METEOROLOGY 1 1 7 



desiccation is in progress in Nigeria at present is unproved and 

 demands the closest scientific investigation before the fact can be 

 accepted.' The real criterion must lie in records of rainfall, but 

 these cover so short a period of years in the area in question, that 

 they contribute no evidence. Whether or not the desert is advan- 

 cing as a whole, all are agreed that the effect of man has been very 

 serious in this danger zone, and that the establishment of forest 

 'shelter belts' would be of great advantage. This subject is con- 

 sidered in connection with forestry in Chapter VII. 



Some authorities are inclined to attribute the apparent desicca- 

 tion to the cyclical changes of climate discussed on page 113. In 

 French West Africa for instance, where the eleven-year cycle is so 

 evident in the rainfall records of Dakar, M. Welter considers that 

 the population tends to move southward during the years of 

 reduced rainfall, but does not subsequently return to the north. 



As regards the meteorological evidence, careful examination of 

 long rainfall records fails to reveal much sign of diminished pre- 

 cipitation in recent years. The best data are from South Africa, 

 given by Schumann and Thompson (1934). Their conclusions are 

 so important that a few points are quoted: '3. The greater part of 

 the Union enjoyed a period of plentiful rains around the year 

 1890. This possibly accounts for the popular idea that South 

 Africa is drying up. 6. Over the last forty to fifty years the annual 

 rainfall in certain parts of South Africa shows a more or less defi- 

 nite, though irregular downward trend. There is no proof, how- 

 ever, of any permanent diminution, and periods of plentiful rains 

 may confidently be expected in the future. 9. The depopulation 

 of the Midland Districts, to which attention is drawn by the 

 Drought Investigation Commission (1923), appears to be funda- 

 mentally attributable to the downward trend of the rainfall in this 

 area over the last forty to fifty years.' There is a widespread belief 

 in the drought-stricken regions that reckless deforestation and burn- 

 ing of grassland have tended to make a larger percentage of the 

 rain come in destructive storms instead of in well-distributed gentle 

 downpours. 



The principal human activities concerned in the process are the 

 destruction of forests and natural vegetation in association with 

 unenlightened methods of farming. Studies of the influence of 



