142 Report S.A.A. Advancement of Science. 



The shape of the Durban curve is here nearly the same as in 

 Table 7, where all the cold spells are included, but the epoch is 

 somewhat earlier. Also the average is rather greater, but that is 

 accounted for by the circumstance that cold winds from the sea are 

 not met with during the winter months. The East London and 

 Kimberley comparative values are elevated somewhat for the same 

 reason. While these peculiar conditions are in progress, pressure on 

 the table-land is moderately high, the averages at 8 a.m. at Kenil- 

 worth being : — 



Fourth Day ... 26" 1 29 inches. 

 Fifth Day ... 26-1 16 „ 



First Day ... 26" 108 inches. 

 Second Day ... 26"l69 „ 

 Third Day ... 2&160 „ 



and the wind on the table-land shifting from easterly components to 

 westerly. Thus the low temperatures at Durban coincide with the 

 following sides of moderate barometric crests* and the incidence 

 of westerly winds ; also the rise of temperature at Kimberley on the 

 fourth and fifth days of the penthemera corresponds to the falling 

 pressure. The average temperature on the table-land for the five 

 days is probably somewhat greater than the mean minimum tempera- 

 ture of the months in which they mostly occur. Now, there is one 

 very jemarkable characteristic of these cold sea wdnds, and that is 

 their dryness. Not one of them carries the quantity of moisture 

 proper to the time of year, the average for them all being fully 15 

 per cent, short of the mean, and, in spite of their low temperature, 

 they appear to be far from humid. One would scarcely expect a 

 cold, dry wind to blow over an ocean whose water is for its latitude 

 among the warmest in the world. Therefore, because of its dryness, 

 it is probably a re-entering land wind which has somewhere moved 

 outwards, but returned after describing a very short path over the 

 ocean. But how, then, is its temperature to be accounted for? If 

 it originate on the central table-land it must have started as a warm 

 wind, and by dynamic heating in its descent should become still 

 warmer, and thus reach the coast as a warm wind. The cause of 

 its low temperature is therefore still uncertain. In this connection 

 it would be interesting to know whether snow is lying on the crests 

 of the Drakensberg while these cold sea winds are blowing. If not, 

 then the only likely place of their origin seems to be in the main 

 westerly atmospheric drift above, from which they slant downwards 

 and inwards in response to the falling pressure. 



*The barometric changes at Durban are appreciably the same as those at 

 Kimberley at these times. 



