settlers, but the names of which are, nevertheless, probably very old, since the words of which they are composed are in many cases not retained in the colloquial tongue, in which they must once have been included, and are in some instances lost from the language altogether, so much so as to be only partially explicable even by scholars. The argument applies likewise, in their degree, to camps, barrows, and other artificial earth-works. Conclusions thus drawn, when established, rest upon a very firm basis. They depend upon the number and appositeness of the facts, and it would be very interesting to pursue this branch of evidence in detail. In following up this idea, the names to be sought for might thus be classed:— I. Names of the great features, involving proper names and actions. Cadair Idris and Cadair Arthur both involve more than a mere name. Idris and Arthur must have been invested with heroic qualifications to have been placed in such “seats.” II. Names of lesser features, as “Bryn y Saeth,” Hill of the Dart; “Llyn Llyngclys,” Lake of the Engulphed Court; “Ceven y Bedd,” the Ridge of the Grave; “Rhyd y Saeson,” the Saxons’ Ford. III. Names of mixed natural and artificial objects, as “Coeten Arthur,” Arthur’s Coit; “Cerrig y Drudion,” the Crag of the Heroes; which involve actions. And such as embody proper names only, as “Cerrig Howell,” the Crag of Howell; “Caer Arianrod,” the Camp of Arianrod; “Bron Goronwy,” the Breast (of the Hill) of Goronwy; “Castell mab Wynion,” the Castle of the son of Wynion; “Nant Gwrtheyrn,” the Rill of Vortigern. The selection of names would demand much care and discretion. The translations should be indisputable, and, where known, the connexion of a name with a legend should be noted. Such a name as “Mochdrev,” Swine-town, would be valueless unless accompanied by a legend. It is always valuable to find a place or work called after an individual, because it may help to support some tradition of his existence or his actions. But it is requisite that care be taken not to push the etymological dissection too far. Thus, “Caer Arianrod” should be taken simply as the “Camp of Arianrod,” and not rendered the “Camp of the silver circle,” because the latter, though it might possibly have something to do with the reason for which the name was borne by Arianrod herself, had clearly no reference to its application to her camp. It appears to me, then, looking back upon what has been advanced:— I. That we have throughout Europe, at an early period, a great body of literature, known as Mediæval Romance, which, amidst much that is wholly of Teutonic origin and character, includes certain well-marked traces of an older Celtic nucleus. II. Proceeding backwards in time, we find these romances, their ornaments falling away at each step, existing towards the twelfth century, of simpler structure, and with less encumbered Celtic features, in the works of Wace, and other Bards of the Langue d’Oil.