Reviews, 239
It must not be thought that this is an idle question of termino- logy. In default of the explanation due to Sir John Rhys, the existence of the Four Branches cycle cannot be accounted for in any rational way. Literature in a society such as that of mediaeval Wales is the product of sociologico-psychological necessities, not, as it often is in advanced civilisations, of individual impulse. It can only exist and survive if it satisfies communal requirements, and plays its due part in the organised social scheme. The first question to be asked of any monument of primitive literature is, — what general need does it serve ? Sir John Rhys' hypothesis gives a satisfactory answer to this question as far as the Four Branches cycle is concerned.
Respecting the chronological order of the tales brought together in Lady Charlotte Guest's collection, Dr. Evans expresses opinions which I find myself unable to accept. First, be it noted that he puts the earliest actual Ms. date of any portion of the Mabinogion, (fragments of the Four Branches), at about 1235. But, as he shows at length, the earliest Ms. approve themselves copies of far older originals. In fact, the " paleographic evidence takes us back at a bound to the first half of the twelfth century. . . . The Four Branches are therefore demonstrably a century older than any manuscript containing them, which has come down to our time," (p. xiii). In my annotated edition of the Mabinogion I assigned the composition of the Foiir Branches cycle in its extant form to the last quarter of the eleventh century. As Dr. Evans' date is that below which the cycle cannot be brought, and as he does not preclude " the possibility of composition being a century or more earUer," (p. xiii), it will be seen that so far there is no quarrel between us. But I, in common with all earlier investi- gators, looked upon the Four Branches 2l^ the oldest portion of the collection. This Dr. Evans will not allow. For him "the Winning of Olwen is the oldest in language, in matter, in simplicity of narrative, in primitive atmosphere," (p. xiv). It may seem a matter of slight importance whether one Welsh fairy romance precedes or follows another. Not so ; if Dr. Evans' contention is admitted, our view of the whole development of Arthurian romance in the iith-i2th centuries is vitally affected.
Let me premise that both Dr. Evans and myself refer in our
