tionary. They presuppose that in a vast body of literature the oldest existing text really belongs to almost the latest stage of development; that the later texts, although unnoticed anywhere, must really have existed in some form long previously; and that in the twelfth century a purely Christian subject was almost entirely de-Christianised, precisely in the actual oldest but hypothetically youngest text, whilst the later ones, hitherto regarded as later in origin as well as in date of composition, show a steadily increasing Christian element.
I cannot criticise these views in detail. An indispensable preliminary is a methodical re-examination of the entire Grail literature, and the testing of every point urged by Prof Heinzel.[1] Until I have been able to do this I prefer simply to direct the attention of all students to his work. I may, however, note one or two points. Prof Heinzel confines himself to the French romances; neither the Peredur, nor the English Sir Perceval, nor Wolfram's version is examined with any thoroughness. Yet it is precisely from these outside versions that light is, I believe, to be obtained respecting the early form of the legend. Prof Heinzel confirms indirectly, but in the strongest way, my contention that the Early History of the Grail, being as it is a conversion of Britain legend, is essentially a British product. Though we only possess it in a French dress, it must have originated and been developed in Britain.
Turn we now to Prof Rhys's Grail studies. I had busied myself exclusively with establishing the Celtic nature of the Perceval portion of the legend, deeming, as I did, the Galahad portion to be derived and secondary.
- ↑ Prof. Heinzel's method of presenting his results is stimulating but arduous. He examines each version by itself, so that one has to consider some twenty pieces of independent investigation, to determine for oneself their relation to each other, and then to recheck the whole by the texts. I may add that one has to make one's own index and analytical summary.
