An Introduction to the Study of South African Rainfall. 13 



amounts of cloud and the corresponding computed mean altitudes 

 of the lower plane of condensation. In Table 23 this is done in 

 two ways : First, for each month of the five years 1898-1902 the. 

 mean monthly amount of cloud is tabulated in order of magnitude,, 

 followed by the corresponding computed altitude of the plane of 

 condensation ; next, the process is reversed, the computed altitudes- 

 being tabulated in order of magnitude followed by the corresponding 

 amounts of cloud. For example, taking the mean results for January,, 

 we have in illustration of the first scheme : — 



Year. 



Order. 





Cloud. 



Plane. 



January 1898 



A 



59 



per 



cent. 



2,838 feet 



1899 



B 



45 







5,538 



1900 



C 



38 







5,331 



1902 



D 



31 







5,123 



1901 



E 



30 







6,438 



And in illustration of the second- 



Year. 



Order. 



Plane. 



Cloud. 



January 1898 



F 



2,838 feet 



59 per cent 



1902 



G 



5,123 



\ 31 



1900 



H 



5,331 



38 



1899 



J 



5,538 



45 



1901 



K 



6,438 



30 



The double process is necessary because a particular month may 

 have the same cloud percentage, but different cloud altitudes, in 

 different years. 



It appears from Table 23 that, generally speaking, the average 

 altitude of the first plane of condensation will be greater as the 

 cloudiness of the sky is less ; and conversely. A moment's 

 thought, and a glance at the Table, will show that this is not. 

 contradictory to the other result that the plane of condensation is 

 lower in winter (when the percentage of cloud is small) than it is 

 in summer (when the percentage of cloud is relatively great). The 

 statement, however, is not a law in the sense that a great percentage 

 of cloud necessarily implies a low cloud level. For obviously the 

 prevailing cloud of any assigned month may be cirrus, or it may be 

 stratus, or what not. But in the long run, when sufficient observa- 

 tions have been accumulated to give the averages their chance, the 

 statement may be exact enough. As we might expect, then, in 

 very dry weather the sky will be clearer than when there is much 



