248 Transactions of the South African Philosophical Society. 
will be found of service to the curves of either maximum, mean, or 
minimum, 
y being the temperature of any day required ; 
R the mean range of the curve ; 
T the mean value of the curve ; and 
«x the angular distance reckoned from the day whose mean tem- 
perature (maximum, mean, or minimum) is also the mean oi 
the year. 
For Durban, we have :— 
y = M = (48) sin « + 80°6 
ye eeee = 6 sin 2+ 712 
j= mM = (io) Sie 7s 
the angles x being reckoned from near Oct. 31st, Oct. 24th, and 
Oct. 16th, respectively. 
For Kimberley we have :— 
y= M = (147) sing 4+ 79°38 
y ae aan e 6s 
y =m = (12:1)sin a + 49°8 
the angles x being reckoned from near Sept. 20th, Sept. 28th, and 
Oct. 4th, respectively. 
The smooth curves derived from these formule are plotted on 
Fig. 1, from the calculated values of y found in Table 26. The 
retardation of the minimum curve for Durban is easily seen, and even 
more so the retardation of all the three curves for Kimberley. It is 
significant, and may indicate some interaction of climatic conditions © 
between the coast and interior that the maximum temperature at 
Durban lags as much behind the minimum as the minimum. tem- 
perature at Kimberley lags behind the maximum. ‘Taking the 
numbers from Table 26 for example, the greatest value of M at 
Durban is 85°4 on Jan. 31st, the greatest value of m being 69°-1 on 
Jan. 16th. But at Kimberley the greatest value of M is 94°-0 on 
Dec. 20th, the greatest value of m being 61°°9 on Jan. 4th. Again, 
determined by observation, AX the mean longitude of the sun computed from the 
first of Aries, for any day of the year, the mean temperature of which is y ” 
(Harvey, Art. ‘‘ Meteorology,” Ency. Metr., 1845). The formula, however, was not 
very successful, the errors in some places approaching 4° F. as compared with the 
mean daily temperatures of Stockholm derived from fifty years’ observations. 
