A NEW TRANSLATION OF THE MABINOGION by Prof. Ifor Williams NINETY years have passed since the Four Branches of the Mabinogion, and the Romances, contained in that fine old manuscript, the Red Book of Hergest (written about 1400), were translated into English by Lady Charlotte Guest and Tegid; and as the present translators admit in their preface, "the faults of that translation are few, though at times important, and its charm and literary qualities will always ensure for it the foremost place in the field which Lady Guest was the first to explore completely." The discovery, however, of an older and better text of most of these tales in an earlier manuscript, the White Book of Rhydd- erch, dating from about 1325, and the excellent diplomatic editions of both versions by Dr. Gwen- ogvryn Evans and Sir John Rhys, have made it possible to improve on the old translation in many ways. 10 quote the preface again, we need "a version more accurate in details than that which has hitherto been available," and that is what Mr. T. P. Ellis and Mr. John Lloyd have attempted to give us in these two volumes.1 Short introductory notes are prefixed to each tale, and occasionally footnotes are utilised for dealing with minor points, but there is no claim to have produced a fully annotated edition. The editors stress the fact that their primary object was to translate and it is by the accuracy of their version in details that they must stand or fall. And on the whole, I think their translation must be acknowledged to be much more accurate than Lady Guest's. After all, this is what one would expect. During the ninety years that have intervened, scholars like Zeuss, Ebel, Rhys, Loth, and many others have worked hard and successfully on Welsh lexicography and syntax, and their researches have smoothed the path of the translator of the Mabinogion. Though there is still room for debate as to the precise meaning of a few words and phrases in these tales, the bulk of the vocabulary has now been investigated, and the meanings determined. Loth's translation into French is very useful; in spite of several "howlers," and the present trans- lators have used his version with profit. But what was not to be expected, and what one finds difficult to condone is the number of inaccuracies left uncorrected in their translation and the num- ber of fresh ones for which they themselves are responsible. Before touching upon these, I should like to say how glad I am to find that the editors have 1 The Mabinogion. A new translation by T. P. Ellis, M.A., and John Lloyd, M.A., Oxford. Clarendon Press. Vol. I., i-xii, 1­232, Vol. II. 1-253. 5/- per vol. accepted the correct explanation of the term mabinogi, and discarded Rhys's old theory of the "bardic apprentice," styled mabinog, whose in- dentures can only be found in the faked docu- ments of lolo. We are well rid also of Mana- wyddan and Llew Llaw Gyffes, though unfortun- ately the wrong form of Rhufawn's name has been adopted (11 9 note) and Plynlymmon for Pumlumon, p. 169. Some of the short foothotes are well done, especially, those dealing with legal matters sound stuff packed in a few sentences. The introductory notes had to be short. and in some instances they are also slight. That, I am afraid, could not be helped. It is almost impos- sible to deal succinctly in a page or so with the general problems raised by tales of such varied character and development. As for the printing and general appearance of these volumes they are worthy of the Clarendon Press, and no more need be said. The rest of the space allotted to me must be devoted to the errors which a cursory examin- ation brought to light. Vol. 1., 5 llannerch (not llanerch) "glade," cannot be a compound of llan and etch "dark," and does not mean a "dark clearing." Annwn does not mean "the Abode of Death," but "Fairy land." It is quite incorrect to translate llithiaw as "set"; it is the usual term for feeding or fleshing hounds after they have brought down their quarry. P. 8 a rodych is the jussive sub- junctive, and the translation offered "one stroke that you give him" is inaccurate. P. 10 macwy is not coincident in origin with the Gaelic mac, but a direct borrowing of the Irish mac coemh. P. 13 neut teruynedic angheu y mi cannot mean "it is the appointed death to me" but simply "I am mortally wounded," my death has been ac- complished. P. 16, n. 44 the bwyta kyntaf is not the first meal of the banquet, for if so, how could the meal on the following day be also styled y bwyta kyntaf? This name was given to the meal partaken of by the lord and his retinue after they had finished, the gwasanaethwyr sat down to enjoy what was left, while their lords went out for a stroll cp. W.M. 32 b. P. 19, gwrthpwythi means "obstinacy," not "haste," cp. M.A. 492 b, 500 a, 528 a, 681. P. 23, tormynnawc kyuoetnaxoc should not be translated "rich in herds," for the first word refers not to "cattle" but to hosts of armed men. The note on p. 24, about the magic bag and the Holy Grail calls for comment As for p, 25, n. 66, may I remind the editors that there was no initial g in the Early Welsh adawaf, and there- fore this note is useless? The sense required is not "promise" but "allow, permit." P. 26, n. 67, ceimat, is derived from camp not from cam,