for disbelieving them. But in these earlier forms--the authority of the
most learned Celticists who have any literary gift and any appreciation
of evidence is decisive on this point--not only are the most
characteristic unifying features--the Graal story and the love of
Lancelot and Guinevere--completely wanting, but _the_ great stroke of
genius--the connection of these two and the subordination of all minor
legends as to the dim national hero, Arthur, with those about him--is
more conspicuously wanting still. Whether it was the Englishman Walter
Map, the Norman Robert de Borron, or the Frenchman Chrestien de Troyes,
to whom this flash of illumination came, has never been proved--will
pretty certainly now never be proved. M. Gaston Paris failed to do it;
and it is exceedingly unlikely that, where he failed, any one else will
succeed, unless the thrice and thirty times sifted libraries of Europe
yield some quite unexpected windfall. In the works commonly attributed
to Chrestien, all of which are well known to the present writer, there
is no sign of his having been able to conceive this, though he is a
delightful romancer. Robert is a mere shadow; and his attributed works,
_as_ his works, are shadows too, though they are interesting enough in
themselves. Walter not only has the greatest amount of traditional
attribution, but is the undoubted author of _De Nugis Curialium_. And
the author of _De Nugis Curialium_, different as it is from the
Arthurian story, _could_ have finally divined the latter.

But at the time when he wrote, Englishmen, with the rarest exceptions,
wrote only in French or Latin; and when they began to write in English,
a man of genius, to interpret and improve on him, was not found for a
long time. And the most interesting parts of the Arthurian story are
rarely handled at all in such early vernacular versions of it as we
have, whether in verse or prose. Naturally enough, perhaps, it was the
fabulous historic connection with British history, and the story of the