complex and often mysterious story of the Marches. It is no mean achievement on Dr. Stanford's part to take us from the Upper Palaeolithic period to the eighteenth century in little more than two hundred pages, in such a way as to appeal to both the layman and the serious student. This is very much an excavator's book, concerned above all with those problems in the solution of which archaeology must play the largest part. Indeed, Dr. Stanford's analysis of 'how communities have striven to survive over the centuries within the geographical and political limitations' of the Marches is sometimes hampered by his unwillingness to look to other disciplines for help. Thus, even a cursory reading of the documentary evidence suggests a degree of over-simplification in the assumption that the distribution and purpose of the medieval Marcher castles can be satisfactorily explained simply in terms of a clash between the Anglo-Norman east and the Celtic west. Again, Dr. Stanford avoids much of the field of industrial archaeology-a label which he dislikes-on the grounds that it is 'a subject that is primarily the province of the economic historian, historical geographer and engineer'. It is for the pre-medieval periods that the book deals most convincingly with the patterns and problems of settlement in the Marches. Here the hillforts assume central importance in Dr. Stanford's discussion. They determine the geographical scope of the book, which concentrates on the lands visible from six of them: Sudbrook, Twyn y Gaer, Midsummer Hill, Croft Ambrey, the Wrekin and Moel y Gaer. It is on the basis of his own work on the hillforts that Dr. Stanford is able to set out the convincing argument that the early-fifth century B.C. 'saw one of the last major acts of colonization in lowland Britain before the arrival of the English settlers'. The author's intimate knowledge of the hillforts of the central borderland also underlies his stimulating discussion of the location of Tacitus's Decangi. Readers of this journal will regret Dr. Stanford's resolve 'not to be drawn too deeply into Wales'. But he had to draw the line somewhere, and it is no discredit to him to wish that his book were longer. Slips of any significance are few, but there was no Welsh rebellion in 1280 (p. 238), and Glyndwr's revolt was not suppressed in 1403 (p. 217). The printing of some of the figures, particularly the distribution-maps, could have been clearer. But the photographs are excellent, adding to our understanding of the text as well as conveying vividly the atmosphere of the Marches. DAVID STEPHENSON Cambridge THE SAINTS OF GWYNEDD. By Molly Miller. The Boydell Press, 1979. Pp. xvi, 132. £ 10. This volume is the first monograph in a series edited by David Dumville under the general title, Studies in Celtic History. Dr. Dumville describes Dr. Miller's book as an 'experimental approach to the difficult problem of the Celtic "Age of the Saints' but the author herself makes a lesser claim for the book, asserting that it consists of lists drawn up for personal use and of revised versions of 'notes which record the ideas that