THE PRESIDENT of the Institute, Professor R. F. Peel has given us an excellent review of the geomorphological aspects of the deserts.2 Although I am myself more a geomorphologist than a human geographer, I shall try to balance his address by some words on the human problems of the deserts, trying to testify, at least for myself, and perhaps at my own expense, to the old and worthy theory of the unity of geography, and to collect together some observations I have had the opportunity to make in the various deserts I have studied. Everybody knows that arid and semi-arid areas result from a deficit of water and humidity balance and that, consequently, the vegetation cover is sparse and poor. That is why these areas can only be occupied, first, by pastoral nomadic people, seeking pasture over the large area occupied by a tribe whenever it has been raining and the soil remains somewhat wet; secondly, by farmers gathered in generally rather small areas, called oases, where water is available for irrigation. When the rain is less scanty and its annual or seasonal distribution more suitable, pasture becomes more abundant and dry cultivation possible, so that either pastors or farmers can adopt new and more diversified types of soil utilization and of life, ranging from nomadism, semi-nomadism and transhumance to dry extensive or irrigated intensive agriculture. In any case, every type of people has to maintain relations, both economic and social, with its neighbours, because they need one another, the more so when they are exclusively nomadic or, on the contrary, gardeners in the oases. So that, when they live either inside the desert or in the semi-arid margins, people must cross these empty areas. The pastoral people, the owners of the beasts of burden, have, or had, to set up caravans; they were carriers and traders, related to the merchants of the towns, the main harbours of the desert, where the oases were halting places and markets along the roads. This, in a few words, is, or rather was till recent times, the traditional life, somewhat different from one place to the other, but everywhere well adapted to a most difficult and determining natural environment. It has been so for a long time; for, in the Old World, desert occupation by either pastoral or cultivating people can be noticed, at least since the last humid period of the Holocene, between 5,500 and 2,500 B.P., and probably before that time. And we know that agriculture could be even older in Asia, and in South America as well. Nowadays, if the people leading a traditional life in arid and semi-arid areas play a valuable role since they occupy very wide surfaces, they do not play the same part in world economic activity. Their number appears reduced if we compare it to the population of other climatic regions, and they appear frequently to the geographer as some of those archaisms which are typical of 'underdeveloped' countries. Now, in our time, natural conditions can be mastered better than formerly, but the new techniques are mainly applied to areas where they have the most