realise in what way the author of x went to work. Having before him the French romance of Aubry, he picked out a particular episode upon which absolutely no stress whatever is laid, which is but one of twenty or thirty other episodes, all possessing equal interest of conception and detail; he carefully eliminated all traces of the original personages and locale, he then modified every detail, and finally worked it into the Arthurian cycle. As Aubry's 100 lines did not give him enough matter, this 12th century writer went back to Julius Valerius and to Pseudo-Callisthenes, and spiced his narrative with a miscellaneous assortment of features, selected now from one now from the other. But even then he was not content, but went on a roving expedition through the Talmudic and Midrashic literature of the day, culling what he thought would fit in with his plan. All this while, with the severest self-denial, he rejected the many marvellous episodes which must have come before him in the course of his reading, and scrupulously refrained from retaining anything that could betray the Eastern origin of his narrative.
The assumption of x is a sufficient tax upon our credulity, but nothing to what is involved in the after development of the legend according to Dr. Gaster's theory. The legend writers fall into two classes: (i) the oldest of them all, Chrestien; (2) all the later writers. But these latter contain a host of details not to be found in Chrestien, argal they must have been present in x, for I do not suppose that Dr. Gaster imagines there was a bevy of writers at the close of the 12th century capable of harmonising Pseudo-Callisthenes and the Talmud. Now, how did these later writers act? They would seem to have gone upon the principle—when in doubt, consult x. For while in the main their presentment of the legend is that of Chrestien, each writer picked out some special feature of x which took his fancy, and added it to Chrestien's account.
All who are acquainted with the methods and nature of
