18 Scientific Proceedings, Royal Dublin Society. 
cool the faster as the air is drier; and this would cool the lower 
air by simple conduction. On the other hand, also, a damp air 
should cool the faster of itself, but check the cooling of the surface 
of the Earth, or of lower layers of air. The meteorological problem, 
then, is a composite one; and whatever be the true interpretation 
of Tyndall’s results, it by no means follows that it may be applied 
immediately, and without hesitation, to such observations as those 
furnished by General Strachey. 
Some years ago I tried to find out whether the cooling of the 
lower air at Kimberley by night could be regarded as depending 
upon the amount of moisture present, by comparing the fall of 
temperature between 8 p.m. and the succeeding minimum upon 
every clear night in one year with the absolute and relative 
humidity at 8 p.m. Kimberley has remarkably clear night skies, 
pretty well or quite one-half of the nights being absolutely free 
from cloud, so that a single year’s observations gave a good series. 
No certain connexion could be traced between the dew-point at 
8 p.m. and the succeeding fall of temperature ; but there seemed to 
be a uniform increase of range of temperature with a decreased 
relative humidity.!. The results could, however, only be looked 
upon as provisional, because, for one thing, the nights are not all 
of the same length, and, consequently, in the winter there is an 
interval of quite two hours longer for the temperature to fall in 
than there is in the summer. To a certain extent this, when 
allowed for, would show an even greater influence on the part of 
the relative humidity, since on the whole the percentage of satura- 
tion is highest when the nights are longest. However, it seemed 
worth while to examine the question afresh, using as a standard 
the fall of temperature in one hour corresponding to the relative 
and absolute humidity at the beginning of the hour. This I have 
been doing at spare intervals for some few years; and the more 
important results are submitted in the present paper. The matter 
needs ventilation, because if laboratory experiments have given 
contradictory results, meteorological opinion has been equally 
indefinite. But first, before calling attention to what I conceive 
1J. R. Sutton, ‘‘ Aqueous Vapour and Temperature,’’ Symons’s Met. Mag., Aug., 
1895. 
