144 EFFECT OF VEGETATION ON THE RAINFALL OF S.A. 



SO that, unless this air goes up very quickly, it will arrive in it 

 state of considerable dryness. 



The rainfall there must be due to another cause, and that 

 is the considerable convectional rise we get there particularly 

 late in summer. Now, according to its latitude, the inland 

 plateau ought to be a region of descending currents of air, and 

 a very dry region such as the horse latitudes. There are two 

 reasons why that should not be the case in South Africa. In 

 the first place, the belts of high pressure at the horse latitudes 

 are comparatively shallow, and at an altitude of 4,000 feet have 

 disappeared entirely, so that at altitudes above that height the 

 pressure there is lower than that at higher latitudes at the same 

 height. The second reason, which is only operative in summer, 

 is the undue heating of the inland plateau, which causes a con- 

 vectional rise, and therefore a reduction in the pressure. That 

 is, monsoonal conditions are established. We may say that 

 generally the rainfall of the inland plateau is due to a convec- 

 tional rise there. Let us examine what the consequences of this 

 will be. When the air rises it cools at the rate of about 1.6° F. 

 for a rise of 300 feet if no heat is lost by radiation and if there 

 is no condensation. Consequently, if air continues to rise, a 

 point will be reached when its moisture will condense in the form 

 of a cloud. 



To calculate the height at which the cloud would appear, 

 supposing the rise of the air to continue, you would need to 

 know the temperature of the air when it commenced to rise and 

 its dew-point then. The temperature at which the cloud con- 

 denses is not, of course, that dew-point, because the water 

 vapour is also expanded by a convectional rise. The approxi- 

 mate formula meteorologists use is : substract the difference 

 between the temperature of the air at the earth's .surface and 

 the dew-point there, and multiply by 300 and divide by 1.27. 



When air is caused to rise by convection, it ceases to rise 

 as soon as its temperature has been lowered to that of the sur- 

 rounding air. As soon as that temperature is reached its vertical 

 velocity is zero, and it spreads out laterally. It is easy to see 

 that if the dew-point is comparatively low— that is, if the air 

 is comparatively dry to begin with — the whole convectional 

 circulation will take place below the cloud-forming level, and 

 consequently will give rise to no precipitation. If the convection 

 current should reach the cloud level, then moisture will be con- 

 densed, and in consequence of this the convection will be further 

 promoted, because — (a) the cloud will absorb the radiant heat 

 falling on it from both the sun and the earth, and this will raise 

 its temperature; (b) the latent heat is liberated on condensation, 

 and if the freezing-point is reached, further latent heat is 

 liberated. 



It is easy to see that if the convection once reaches the 

 cloud level, it will, as a rule, go much higher. In the case of a 

 thunder cloud it may reach a height of Unw or five miles. The 



