Zahn goes fully into these various forms (_Einleitung_, ii. 291-293). He begins by saying that it is impossible, except on a very loose view of the facts, to conclude that the Sinai-Syriac here preserves the original text, which was gradually displaced for dogmatic reasons by the modified form presented in (2), and ultimately by that given in (1). On the contrary, the Curetonian-Syriac preserves an early form of text, and one that had a pretty wide circulation, so that it cannot be due to an orthodox alteration of the Sinai-Syriac. “If it be the case that the latter, like the former, is derived from a Greek original, and that these two earliest versions of the ‘Distinct’ Gospel are not independent of each other but are two recensions of a single version, then it follows that the recension which agrees exactly with a demonstrably old Greek text (in this case the Curetonian Syriac) preserves the original form of the Syriac version; while, on the other hand, the one which deviates from all the Greek, Latin, and other forms of the transmitted text (in this case the Sinai-Syriac) is derived from the other by a process of intentional alteration.” There would be nothing to object to this reasoning were it not that, as it seems to me, there is a flaw in the second of the premises stated above, which of course vitiates the conclusion. In the main, it is true that the Sinai-Syriac and the Curetonian are not independent, but two recensions of a single version, but their common original was, as Zahn himself was the first to suggest, Tatian’s Diatessaron, which did not contain the first chapter of Matthew’s Gospel. So that the Sinai-Syriac may also go back to a Greek text (such as has been discovered in the _Dialogue of Timothy and Aquila_, see above, p. 99), and be earlier than the Curetonian. Zahn concludes his examination of this passage by saying: “We may give up all hope of finding in early manuscripts and versions any indication that Joseph was regarded as the natural father of Jesus by the writers of lost Gospels which may have been employed in the composition of the canonical Matthew and Luke. A writer like Matthew, whose purpose was to silence the calumnies raised against the miraculous birth of the Messiah, and who knew how to utilise the smallest details of an intractable genealogy to this end, cannot at the same time have accepted in his narrative statements directly contradicting his view of that occurrence. Any text of Matthew’s Gospel containing such features would be pre-condemned as one that had been tampered with in a manner contrary to the conception of the author.”