26 Transactions of the South African Philosophical Society. 



station, and that not under the control of the Meteorological Com- 

 mission, we know next to nothing of the wind system of any place in 

 Cape Colony. The statements that connect general wind changes 

 with precipitation on the central tableland are probably misleading 

 because they ignore the undetermined diurnal variation of the vane. 

 So far as the coast stations are concerned it is still an open question 

 as to the manner in which the wind-directions are modified by the 

 great bulk of the tableland. 



In the light of these drawbacks the numbers given in Table 30 

 must be discounted according to taste. They show, to the limits of 

 our knowledge, the frequency and average fall of rain corresponding 

 to particular wind-directions, at Durban, for the eighteen years 

 1885-1902, and at Kimberley for the nine years 1894-1902 (but 

 reduced to eighteen years for purposes of comparison). The 

 argument for Durban is the quantity of rain exceeding half an 

 inch in twenty-four years for any day, and the wind direction at 

 9 a.m. entered to the same day. The Durban rain-day ends at 

 3 p.m. The argument for Kimberley is the quantity of rain for the 

 twenty-four hours ending 8 p.m. on any day, and the wind-direction 

 at 8 a.m. on the same day. Thus the wind is observed at the middle 

 of the rain- day. 



It appears from the Table that a south-west wind brings more 

 rain to Durban than any other direction, the next most important 

 direction being the south. A considerable quantity, also, falls in 

 calms. Upon the whole, if we neglect calms, it may be affirmed 

 that the rain winds of Durban tend to blow nearly parallel with the 

 coast, the resultant direction being almost S.W. by S. The third 

 most important direction is west. It is remarkable that so much 

 should come from this point, which is directly from the considerable 

 mountains of Basutoland. 



At Kimberley the scheme is very simple. The resultant direction 

 is appreciably from N. by E., practically nothing coming from any 

 point having a south-westerly component." 



Thus two places not far from the same parallel of latitude, and in 

 the same system of summer rainfall, have their chief rain-bearing 

 winds from entirely opposite points Of the compass. This result,, 

 however, is really what might have been expected from the opposite 

 relationship to the variations of atmospheric pressure manifested by 

 the rains of the two places. It seems to follow that the direction of 

 the wind is only important in so far as it relates to the baric 



* If we observe the wind at 8 p.m. on the day before, i.e., at the beginning of the 

 rain-day, the resultant direction becomes nearly E. by N. This alteration is due to 

 the influence of the diurnal oscillation of the vane. 



