An outline is given of some aspects of the contemporary dynamics of large cities which are deserving of more study and research by geographers. There appear to have been cycles of movement of manufacturers into and out of large cities. These have not interrupted urban growth but have affected its form. The dynamism of large urban growth at present is mainly due to the growth of the 'quaternary' occupations. The rise of these quaternary occupations con- solidates urban concentration at selected locations owing to the interweaving between activities that use quaternary personnel and information, and also to such corollaries as the swelling flow of transients to large cities and the increasing role of collective rituals. In conclusion, an 'Alexandrine model' of complementarity between large cities is offered as an aid to understanding the structure of modern urbanization. The location of manufactures and urban growth My first point concerns the relationship between manufacturing production and the dynamics of large cities. It has generally been accepted that since the Industrial Revolution the concentration and growth of manufacturing plants has been the prime mover of modern urban dynamics. Only in recent decades has it been acknowledged that manufactures were moving out of the large cities, in any case out of their central sections, towards the periphery of the agglomeration, to the suburbs or even beyond to small towns and to rural places at convenient crossroads. In the larger cities of the Western world the out-migration of manufacturing production has become a charac? teristic of our time, although in the Third World large plants still cluster in and around substantial agglomerations. In the more advanced countries, especially on the two sides of the North Atlantic, experts have used the terminology of the pre-industrial, industrial, and, now, post-industrial city. I am not satisfied with this tripartite sequence of the historic evolution. The relationship between manufacturing produc? tion and the dynamics of large cities appears to have followed, at least in the West and for the last millennium, a pattern that may be described as cyclical. Economic historians have shown that in Flanders, northern Italy and elsewhere, the cities concentrated manufacturing work within their walls from the tenth to the twelfth centuries. This trend of agglomerating crafts and guilds in urban places was almost complete by the thirteenth century. From the fourteenth on and until the latter part of the seventeenth century, an outward migration of manufacturing work developed, scattering production of manufactured goods in villages and throughout the countryside. This dispersal is explained by the economic historians as being a result of the increased burden of costs and regulations brought about by concentra? tion in the larger cities, under the heavy hand of local rulers, guilds and corporations (Braudel, 1967). The escape extra muros was aided by the improvement of security in open country. This out-migration did not cause cities to lose their momentum: they continued to direct, finance and manage manufacturing production, controlling ?? Professor Jean Gottmann is Professor of Geography in the University of Oxford. This paper was given at a joint meeting held with the Institute of British Geographers and