Abstract
The late Sir Roger Mynors, in a letter to Sebastiano Timpanaro quoted in the latter's Contributi difilologia e di storia della lingua latina , p. 543 n. 15, states that he had wondered ‘whether it might be a habit of Latin writers, when they were putting only one or two “parolette” between two pieces of Greek’, to use Greek rather than Latin: he invents as an example ‘θος κα πθος where logic demanded θος et πθος’. The answer is that they sometimes did: the present paper will concentrate on the type instantiated by his imaginary example, the use of κα for et. I do not claim to have recorded every case, but those I have observed are the following