EFFFXT OF VEGETATION ON THE RAINFALL OF S.A. I43 



much as the more elevated one. Now, if the rain gauge was 

 a perfectly accurate measurer of rainfall, you would expect the 

 upper gauge to register slightly more than the lower gauge, as 

 the drops get slightly less in size as they fall because of evapora- 

 tion, and the level of rain-bearing clouds is quite high, even 

 in winter. In fact, we should expect the amount of rainfall 

 to increase with elevation till the level of the rain-bearing 

 clouds was reached. 



In the above experiment, the reason for the discrepancy is to 

 to be found in the increase in wind velocity with elevation. This 

 increases rapidly as you rise from the ground. 



Now there is another circumstance besides the height of 

 the rain gauge above the ground, which would influence its read- 

 ings, and that is the character o^f the rainfall. Large raindrops 

 would not be deflected very much by the eddies, because in the 

 first place they fall through the eddies quicker than smaller 

 drops, and in the second place they expose less surface in pro- 

 portion to their weight than the smaller drops, to be acted on 

 by the eddies. The rain gauge, on this account, would have 

 very little error in registering the heavy rainfall, but would 

 considerably underestimate the drizzling, soaking rains. An ex- 

 treme case is that of snow : here the average density is very 

 small, and it is impossible to valuate the amount of the snowfall 

 by means of the rain gauge : 



For these reasons the writer rejects the rain gauge as an 

 instrument for determining whether the amount of rainfall in 

 South x\frica has been undergoing an alteration — as the charac- 

 ter of the rainfall has undoubtedly altered over large tracts of 

 the country — a large portion of the rainfall being torrential 

 downpours where formerly gentle soaking rains were the rule. 



If we except the Western Province, all portions of South 

 Africa have a considerable proportion of rainfall in summer. 

 Since the rainfall of the Western Province has not altered in 

 character during historical times, this portion will not be con- 

 sidered at all in this communication. 



In summer the region in Central Africa at the equator, and 

 just south of it, is enormously heated, and a convectional rise 

 of the air there takes place on an enormous scale, and heavy 

 rainfall occurs in consequence. This convection and the large 

 humidity causes the pressure to be very low there, and conse- 

 quently air rushes from the south, where the pressure is much 

 greater, to take its place. Owing to the rotation of the earth, 

 this air is deflected towards the west, and forms the south-east 

 trade wind. Unfortunately for South Africa, the relative hu- 

 midity of this wind is not very high. The inland ])lateau of 

 South Africa is unduly heated in summer, and consequently the 

 wind, when forced to rise when leaving the ocean to enter it, 

 does not have its temperature lowered perceptibly, consequently, 

 as a general rule, no moisture is precipitated. . 



The temperature of the inland plateau is probably warmer 

 than the moisture-laden air which starts from the Indian Ocean, 



