That the vitality of the Legend is in part, if not wholly, due to the strange crossing and blending of its sources, I at least have no doubt. To discuss these sources at all, much more to express any definite opinion on the proportions and order of their blending, is a difficult matter for any literary student, and dangerous withal; but the adventure is of course not to be wholly shirked here. The matter has, both in England and abroad, been quite recently the subject of that rather acrimonious debating by which scholars in modern tongues seem to think it a point of honour to rival the scholars of a former day in the classics, though the vocabulary used is less picturesque. A great deal of this debate, too, turns on matters of sheer opinion, in regard to which language only appropriate to matters of sheer knowledge is too often used. The candid inquirer, informed that Mr, or M., or Herr So-and-so, has "proved" such and such a thing in such and such a book or dissertation, turns to the text, to find to his grievous disappointment that nothing is "proved"--but that more or less probable arguments are advanced with less or more temper against or in favour of this or that hypothesis. Even the dates of MSS., which in all such cases must be regarded as the primary data, are very rarely _data_ at all, but only (to coin, or rather adapt, a much-needed term) _speculata_. And the matter is further complicated by the facts that extremely few scholars possess equal and adequate knowledge of Celtic, English, French, German, and Latin, and that the best palaeographers are by no means always the best literary critics. Where every one who has handled the subject has had to confess, or should have confessed, imperfect equipment in one or more respects, there is no shame in confessing one's own shortcomings. I cannot speak as a Celtic scholar; and I do not pretend to have examined MSS. But for a good many years I have been familiar with the printed texts and documents in Latin, English, French, and German, and I believe that I have not neglected any important modern discussions of the subject. To have no Celtic is the less disqualification in that all the most qualified Celtic scholars themselves admit, however highly they may rate the presence of the Celtic element in spirit, that no texts of the legend in its romantic form at present existing in the Celtic tongues are really ancient. And it is understood that there is now very little left unprinted that can throw much light on the general question. I shall therefore endeavour, without entering into discussions on minor points which would be unsuitable to the book, to give what seems to me the most probable view of the case, corrected by (though not by any means adjusted in a hopeless zigzag of deference to) the various authorities, from Ritson to Professor Rhys, from Paulin Paris to M. Loth, and from San Marte to Drs Foerster and Zimmer.