424 Transactions of the South African Philosophical Society. 



Table 3 gives the monthly averages of earth temperature at 

 depths of 1, 2, 4, and 6 feet respectively, derived from regular 

 observations throughout ten years. Three readings a day are taken 

 of the thermometers at depths of 1 and 2 feet, although at least four 

 readings a day seem to be wanted ; at 4 feet and more there is no 

 diurnal variation, so that one reading a day is sufficient. It is quite 

 evident from Table 3 that ten years is too short a period to give true 

 monthly means in every case, and to eliminate all the casual irregu- 

 larities, for the Table makes the January 1-foot averages appear 

 lower than those of either December or February. Apart from that 

 it is very interesting to note how the ranges get less, and the phases 

 come later as the depth increases. 



TABLE 4. 



Diurnal Variation of Earth Temperature. 



1904. 



One foot. 



Two feet. 



Hour. 



April. 

 17 days. 



May. 

 13 days. 



April. 

 17 days. 



May. 

 13 days. 



VIII 



XI 







65-1 

 65-1 



66-8 

 68-8 

 69-2 

 68-5 



o 

 57-5 

 57-2 



59-1 

 61-1 

 61-6 

 61-0 



o 

 69-1 

 69-0 

 68 9 

 68-8 

 68-9 

 69-0 



o 

 62-8 

 62-7 



XIV 



XVII 



XX 



XXIII 



62-6 

 62-5 

 62-7 

 62-8 



Table 4, derived from some observations made from April 14 to 

 May 13, 1904, serves to show how the temperatures vary at shallow 

 depths throughout the day. Table 5, derived from hourly observa- 

 tions between 8 a.m. and 11 p.m. made upon sixteen fine days 

 scattered at intervals of a fortnight or three weeks throughout the 

 year 1899, gives somewhat more elaborate diurnal curves. They 

 refer, however, to the earlier (and somewhat inferior) positions of 

 the thermometers. 



