164 Transactions of the Boyal Society of South Africa. 
A comparison between the temperatures observed at Mochudi and 
those of some other inland places is worth attention. In Table 2 this 
comparison is made for Kimberley, Elim, and Bulawayo. Evidently the 
curve of maximum temperature has its greater value earlier or later as 
the latitude is lower or higher. We see, moreover, how slowly the days 
cool off as the summer progresses into autumn inside the tropics, com- 
pared with what happens outside even so close up as Mochudi is. It 
is curious, however, that the turning-points of the curve of minimum 
temperature are practically unaffected by the latitude, albeit the range 
of the curve varies from place to place. 
It is not unlikely that the maximum temperatures at such places as 
Elim and Bulawayo are dependent not only upon the altitude of the 
sun, but also upon the rainfall of tropical Africa. In Northern Ehodesia 
the rainy season has settled in in earnest in November, and with it a 
mitigation of the fierce summer heat. Bulawayo feels the benefit of this 
mitigation very quickly ; and it may be that the period of time by which 
the highest point of the Mochudi curve of maximum temperature 
anticipates the December solstice represents the influence of the tropical 
rainy season upon the climate of the Bechuanaland Protectorate. 
The range of temperature at Mochudi is considerable. A temperature 
exceeding 100° is possible — one might say probable — in any of the seven 
months September to March, while freezing temperatures are common 
from May to September. This last month, so far, has shown the greatest 
extreme range of any, namely, a maximum of 108° in 1910 and a> 
minimum of 28° in 1912. 
In various previous papers it has been shown that the curves of mean 
maximum temperature for places as far apart as Durban, East London, 
and Kimberley show a pronounced dip about the middle of July. Cold 
days are to be expected in most years at this period ; and occasionally 
this cold penetrates at least as far north as Bulawayo ; while curiously 
enough the nights are not very much affected. Stewart finds the same 
condition at Cape Town, and at Colonies Plaats, near Graaff Eeinet. It 
is interesting, therefore, to find that it also appears to hold at Mochudi. 
In Table 3 will be found the mean daily maximum and minimum 
temperatures for Mochudi during July, showing quite plainly the average 
coldest day on the 16th. This cold wave may, then, be regarded as 
proved common to the whole of South Africa outside the tropics, so far 
as the maximum temperatures are concerned. There is a similar, though 
less important, cold spell affecting the nights, however, as much as, if 
not more than, the days, during the last week of May. 
