthy identity accords well with the disjunctive asyndeton, occupat amnem.| pleni saepe: the humble homo sitiens dreams of gulping a river of water, the rich pleni homines have had their fill of convivial wine.

As for the replacement of pleni by puri, an intrusive marginal or supralinear gloss, pueri (attesting to the prejudice, commonly held today, that children are particularly prone to bed-wetting), will have ousted the adjective, especially if it had been miscopied as pleri; for confusion of ‘r’ and ‘n’, compare 4.143 genantur (Lambinus, gerantur MSS). pueri then was corrected, metri gratia, to our manuscripts’ puri.