546
leave room for more than one interpretation. In what manner the genealogical principle can be applied to these more difficult cases will appear presently.
The documentary evidence for the text of the New Testament consists of Greek MSS dating from the fourth to the sixteenth century, most of the earlier being in a fragmentary state ; of ancient Versions in different languages ; and of quotations found in the extant remains of the Fathers, written in Greek, in Latin, and to a small extent in Syriac. In order to understand fully the history of the text, documents of all kinds and ages have to be taken into account ; though, as soon as the history is known, a vast numerical majority of documents must be treated as of no primary authority in ordinary variations. Since even the two earliest Greek MSS do not carry us back further than to the middle of the fourth century, the fixing of historical landmarks is chiefly dependent on the evidence of patristic quotations, which are for the most part definitely chronological, and also of the versions, three or four of which can hardly have been later than the second century. Each kind of evidence has its own imperfections. Quotations are often made from memory, and therefore liable to be loose and confused : different forms of text are used at different times by the same writer : and another kind of uncertainty is introduced by the diversity of text often exhibited by the MSS of patristic writings in quotations, which betrays the liability to corruption from the in- fluence of late current texts of the New Testament, and by the uncritical handling from which the text of most Fathers still suffers. Versions are affected by the genius and grammatical peculiarities of their language, and in other respects are not equally or uniformly literal ; while some have as yet been insufficiently edited. But all these drawbacks, however they introduce ambiguity into the evidence for single passages, do not materially impede the arrival at secure conclusions about the history of the text at large.
Comparison with patristic quotations discloses at once the striking fact that all the more considerable variations of reading must have arisen before the latter half of the fourth century. Variations of later origin are for the most part of little moment, and the changes which took place after that period were mainly changes in the distribution of readings already existing. A text virtually
