Welsh Journals

archaeologist, which is complex enough in itself: the comparative study of man's past handiwork, as a guide to his technical, economic, and even cultural and spiritual life. Later on, indeed, Atkinson himself advocates a more sensible arrangement, which would bring scientists into archaeol- ogical institutions to work full time on archaeological applications of their theoretical knowledge, and, in fact, to go on taking the initiative in such matters. This is, in fact, what some large institutions, like the British Museum and the London Institute of Archaeology, have already begun to do, and the ordinary archaeologist, in his provincial museum, should still have time left for archaeology. H. N. SAVORY. National Museum of Wales. LLYFR IORWERTH. Edited by Aled Rhys Wiliam. University of Wales Press, Cardiff, 1960. Pp. 164. 35s. This work is one of the most recent evidences of a long-delayed but welcome awakening in the study of the medieval Welsh lawbooks. It is described as 'a reconstruction from extant MSS. of the thirteenth-century lawbook of Iorwerth ap Madog'. This reconstruction has been undertaken because the 'Book of lorwerth' itself no longer exists: 'no MS. of those now extant', the Editor points out, 'can be regarded as the original Book of Iorwerth'. So the lost original can be recovered only by piecing it together, so to speak, from the group of still existing exemplars which were evidently derived, directly or indirectly, from it. Medieval Welsh lawbooks are made up of a varying number of sections, each of which 'treats'-the Welsh word is traethu, derived from the Latin tractare-of some specific subject, such as 'the laws of the court' or 'the laws of women'. These sections, though sometimes called 'books' in the Welsh texts, are not denominated invariably by that or any other single name; but since they are not infrequently described in Welsh as 'treating' their respective topics, they may conveniently be designated in English by the word 'tractate'. These tractates are of very varying lengths: in the present edition, some of them run to over thirty pages, while some barely extend to three. A lawbook so constructed may be 'reconstructed' at either of two levels. The 'reconstruction' may be at the level of merely trying to establish what were the tractates included in the lost book, and what was the order in which they were placed. Or it may be at the much more difficult level of trying also to establish, word by word, the text of the various tractates exactly as they figured in the lost book. The 'reconstruction' attempted in the present edition is at the former of these two levels. The Editor has concerned himself with the problem- Which tractates were included in the lost 'Book of Iorwerth' ? He has not attempted to reconstruct the word-by-word text of that lost book- in the way, for instance, that W. H. Stevenson reconstructed the text of