Page:The New Testament in the original Greek - 1881.djvu/653

555 this or that principal ancient text is open to doubt; or in which there is little or no reason to suppose that the divergence of reading has any connexion with the divergence of the principal ancient texts. Here too however the genealogical principle can he applied by an extension of 'internal evidence of documents' to the lost ancestors of groups of documents. The general internal character of distinctively Western and of distinctively Alexandrian readings was ascertained in precisely the same manner as the general internal character of any single document is ascertained, namely by consecutive examination of the whole body of readings; and the power thus given of employing easy variations as a key to difficult variations is of universal range, the same mode of testing general internal character being applicable to the whole body of readings of any other group of documents which frequently stands out in opposition to other documents. In every place in which two or more documents have the same reading, unless the reading is such as can naturally be accounted for by accidental coincidence, they must by the nature of the case have had a single common ancestor, whether it be the autograph or some later MS. If the same group of documents is found standing by itself in a considerable series of readings, sufficient material is provided for generalisations as to the common ancestor in all these places, which ancestor is virtually a series of fragments of a lost MS. This 'internal evidence of groups', by rendering it possible to estimate as wholes the documentary arrays by which rival readings are attested, independently of any estimates that may be formed of the character of their constituent members individually, escapes the difficulties caused by mixture which beset every attempt to treat individual documents of the New Testament as so many ' authorities ' of constant value.

The number of groups that deserve serious attention is soon found to be comparatively small. Neither Greek MSS containing a large amount of distinctively Pre-Syrian text nor early Versions nor early Fathers are numerous, and to a great extent they are fragmentary or discontinuous; and combinations into which none of them enter may evidently in most cases be safely neglected. It is likewise soon found that various groups practically identical are somewhat variable in their limits through the defection of one or another of the documents which arc habitually their members.