Simrock's speculations, though marred by his standing tendency to claim over much for German tradition, are full of his usual acute and ingenious, if somewhat fanciful, learning. His ignorance of Celtic tradition unfortunately prevented his following up the hint given in the passage quoted above which I have adopted as one of the mottoes of the present work.
In 1855 Rochat published ("Ueber einen bisher unbekannten Percheval li Gallois," Zurich) selections from a Berne MS. containing part of Gautier de Doulens' continuation of Chrestien (v. 21,930 to end, with thirteen introductory and fifty-six concluding original lines, _cf._ p. 19), and entered at some length into the question of the origin and development of the Grail legend. The Mabinogi, contrary to San Marte's opinion, is placed after Chrestien. Villemarqué's ballad of Morvan le Breiz is the oldest form of the Perceval _sage_, then comes the Thornton Sir Perceval, a genuine popular production derived probably from a Welsh original. In spite of what San Marte says, the Grail incident is found in the Mabinogi, and it might seem as if Chrestien had simply amplified the latter. On San Marte's theory of the (Southern) origin of the Grail myth, this, however, is impossible, and the fact that the Mabinogi contains this incident is a proof of its lateness.
Up to 1861 all writers upon the Grail legend were under this disadvantage, that they had no complete text of any part of the cycle before them,[61] and were obliged to trust largely to extracts and to more or less carefully compiled summaries. In that year Mr. Furnivall, by the issue for the Roxburghe Club of the Grand St. Graal, together with a reprint of Robert de Borron's poem (first edited in 1841 by M. Franc. Michel), provided students with materials of first-rate importance. His introductory words are strongly against the Celtic origin of the story, and are backed up by a quotation from Mr. D. W. Nash, in which that "authority who really knows his subject" gives the measure of his critical acumen by the statement that the Mabinogi of Peredur can have nothing to do with the earliest form of the legend, because "in Sir T. Malory, Perceval occupies the second place to Galahad." In fact, neither the editor nor Mr. Nash seems to have tried to place the different versions, and their assertions are thus of little value, though they contributed, nevertheless, to discredit the Celtic hypothesis. San Marte, in an essay prefixed to the first volume, repeated his well-known views respecting the source of Wolfram's poems, and, incidentally, protested against the idea that the Mabinogi is but a Welshified French romance.