10 J.K. Elliott, Christian Amphoux and Jean-Claude Haelewyck papyri of Mark where extant, especially of course P45, ‫ א‬B C L ∆ 33 579 892 fam1 (= 1, 118, 131, 209) fam13 (mainly 13 69 124 346 543 plus 174 230 788 826 828 983) 28 565 700 1424 (all of which are well known to be allies of manuscripts from a much earlier date). There will be no attempt to exhibit an exhaustive critical apparatus, but merely to show significant variants as appropriate on each page between the normal allies of the six representative text-forms. Patristic citations are not included in the apparatus, but readers are to be encouraged to relate patristic quotations and allusions to the wording of a particular text form. It is, however, well known that unambiguous citations from the Gospel of Mark are few and far between. Two tools may help: the SBL series, The New Testament in the Greek Fathers attempts to relate patristic citations to known manuscripts and text-types of the New Testament. The series Biblia Patristica serves to highlight quotations in certain periods. So far we have concentrated on the Greek text of Mark - and for most users of Mark it will be the Greek text that is of paramount importance. The early versions of the New Testament text are, however, supremely important. The arguments for this do not need rehearsing in this context. The early versions are paramount in this project, hence the name Marc multilingue. Rather than combine the evidence found in all these early translations and subsume them in the apparatus to the Greek text, each language will have a volume on its own - ten in all, including the Greek. It is recognised that each version has its own distinctive history often quite independent of the Greek original, on which many are directly based. Table 2 shows sample pages from the projected fascicule on the Latin. In several ways the situation in respect to the Latin is more favourable than it is in the case of the Greek, especially as far as the Vetus Latina is concerned. The manuscript tradition as a whole is of manageable propor- tions; it consists of only thirteen witnesses17 and the texts are accessible, thanks to photographs made available by the Vetus Latina Institut in Beuron. The Old Latin witnesses may be divided as follows18. The Afri- can tradition is in two forms. Manuscript k (attested in the second half of Mark, but with lacunae) contains an ancient and pure form of the African text and one that antedates Cyprian; manuscript e (lacunose in the second half of Mark) also reproduces an African type of text, but one that betrays several levels of revision during which European readings 17 This is set out in Mélanges de science religieuse 56, 42-5. The mss aur and l should not be retained among the Vetus Latina witnesses as their text is Vulgate in Mark, and f ought also be set aside. 18 See Table 2.