MIRACLE CYCLES. 315 It appears to me that the compiler of is selecting and recombining at will matter that lies before him in a written text. If so, that text was a manuscript intermediate between F and C, and we will call this K as above. Consequently an ex- amination of x will throw light upon the relation of C to its immediate source, and we find that K must have approximated far more closely to the Y-W text than does C. This explains why x, although nowhere parallel to W where C is not, yet retains in a number of passages more or less the wording of the original where this has wholly disappeared from C. But we still have to account for the fa<5t that x is nowhere parallel to C where C is not parallel to W. It is obvious that C con- tains, and we may infer that /c contained, various passages that were not in F, yet of these passages X shows no knowledge whatever. Now, the most important of these, and the only ones of which we can say with certainty that they were present in K, are scene 2 and the Commandments. Scene 2 is wholly absent from x ; *h e Commandments are rehearsed in a quite different form. Why ? The fac~l that the compiler of x is constructing a short piece to form an appendix as it were to the c Purification/ and is therefore compressing his material as much as possible, will account, I think, for his omission of scene 2. The case of the Commandments is more difficult, but may, I think, be explained on the same ground. Y and W, though different, both endeavour to combine what I may call the ' duologue ' honour thy God, love thy neighbour with the ordinary decalogue, by
