Do the Mining Operations affect the Climate of Kimberley? 18 
The tendency of these numbers is to show that the lower the tem- 
perature of the air the nearer does it approach the temperature of 
the ground. And doubtless if the temperature of the mathematical 
surface of the ground could be ascertained the effect would be still 
more marked. 
It is obvious that if we wish to get a true idea of the relative tem- 
peratures of Kimberley and Kenilworth sources of error due to faults 
of exposure must, as far as possible, be allowed for. The method 
adopted after sundry trials was to consider the mean monthly error 
at both places to be a constant throughout any. assigned month. 
And thus the mean difference d between the Kimberley and Kenil- 
worth maximum temperatures for that month shall be a constant 
quantity, which, if applied with its proper sign to the Kenilworth 
maximum on any day in that month, will give the true relative 
difference for the same day between the maxima of the two places. 
The method answers equally well for the months before or after the 
change in the exposure of the Kenilworth thermometers. The pro- 
cess will be best understood by the annexed Table IV., representing 
a specimen month of readings in full. 
In this table— 
Column 1. Contains the dates. 
,,  &. The direction of the wind at 2 p.m. 
3. The maximum air temperatures of Kimberley (K). 
, 4. The maximum air temperatures of Kenilworth (K’). 
}. The differences between the maxima of Kimberley 
and Kenilworth (K-—K’) + or — according as 
KS <I 
,, 6. The corrected differences (cor. K — K’) obtained from 
column 5 by applying to each of the differences 
shown in column 5 a correction equal to the mean 
value of column 5. For example: the mean value 
of column 3 in the table is 67°7°, and of column 4 
is 68.2°. The mean difference K—K’ of these 
is —0-'d°, which is manifestly the mean value of 
column 5. Deducting this quantity from each of 
the numbers in column 5 we get column 6, 2.e. :-— 
— 0:3° —( -—0°5°) = + 0.2° 
+0°4°—(-—0°5°) = + 0-9° 
and so on. 
These calculations were made for each month, and the numbers 
from column 6 then arranged into groups according to the direction 
of the wind. The derived mean monthly differences are shown in 
Table V. 
