Reviews of Books.

Y LLYVYR COCHO HERGEST. Y Gyvrol I. The Text of the Mabinogion and other Welsh Tales from the Red Book of Hergest. Edited by JOHN RHYS, M.A., Professor of Celtic in the University of Oxford; and J. GWENOGVRYN EVANS. Oxford, 1887.

THIS is the first instalment of a very important series of Welsh Texts. Most of the readers of the CYMMRODOR have already heard something of this series, and what they have heard has led them to anticipate some very good work. They will not be disappointed. The value of this edition, which is of course dependent upon the perfect accuracy with which it is worked out, is very great. We cannot all have access to the Red Book of Hergest, and many of us could not read it if we could get at it; but the Editors of this series, if they have carried out their work thoroughly well, will have brought its contents, for all practical purposes, within the reach of everyone. And the practical purpose for which this edition is suited is not the plain and simple one of reading the stories for amusement, but that of discovering, by a comparison of peculiarities of writing, mistakes, corrections, and spellings, the unknown quantity of literary history, philology, and archæology which may lie hidden among the lines of a carelessly written medieval manuscript. The stories themselves, as stories, matter very little in the present case; but if, by careful comparison of forms of letters, substitutions of one letter for another, and delicate little points of that sort, we can trace back the pedigree of such tales Peredur" a few steps beyond the Red Book of Hergest

as

and the "Parcival" of Wolfram of Aschenbach, perhaps into pre-Norman Wales, we shall have learnt something worth learning in the history of the literature of romances, and perhaps in the history of the Welsh people. But more of this anon.

The former edition of the Mabinogion, that of Lady Charlotte Guest, though excellent for its time, and though, as Professor Rhys remarks, approximating " to the original more nearly than that of any other Welsh text of any length", could not be relied on for perfect accuracy, and it would have been of no use to try to base theories upon it; while the portions of the Red Book printed by the Editors of the Myvyrian Archaiology were printed with little attempt. at accuracy of copying, and apparently not too much correction of proofs. In neither case was there any idea of reproducing the manuscript letter for letter, and the chances were, that if the original scribe had made a mistake, the editor corrected it. The present edition is based on quite another plan. There is a Triad, not found in the Red Book (or elsewhere), which says: "Three ways of printing an early text: the first, by making an exact facsimile, photographic or otherwise; the second, by constructing a critical text by comparison of the best MSS. and the correction of mistakes; the third, by contriving a diplomatic reproduction, wherein, though the exact forms of the manuscript characters are not given, the manuscript is transcribed character for character, letter for letter, word for word, spacing for spacing, error for error, correction for correction."

It is the last of these three that the Editors have adopted, and a plan has been adopted for which Mr. Gwenogvryn Evans claims some originality. Nine different alphabets are used, Gothic, Roman and Italic of various forms, each having a separate meaning. Ordinary capitals, semi-capitals, and small letters are represented by Roman type of different

sizes, rubricated letters by Gothic or thick letters, passages - retraced in modern times by Italics, and very faint passages by "hair-line" letters. Besides these, there are numbers of signs of contraction, puncta delentia, etc., all of which are carefully and consistently noted; and the two forms of the letters r, s, and w which occur in the MS. are carefully preserved. But one great feature of the system is the method of spacing. In the MS. the words are spaced in various ways. There may be meaning in them or there may not. That was not the Editors' business: all that they had to do was to reproduce them. This Mr. Evans claims to have done, so that "every scholar who has any knowledge of manuscripts will be able to restore in his mind's eye the exact spacing of the original, while the beginner will not be bewildered by treating simple words as compounds." Perhaps Mr. Evans need not have been so particular about bewildering beginners. His care for the beginner is his one deviation from perfect exactness. When two or more words are written as one in the MS.-e.g., "aoruc","aphandeuth", "acynylle”,—he separates them by what he calls space 1, a very small space. One does not precisely see why he should not leave them as they are in the MS., and leave the beginner-who, after all, is not likely to begin his study of Welsh with this book-to take care of himself.

This reproduction is certainly remarkably complete, and there are few things of the sort to rival it, except perhaps the transliterations in some of the palæographical publications of the British Museum; and if, when this book comes to be used by students, it should be found that the Editors have carried out their plan with the accuracy which they have set up as their ideal, and which a careful comparison of the facsimiles leads one to hope that they have attained to, students will not be able to praise it too highly.

One word as to the facsimiles. Photolithographs are not so expensive as other processes, but at the same time are less perfect; and, retouched or not, they are not always reliable. Good autotypes would be far better, and the risk of blurred and broken letters would be less.

There is one sentence in Mr. Evans's introduction which should not pass unnoticed.

"A few calligraphic peculiarities tend to show that the scribes were copying an original in the old Kymric hand which prevailed till the time of the Normans." The reason given for this opinion is perfectly sound, as any palæographer will understand. There is room for inquiries of remarkable interest in that direction. If the constant mistakes between s and r point to an original MS. in which those letters were very similar to one another, what and where may that MS. have been? There is no doubt that similar mistakes in the Cornish Vocabulary in the Cottonian Library have proved very clearly that the thirteenth century MS. contains specimens of a much earlier stage of the language. If the Mabinogion were really copied from a manuscript of preNorman date, it goes a long way to prove the Celtic origin of romances which were common to all mediæval Europe. "Diplomatic" texts like this may set distant critics to work on this very interesting point with some chance of

success.

Finally, the book is well printed, well got up, and on good paper, even in the less expensive "Student's Edition".

DICTIONARY OF THE WELSH LANGUAGE. By the Rev. D. SILVAN EVANS, B.D., Rector of Llanwrin, Machynlleth, North Wales. GEIRIADUR CYMRAEG. Gan y Parch. D. SILVAN EVANS, B.D., Rheithor Llanwrin. Carmarthen William Spurrell. London: Simpkin, Marshall, and Co.; Trübner and Co. 1887. (Part I. "A-Awys".)

WE welcome with much pleasure the completion of the first part of Mr. Silvan Evans's great undertaking. The magnitude of the whole work will be appreciated when we say that the first instalment, of 420 closely printed pages, does not quite exhaust the letter A. We are not able at present to do more than announce the issue of this Part, of which an extended notice will appear in our next number. The printing appears to be well executed, and the style of the work reflects great credit on the Carmarthen press.