the shorter and earlier Greek text, which was rendered into Syriac and only survives in the Armenian abridgement. But who knows what changes this apology may not have undergone at the hands of the author of our existing Greek. The Armenian proves that some characteristic touches borrowed from The Preaching of Peter were obliterated when the earlier Greek was Avorked up into its existing form.
Till we have the complete Georgian text before us, it is rash to be too positive about its exact place in the development of the Christian form of the legend. But it is clear from Marr's account of it, as also from his printed excerpts which I have translated, that it had much in common with the simpler Armenian text. This is seen if we compare the forms of creed given in the Georgian and Armenian with the corresponding passage in the Greek (Boissonade, 83-84). In the old Arabic and Hebrew texts there is of course no corresponding passage, no counterpart at all. So a long extract from the Georgian about design in nature is closely similar to the Greek text, while the old Arabic form has barely the rudiments of the argument, which the Armenian omits altogether. Indeed, a careful comparison of the Greek and Georgian reveals that they have many points in common, which are absent from the old Arabic and Armenian, though this last fact has little significance, as the Armenian is a confessedly abridged text.
We must recognise, then, that already in the Georgian text, in spite of its comparative nearness to the non- Christian forms of the legend, evinced by its arrangement of matter and spelling of the proper names, the development of the latest Christian text as revealed in the Armenian and Greek has begun and even made fair progress. It cannot be regarded then as a Christianised form of the legend independent of the common parent of the Syrian, Armenian, and Greek texts, but must take its place far back in their common pedigree as the earliest stage of the Christian development of the tale. Next after it, longo sed proximus
