As is well known, the first edition of Lucan's De bello civili that offered a critical apparatus in the modern sense of the word was Carl Hosius' edition, first published in 1892, then revised and reprinted twice, the last time in 1913. It is the editions of Hosius and Bourgery-Ponchont that give us the fullest information about MSS readings. The most important edition of Lucan in this century, that published by Housman in 1926, contains, on the other hand, only a selection of readings taken over from Hosius. In his preface as well as in his apparatus Housman is eager on any occasion to demonstrate the unreliability of the codex Montepessulanus, M, the favourite of his predecessor Hosius, and a scholar who has recently re-examined an essential part of the tradition of Lucan says about him: ‘The slighting of M and the MSS in general was Housman's peculiar contribution to the study of the transmission of Lucan's text. Rather than attempt to understand the nature of the evidence, he preferred to ignore it. It would be relevant to ask an editor with such an attitude why he bothers to provide an apparatus criticus.’ These hard words come from Harold C. Gotoff, who in 1971 published the monograph The transmission of the text of Lucan in the ninth century. Gotoff's book will certainly give valuable help to a future editor of Lucan as far as the composition of the critical apparatus is concerned, but not at all, I think, to the same degree when it comes to the constitution of the text. This may perhaps sound somewhat peculiar, but I will try to make clear what I mean.