2. It will probably be admitted that in external proof Cureton’s theory is not strong, while yet the internal character of the version may be deemed by many powerfully to favour his view. Negligent or licentious renderings (and the Curetonian Syriac is pretty full of them) cannot but lessen a version’s usefulness as an instrument of criticism, by increasing our difficulty of reproducing the precise words of the original which the translator had before him; but in another point of view these very faults may still form the main strength of Dr. Cureton’s case. It is, no doubt, a grave suggestion, that the more polished, accurate, faithful, and grammatical of the two versions—and the Peshitto richly deserves all this praise—is more likely to have been produced by a careful and gradual revision of one much its inferior in these respects, than the worse to have originated in the mere corruption of the better (Cureton, Pref. p. lxxxi). _A priori_, we readily confess that probability inclines this way; but it is a probability which needs the confirmation of facts, and by adverse facts may be utterly set aside. Cureton’s remark that “upon the comparison of several of the oldest copies now in the British Museum of that very text of the Gospels which has been generally received as the Peshitto, the more ancient the manuscripts be, the more nearly do they correspond with the text of these Syriac fragments” (Pref. p. lxxiii), is confirmed by other, and subsequent, labourers in the same field. The received text of the Peshitto was printed from MSS. of a late type. It was the opinion of P. E. Pusey (whose name has already been mentioned in these pages) that a revision of the Peshitto text was made in the eighth century. The oldest Syriac Massoretic MS. which we possess is dated A. GR. 1210 = A.D. 899(20), but a copy of the Gospels (Add. 14,448), the date of which appears to be A. H. 80 = A.D. 699-700, contains a text which approximates to the type of the printed Peshitto, but exhibits marginal notes in a later hand, referring, however, chiefly to pronunciation and accentuation. There is no evidence that any formal revision took place; but it would appear certain that as questions of orthography, of grammar, and of pronunciation were fixed by the decisions of the Massoretes and grammarians, the faults (as they were deemed) of the older readings were emended by scribes. Hence it is, that if we open a codex of the Peshitto Gospels of about the date of the Codex Curetonianus, we find many resemblances of the kind indicated by Cureton, between the fifth century Peshitto text and the Curetonian text, because both belong to an early, and perhaps less accurate era of transcription(21). But the resemblances only extend to matters of grammar and spelling. In more important readings, the fifth century form of the Peshitto does not approximate to the Curetonian text. This was clearly seen by Pusey, as a result of the collation of a large number of Peshitto MSS.