'210 A. G. Iloivard. — The Winter Storms of S. Africa, [Sept. 28, 



mornino- after the change at Cape Point to that at the other stations. 

 This has been put down as a simultaneous change, while in reality 

 ;the wind might have changed at Cape Point at a totc^Uy different 

 time to the other stations. Twenty-four hours is a long period to 

 pass over without an observation. 



Roughly speaking the following are the wind changes : — 



On fifteen days the wind changed at Cape Point before it did at 

 the other stations. 



On fifteen days the change was simultaneous. 



And on three days the change did not take place at Cape Point 

 i\\\ after the lower current had changed. These last three I am 

 inclined to consider as due to w^rong observations or to temporary 

 whirls. The remaining five instances are very doubtful as the wdnd 

 Avas blowing from the north-west all over the Cape Peninsula most 

 of these times ; hence no change seemed to take place. 



From the foregoing, the following inference may be drawn. 



When there is either an undue increase of pressure or heat at 

 Cape Point, a westerly depression may be looked for, especially 

 if the wind is blowing from the north there, no matter w^hat the 

 direction may be below. 



I have already mentioned that one of the forerunners of a storm 

 :is the appearance of a number of thread-like cirrus forms, called 

 by the Rev. Clement Ley, Cirro-filum, but noted down by me long 

 -before I ever heard of Ley's name as thread cirrus. 



This cirro-filum is formed at the junction of the two strata of the 

 upper current. The upper stratum of this current always flows from 

 some point between west and north-west, but the lower stratum 

 is more or less affected by the cyclonic disturbances, very often 

 travelling from south-west, but more aften travelling from a similar 

 direction to its upper component, although I have noticed it to travel 

 ,from the south, when any great disturbance was at hand or just passed. 



Now if a cyclone is advancing from the west or south-w^est, the 

 ;advance edge of the cone w^ill affect the motion of this current, the 

 general direction of which will be more westerly. It will also become 

 .more charged with moisture. This moisture will form into cumulus 

 clouds with perfectly level bases and sharply defined rounded tops. 

 The more moisture is added to these clouds the higher they will 

 become till, impinging on the upper current, their upper portions 

 will be drawn out into long threads by this upper current travelling 

 iuiore quickly. 



