inductive criteria of texts, documents, and readings, we have at the same time found it alike undesirable and impossible to take any intermediate text, rather than that of the autographs themselves, as the pattern to be reproduced with the utmost exactness which the evidence permits.
376. Two qualifications of this primary aim have however been imposed upon us, the one by the imperfection of the evidence, the other by the nature of the edition. Numerous variations occur in which the evidence has not appeared to us decisive in favour of one reading against the other or the others; and accordingly we have felt bound to sacrifice the simplicity of a single text to the duty of giving expression to all definite doubt. In this respect we have followed Griesbach, Lachmann, and Tregelles: and it is a satisfaction to observe that Tischendorf's latest edition, by a few scattered brackets in the text and occasional expressions of hesitation in the notes, shewed signs of a willingness to allow the present impossibility of arriving every where at uniformly certain conclusions. Secondly, it did not on the whole seem expedient, in a manual text of the New Testament intended for popular use, to give admission to any readings unattested by documentary evidence, or to give the place of honour to any readings which receive no direct support from primary documents. Since then the insertion of any modern conjectures would have been incompatible with our purpose, we have been content to affix a special mark to places where doubts were felt as to the genuineness of the transmitted readings, reserving all further suggestions for the Appendix: and again, by an obvious extension of the same principle, the very few and unimportant readings which have both
21
