Cylchgronau Cymru

grateful to know what reliable authority exists for the confident assertion that Swansea was a 'Danish settlement' where 'Harry Beaumont' in 1099 built a castle 'on a mound of glacial gravel just west of the river mouth' (p. 397). The editor rightly draws attention to the difficulties of Welsh place-name spellings. Nevertheless 'Graig Llwyd' (p. 133) and 'Barcloddiad y Gawres' (p. 312) ought never to have appeared, and 'Presceli' (p. 330) is the most horrible hybrid imaginable. To end on a note of minor criticisms would be ungenerous. The compilation of this book was an onerous task. An advanced regional geography of Wales has long been needed. Historians, along with many others, will be grateful to Professor Bowen and his colleagues for so useful and comprehensive a survey. G. W. Swansea. THE STONES OF LLANDAFF CATHEDRAL. By F. J. North. University of Wales Press, Cardiff, 1957. Pp. viii, 122. 15s. By examination of the terrain Dr. North shows that the site later occupied by Llandaff cathedral was probably chosen, in part at least, because of its position on a Roman road running south-westward from Caerleon. No more than conjecture is possible with regard to the appear- ance of the monastic church, 28 'feet' long, 15 wide, and 20 high, to which, about 1120, bishop Urban, in order to acquire the reputation (and, perhaps, the takings) of a reliquary church, brought from Bardsey the corpse of St. Dyfrig. The monasterium majus in honore Petri apostoli which Urban began was extended by his successors, provided with a tower by Jasper Tudor, neglected by later ages, restored in the nineteenth century and, finally, in large part destroyed by German airmen in 1941. In the absence of fabric rolls the historian can know nothing with certainty about the conduct and cost of the building operations or the men who shaped and set the stones; but Dr. North's researches have greatly extended knowledge on at least one matter, the provenance of the stone, for much of which water transport was available, used or reused in the various stages of the work. His interest in the formation and vast antiquity of the building materials will, perhaps, be shared by few, but at least one geological ignoramus is grateful for his useful account of the manufacture of plaster of paris (p. 92), his remarks on the effect of limestone on sand- stone (p. 72), and the excellent illustrations of rocks and building work in this volume. One may doubt whether he is quite correct in connecting wrong notions about the composition of Purbeck stone with disuse of that material, for, though it no doubt went out of fashion for pillars, it was very commonly used in seventeenth-century London for other purposes, especially paving. Grange-over-Sands, G. P. JONES.