THE CLIMATE OF LOURENQO INIAROUES. 3CK; 



has also to consider the irregular variations arising from the 

 passage of areas of high or low pressure. 



As an instance of one of these disturhances and its etfects on 

 the other meteorological elements, I shall mention the respective 

 hourly ranges of a depression on the 28th September, 1912, which 

 may be of very little interest to those who have studied similar 

 phenomena in any other locality in South Africa, but which will 

 serve to give evidence as to the analogy of the meteorological con- 

 ditions of Lourengo Marques with those of such places. 



The mere observation of Table No. V'l shows all the effects 

 on temperature, the direction and velocity of the wind, vapour 

 pressure, and relative humidity. Depressions, by causing NNW 

 and N hot zvinds, practically in every case have some influence on 

 evaporation, and render it excessive, thus requiring special atten- 

 tioii to avoid the shortage of water in the evaporimeter and 

 psychrometer. 



In this depression the duration of the hot zciiid was as usual ; 

 but the rapid decrease of temperature, when the wind, after a 

 short calm, veered to the south, was not very pronounced. This 

 calm generally lasts a maximum of 20 minutes, when the south 

 wind breaks out with violence, sometimes with a velocitv of 70, 

 80, and even 100 kilometres per hour, and brings about the so- 

 called " dust storms," which are less fre([uent here than in other 

 localities of South Africa. 



Depressions are not always preceded by cirrus clouds, and 

 when this occurs it is chiefly noticed two or three days before. 



A few hours before the wind veers to the south the horizon 

 frequently becomes cloudy, and the sky gradually obscured by 

 cirro-stratus, followed by cumulus and cumulo-nimbus. The 

 south wind is at times accompanied by rain, which is more fre- 

 quent on the second and third days, especially in the rainy season, 

 but it never rains when the wind comes from the north. 



(Ju other occasions it is accompanied by thunderstorms. The 

 ESE wind, which on this occasion was registered at XII on the 

 29th, prevailed for a good many hours, as is frequent, with vari- 

 able velocity, being sonietimes quite fresh, and it finally 

 veered to E and ENE. 



I had the opportunity of ascertaining, in 1912, that depres- 

 sions are felt in Durban from eight to 20 hours before they 

 occur at Lourenco Marques. They are generally more frefjuent 

 in May and from August to December, and last a shorter period 

 in the dry season than in the rainy season. This example seems 

 to prove in a general way the following conclusions which Dr. 

 Sutton arrived at in his paj^er " On the variation of the hourly 

 meteorological normals at Kimberley during the passage of a 

 barometric depression " : — That depressions greatly increase the 

 normal variation of pressure and temperature ; that they influence 

 cloudiness, especially in the case of low level clouds ; that they 

 originate rains, which are more abundant on the second day ; and, 

 finally, it seems to me that it can also be concluded that they 



