(137) 
A NOTE ON THE RELATIONSHIP BETWEEN CLOUD AND 
SUNSHINE. 
By J. R. Sutton. 
On a cloudless day a perfect sunshiue recorder would show 100 per cent, 
of the optimum of sunshine. The ordinary sunshine recorders in use show 
less, because the record ceases when the sun is very low down, rising or setting. 
The recorder in use at Kenilworth (Kimberley) is of the Jordan hemi- 
cylinder photographic type. The paper used with it is not sensitive to the rays 
from a low sun ; and besides, it has not a good horizon, so that at the best it 
does not record more than 96 per cent, or so of what a perfect instrument would 
do. When the sun is high enough the instrument gives an excellent record, 
indicating with a considerable approach to exactitude, and with very little 
overlapping, the duration of the transit of the lower clouds over the sun. 
The higher, and thinner, clouds do not interrupt the sunshine very much. 
Now any trace of sunlight on the sensitised paper is counted as sunshine ; 
therefore if the sky were covered with thin cirrus all day, transmitting some 
sunshine, it would be possible to get this condition. 
Sunshine + Cloud = Total 
96 + 100 =: 196 per cent. 
Again suppose 40 per cent, of the sky to the southward to be covered 
with either cumulus or stratus all day, but no cloud to cross over the sun, 
then we should have — 
Sunshine + Cloud = Total 
96 + 40 = 136 per cent. 
These are extreme cases of course, but we may nevertheless expect in 
general the monthly and yearly sums of sunshine and cloud to exceed 96 per 
cent. ; and this proves to be the case. The following table shows the mean 
percentages of sunshine and cloud at Kenilworth for each month of the year, 
the cloudiness being deduced from daylight observations made at 800, 1100, 
1400 and 1700 each day. The period from which the means are obtained is 
the twenty years 1900 to 1919.* Annual means for Hong-Kong, Cordoba and 
Mauritius are added for purposes of comparison. The Hong-Kong 
cloudiness is that of the mean of hourly observations, day and night, that of 
Cordoba is deduced from the daylight hours only, that of Mauritius from 
observations at 600, 900, 1300, and 1500. 
* January, 1917, is omitted. 
14 
