126 SCIENCE IN AFRICA 



of the continent, it is to them that remarks about poverty of African 

 soils generally apply. In contrast, the soils of volcanic areas can be 

 very rich, and so also can the alluvial soils. Conditions vary so 

 much from place to place, however, that even these rough state- 

 ments have their exceptions; for instance in Uganda some of the 

 richest soils, especially those of Busoga, are developed from rocks 

 of the basement complex, and many examples could be found of 

 poor soils derived from volcanic rocks. 



The value of soil surveys must be considered in relation to the 

 other subjects treated in this volume. In order to obtain satis- 

 factory results from soil surveys the field conditions of water- 

 supply, vegetation, agricultural practice, numbers of stock, etc. 

 must be taken into the fullest account. Vegetation may be an 

 important indicator of soil types, and, as has been pointed out, 

 the plants themselves and their destruction by man affect the soil 

 very materially. For this reason many workers believe that all 

 such surveys should be based on the soil vegetation unit, a method 

 which is being developed particularly by Mr. G. G. T. Morison 

 at Oxford. 



A second and perhaps still more important consideration is 

 that surveys of soil, or of soil and vegetation combined, over the 

 vast areas of Africa must always be extremely rough, in view of the 

 limited resources available. The question arises therefore, to what 

 extent they can be of value in the allocation of land for different 

 uses. In some parts of Africa surveys over a wide area would be 

 of value to the European farmer in the choice of a farm, but in the 

 greater part of the continent the unit farm is that of the African 

 peasant, having an area of only a few acres, and differences in 

 soil from one plot to the next, though often pronounced, could 

 never be shown in the kind of survey contemplated. It would be 

 of value only if combined with agricultural experience obtained 

 from many small-scale experimental farms, while for opening up 

 new areas it might be claimed that the study of vegetation, especi- 

 ally of plant ecology, checked by a few soil analyses, is sufficient 

 for most purposes. Moreover, many other factors are involved in 

 the agricultural potentialities of a territory, particularly the atti- 

 tude of the native population, the cost of transport of crops and the 

 possibilities of irrigation. 



