556
This is the natural result of the casual eclecticism of miscellaneous mixture, which tends to disguise the simplicity of the primitive relations of text under a superficial complexity of existing attestation. Before investigation has proceeded far, it becomes manifest that the groups which can by any possibility carry authority in doubtful variations are sure to contain one or more of a very small number of primary Greek MSS. In strictness the earlier Versions and Fathers should be included in the list of primary documents, and the process would certainly be incomplete if no account were ultimately taken of readings attested by them without the support of any primary Greek MS; but nothing is lost and much simplicity is gained by treating them in the first instance as accessory to Greek MSS.
The next step is to determine how far there is a common element in all or most of those groups which shew the best character when tried by 'internal evidence of groups'. Here two remarkable facts come out successively with especial clearness, the constant superiority of groups containing both B and א to groups containing neither, wherever internal evidence is tolerably unambiguous, and the general but by no means universal superiority of groups containing B to opposed groups containing א. These facts exactly correspond the one with the immunity of both MSS from Syrian readings, and the other with the almost complete immunity of B from the mixture with the chief aberrant Pre- Syrian texts which has largely affected א; while they are elicited from a different kind of evidence. They are moreover independent of the size of the groups. Thus the cases in which אB have no support from other Greek MSS, or no documentary support at all, are connected by every gradation with the cases in which they stand at the head of a considerable group. If B and א were for a great part of their text derived from a proximate common original, that common original, whatever might have been its own date, must have had a very ancient and a very pure text. There is however no tangible evidence for this supposition; while various considerations drawn from careful comparison of the accessory attestation of readings supported by אB together, by B against א, and by א against B respectively, render it morally certain that the ancestries of B and of א diverged from a point near the autographs, and never came into contact subsequently; so that the coincidence of אB marks those portions of text in which two primitive and entirely separate lines of trans-
