272 Proceedings of Royal Society of Edinburgh. [sess. 
No doubt, we also have similar currents at continental stations, 
and they are distinctly shown by different temperature curves at 
different places, according as the direction of the wind is from a 
hotter or cold quarter. But in this case it is quite impossible to 
pronounce a distinct law about such currents, the probability of 
their carrying air from colder places being generally as great as the 
contrary, in a theoretical point of view. Nevertheless, in every 
instance where a distinct convection current influences the problem, 
the mathematical explanations just given will suffice to represent 
the peculiarities of the daily curve. 
Our deductions, so far, would be correct on the supposition of a 
perfectly clear sky, — that is to say, in the absence of visible clouds, 
as well as of invisible aqueous vapour in the atmosphere. Unfor- 
tunately, we are assured by actinometric observers, that even on 
so-called bright and cloudless days the solar radiation is subject to 
considerable variations, most probably caused by the continuous 
changes in the amount of atmospheric moisture. The observations 
of Crova at Montpellier show that, for instance, in the summer the 
intensity of solar heat at noon is about a third of its whole amount 
smaller than it ought theoretically to be, even on days which the 
eye judged as exceedingly bright. Now, Langley , on whose labours 
these considerations, as we have said, are principally founded, had 
chosen a place for his observations, the atmosphere of which was 
very much drier than that of the average places we have to deal 
with ; so that the above formula for solar radiation, if applied to 
ordinary conditions, has to be considerably corrected on account of 
the discrepancies just mentioned. 
The best method of studying this remarkable influence of the 
daily changes of the moisture seemed to be to carefully investigate 
the daily curve of cloudiness, which had the following result : — 
In the winter there is a distinct decrease of clouds from the 
morning to the afternoon, the minimum being at about 3 o’clock 
in the evening ; in the summer, on the contrary, we have a maxi- 
mum of clouds very near the culmination of the sun ; while in 
spring and autumn, we observe two epochs, when there is no 
apparent daily change at all. These three different features, which 
are very characteristic, and which we believe to give a correct 
account of the changes of aqueous vapour at the same time, 
correspond in a remarkable manner to synchronous variations of 
