the Banwy bridge, which seemed to threaten the future of the line, was overcome. The full restoration of services to Welshpool is now the goal, even though this will only be to Raven Square; the quaint journey through the centre of Welshpool will be lost for ever. This book is well-presented and illustrated, and the appendices comprehensively detail the minutiae of the line and its operation. It may be warmly recommended to the railway historian and to the general reader. R. V. BARNES Swansea. MIGRATION AND URBAN DEVELOPMENT. By Brinley Thomas. London, Methuen and Co., 1972. Pp. xvi, 259. £ 3.80 (paper-back, £ 1.90). The title of Professor Thomas's most recent book is modest to the point of being misleading. The subject matter is not confined to the influence of migration on the development of towns, but rather ranges over the whole question of the development of the international economy-and particularly its Atlantic section-since the mid-nineteenth century and up to the present. This international economy is, moreover, viewed as a system with a pattern of internal dynamics which is something more than the unpredictable resultant of forces generated independently from within separate national economies. The book falls into three main sections. The first four chapters, comprising over half the total, deal with the pattern of interaction between creditor and debtor nations in the period from 1870 to 1914. Two hypotheses are advanced: first, that the various national segments of the international economy all displayed long waves of the 'Kuznets' variety in their growth rates, and, secondly, that the cycles of debtor countries moved inversely to those of Britain. Underlying both hypotheses is the conviction that the great international migrations of this period were the primary motive force generating these fluctuations, although the author is by no means guilty of proposing an over-simplified causal relationship, since he devotes considerable space and ingenuity to teasing out the complex relationships between migration and other international flows, such as capital and commodities. That Professor Thomas is doing more than ploughing through an exhausted field of enquiry can be seen from the range of other researchers with whom he must take issue in order to sustain his hypotheses. These include those, such as Adelman,1 who dismiss Kuznets cycles as statistical illusions thrown up out of the averaging techniques used to eliminate short-term fluctuations from time series; those like Habakkuk2 and Saul3 who dispute the existence of systematic relationships between the long-term fluctuations in the various national economies; and finally 1 Irma Adelman, 'Long Cycle-Fact or Antifact', American Economic Review, June 1965. H. J. Habakkuk, 'Fluctuations in Housebuilding', Journal of Economic History, XXII, part 2 (1962). • S. B. Saul, 'House-building in England, 1890-1914', Economic History Review, 2nd series, XV, part 1 (1962).