226 Transactions of the South African Philosophical Society. 



that the friction of a dust-laden air may account for some, at any 

 rate, of the 20° we want. 



The surface temperatures of the sea have been observed regularly 

 at East London since 1897. They fall upon a curve very similar to 

 that of the maximum temperatures of the air, with the same long- 

 drawn-out minimum from July to September. * Monthly averages 

 are : — 



Jan. . 



Feb.. 



Mar. 



April 



May., 



June 



67° 



66 



64 



64 



63 



62 



July. 



Aug. 



Sept. 



Oct.. 



Nov. 



Dec. 



61° 



61 



61 



63 



65 



66 



Year 



64 



The rapid rise during the late spring and early summer is very 

 marked. Hydrometer observations show a slight — very slight — 

 increase in the density of the sea from July to September. 



The cloud averages show two maxima in the course of the year, 

 in February and in October, agreeing in this respect with Kimberley, 

 and, in fact, with the greater part of the whole country lying between 

 Natal and Pondoland on the east and Namaqualand on the west. 

 The minima occur at the solstices. Neither of these turning-points 

 show any special agreement with the monthly averages of rainfall, 

 one reason possibly being that clouds of different types prevail at 

 different seasons. Monthly averages are : — 





East London. 



Kimberley. 





East London. 



Kimberley. 





8 a.m. 



Mean. 



8 a.m. 



Mean. 



Jan 



53% 



56 



50 



45 



39 



35 



40% 



42 



35 



32 



21 



17 



July 



Ausf 



34% 



40 



47 



56 



55 



52 



17% 

 17 



Feb 



Mar 



May 



June 



Sept 



Oct 



Nov 



Dec 



25 

 31 

 26 



37 





47 



28 



In the East London Eegister the species of cloud does not appear 



Cf. 0. Stewart, Science in South Africa, pp. 25, 48, 51. 



