R. Renehan's ingenious solutions to the problems of Symphosius 42. 1 and Anth. Lat. 207 in this journal (n.s. 31 (1981), 471 f.) are much to be welcomed. On the other hand, I do not think that his defence of the manuscript reading in Anth. Lat. 24. 3 marcent post rorem violae, rosa perdit odorem holds water. Taking rorem as = rorem marinum he explains that ‘the poet is not presenting us with a piece of botanical information about the relative seasons of the violet and rosemary; he means rather that all flowers wither and fade’. Actually, however, the poet on this showing does present information; and whether the information is botanically correct or not (I am not enough of a botanist to know), that is an odd way to make his point. Stranger still is his choice of rosemary out of all the fading flowers of field and garden. It was bound to be I misunderstood. As Renehan indicates, ros = ros marinus is supported only by Virg. Georg. 2. 212–13, where the identity has been doubted, and Plin. HN 24. 100, where ex rore supra dicto refers back to ros marinum in 99. Renehan may well be the first reader not to take rorem as dew. And an evergeen shrub (see Shorter Oxford English Dictionary) makes a singularly unfortunate illustration of floral decay, even though the shrub does produce a flower. Pliny classes it with herbs (19. 187; cf. 24. 99 ros marinum dictum est. duo genera eius: alterum sterile, alterum cui et caulis et semen rosinaceum, quod cachrys vocatur).