558
of such groups are, it can hardly be doubted, Western, and many others Alexandrian. Still more unfavourable results are obtained by a similar testing of other single MSS.
These general results are such as might naturally be anticipated from the relations of א and B to other documents and to each other. It was to be expected that the text of the extremely ancient common source of B and א, which is shown by the concordant readings of NB to have been of singular purity, should as a rule be preserved in one or other of the two MSS where they differ; and further that B should usually, though not always, be its faithful representative. The wrong readings of B, with whatever amount of accessory attestation, being for the most part due only to sporadic corruption, it would naturally preserve a much larger amount of the common ancestral text than a MS so largely affected by Western and Alexandrian influences as א; and, as regards readings in which each of them stands alone, the different types of transcription characteristic of their respective scribes would naturally have similar consequences.
Although however a text formed by taking B as the sole authority, except where it contains self-betraying errors, would be incomparably nearer the true text of the autographs than a text formed in like manner from any other single document, it would certainly include many wrong readings; and the only safe criticism is that which throughout takes account of all existing evidence. The places in which the true reading appears to have been lost in both B and א are extremely few; but certain or possible exceptions to the usual superiority of B to א are many; and thus the various presumptions afforded by the internal character of various groups of documents are invaluable, while ' internal evidence of readings ' is often a helpful instrument of verification in the last decision, removing many uncertainties which must otherwise have continued unresolved, and again occasionally suggesting uncertainties which claim recognition. Such also, wherever the ancient texts are difficult to identify, are virtually the resources on which criticism depends in those parts of the Epistles which have perished in B, namely in the latter part (ix 14—end) of the Epistle to the Hebrews, in the Pastoral Epistles, and in the Epistle to Philemon. In the Apocalypse the authority of single documents is merged still more in that of grouped documents and in internal evidence; and the leading
