552
All other Greek MSS contain a greater or less Syrian element, and their Pre-Syrian elements almost always exhibit readings of all three Pre-Syrian types, though in different proportions. Nor is the general proportion of mixture by any means uniform throughout each document: thus the Syrian element of A is very large in the Gospels, much smaller in the other books, the transcription having probably been made from different smaller exemplars in different parts of the New Testament. The Western character of the Old Latin version in its earlier forms and apparently of the Old Syriac has been already noticed. The other early versions, the Memphitic and Thebaic, both Egyptian, are apparently altogether Pre-Syrian: they certainly are for the most part sometimes neutral, sometimes Alexandrian, though not without a Western element, which in the Thebaic is considerable. A revision of the Old Syriac version appears to have taken place early in the fourth century, or sooner; and doubtless in some connexion with the Syrian revision of the Greek text, the readings being to a very great extent coincident. All subsequent versions and revisions of versions are much affected by Syrian influence, more especially the Gothic and the 'Italian' Latin: but the Pre-Syrian elements of the Ethiopic, the Armenian, and the Jerusalem Syriac are large and important.
The textual elements of each principal document having been thus ascertained, it now becomes possible to determine the genealogy of a much larger number of individual readings than before in relation to the several ancient texts. The process can hardly be reduced to rule: but after a while the contrasted groupings of attestation become for the most part easy to interpret with patience and care. When once the ancient distribution of a reading has thus been ascertained, the characteristics of the several ancient texts furnish presumptions of the highest value as to its genuineness or spuriousness.
A reading marked as Syrian or Post-Syrian by the range of the documents which attest it may be safely rejected at once. If it has but one rival, that rival reading will be sustained by the united authority of all Pre-Syrian texts, Western, Alexandrian, and neutral alike. If there are two or more rival readings, this circumstance leaves untouched the antecedent improbability of all distinctively Syrian readings as deduced from the historical relations of the Syrian text as a whole to other texts. On the other hand it is a
