560
is scanty on one side or obscure. The ultimate determination must evidently be here left to personal judgement on a comprehensive review of the whole evidence. But in a text so richly attested as that of the New Testament it is dangerous to reject a reading clearly commended by documentary evidence genealogically interpreted, though it is by no means always safe to reject the rival reading. Here, as in the many variations in which documentary and internal evidence are both indecisive, it is manifestly right to abstain from placing before the reader an appearance of greater certainty than really exists, and therefore to print alternative readings, so as to mark the places where an absolute decision would at present be arbitrary, and also to mark the limits within which the uncertainty is confined.
The office of criticism thus far has been to discriminate between existing various readings, adopting one and discarding another. But it is at least theoretically possible that the originality of the text thus attained is relative only, and that all existing documents are affected by errors introduced in the early stages of transmission. Here there is no possible ultimate criterion except internal evidence: but the history of the text of the New Testament shews the meetingpoint of the extant lines of transmission to have been so near the autographs that complete freedom from primitive corruption would not be antecedently improbable. As far as we are able to judge, the purity of the best transmitted text does in all essential respects receive satisfactory confirmation from internal evidence. We have never observed the slightest trace of undetected interpolations or corruptions of any moment, and entirely disbelieve their existence. There are however some passages which one or both of us suspect to contain a primitive error of no great importance, and which are accordingly indicated as open to question, all suggestions for their correction being reserved for the Appendix.
This brief account of the text of the New Testament would be incomplete without a word of caution against a natural misunderstanding. Since textual criticism has various readings for its subject, and the discrimination of genuine readings from corruptions for its aim, discussions on textual criticism almost inevitably obscure the simple fact that variations are but secondary incidents of a
