On some Meteorological Conditions controlling Nocturnal Badiation. 387 
namely, that the rise in the value of the difference of temperature (where 
there should be a fall) is caused by a partial overriding of the relative- 
humidity effect by the varying velocity of the wind. 
Table 7 supports this view. It gives for each grade of relative 
humidity the average temperature differences, smoothed in threes, cor- 
responding to grades of two miles an hour in the wind. The first column 
contains the average value of the relative humidity ; at the head of each 
column after the first is the velocity of the wind for that column. 
TABLE 7. 
Radiation Temperature Differences corresponding to Grades of 
Relative Humidity and Wind Movement. 
Mean Relative 
Humidity. 
Wind in Miles an Hour. 
0-1-9. 
2-3-9. 
4-5-9. 
6-7-9. 
8-9-9. 
10 & more. 
% 
0 
o 
o 
o 
o 
o 
92 
61 
5-7 
4-2 
84 
6-2 
5-5 
4-4 
2-7 
77 
6-7 
5-9 
5-0 
4-1 
5-0 
72 
7-4 
6-9 
5-6 
4-4 
5-0 
67 
7-8 
7-5 
6-3 
3-7 
5-0 
4-5 
62 
7-8 
7-8 
6-8 
3-9 
4-1 
4-0 
57 
7-6 
8-0 
6-9 
4-6 
4-3 
3-8 
52 
7-4 
7-9 
7-0 
5-8 
4-3 
4-0 
47 
7-5 
7-9 
6-9 
6-1 
5-2 
4-1 
42 
7-8 
7-7 
7-2 
7-2 
6-0 
4-1 
37 
8-1 
8-0 
7-4 
7-4 
5-9 
4-3 
30 
8-3 
8-5 
7-7 
7-2 
5-2 
5-0 
Thus as the mean relative humidity falls from 92 per cent, to 30 per 
cent, the mean value of D rises from 6"*1 to 8°' 3, while the velocity of the 
wind is between zero and 1-9 miles an hour ; from 4°"2 to 7°"7 while the 
velocity is between 4 and 5*9 miles an hour ; and so on. Again, examin- 
ing the lines, we see that for any assigned relative humidity the values of 
D fall as the wind gains strength. For a mean wind velocity of 8 miles 
an hour or more the relative-humidity effect is practically lost.* 
As a final test for any possible dew-point effect Table 8 has been 
constructed. It shows for each grade of relative humidity the tem- 
perature differences corresponding to rising grades of two degrees each 
* The wind at Kimberley is gusty as a rule. Consequently its velocity for the hour 
preceding an observation cannot be expected to closely represent the wind factor at the 
time of observation. 
