1886.] illustrating the value of Cape Point as a Waiming Statio?i. 209 



visit the Cape. Now if further investigations prove this of anj 

 practical value, so that we shall be able to tell for a definite period of 

 time beforehand what nature of a storm to expect at even an approxi- 

 mate date, what an important warning station might not Cape Point 

 be converted into. The sun spots will tell to a day or two when the 

 storm is to be expected, and the observer at Cape Point couaI 

 telegraph atJeast, and often more than, 24 hours, before its arrival, that 

 it was close at hand. 



Having given a general idea of the subject, I now wish to phice 

 before you some dry facts and figures. I will confine myself to this 

 year alone. From the first of January to the end of August there 

 Avere 38 separate cyclones from the north-west, west, or south, purely 

 of the winter type. Taking the state of the barometer on the day 

 of the storm, and on 1, 2 and 3 days before, the following results 

 come out : — 



The barometer was high, 11 times 3 days before the storm. 



7 2 



1 ^ 1 



>> 7» '-^ 5? ^ 5? '5 11 J1 



and 17 times on the morning of the storm. 



Thus we find that during this period, twenty-three storms were 

 foretold bj an increase of pressure from one to three days before the 

 storm, while seven had an increase ouli/ on the morning of the storm. 

 The total of storms so foretold is thirty. This leaves eight storms 

 un-indicated by increased pressure within three days of their appear- 

 ance. 



But increased pressure is not the only thing to be guided by ; 

 increase of temperature is another indication, which I think the 

 following will pretty well demonstrate. 



Taking the same time as before we find that the maximumi 

 thermometer reading at 8 a.m. was 3 days before the storm 7 times. 



9 ^ 



^ ?> 5) iy 5? ^ ?? 



^ »> J5 5? J? it) ,, 



and on the day of the storm 12 ,, 

 Two of the maxima for three days before must be omitted as they 



were lower than the readings of the days before, but on these days 



very strong barometer indications are noticeable. 



With reference to the wind changes, one reading a day, and that 



at the very worst time, is hardly enough to construct anything like 



tables with. 



On several occasions the wind blew from a similar direction on the 



