1887 .] Mr John Aitken on Thermometer Screens. 
79 
a condition of weather in which there is no wind, then in summer 
the effect of heating the air on the ground is to cause ascending 
currents of hot air. There is thus an influence at work in summer 
tending to keep the air near the ground at nearly the same tempera- 
ture at all places within a considerable distance ; while the effect of 
cooling the air in winter is precisely the opposite, as the cooled air 
tends to sink down into hollows and remain there, where its 
temperature is further reduced hy contact with radiating surfaces. 
The cooling of the air in winter, therefore, in the absence of winds, 
will tend to give rise to differences of temperature, which may 
amount to some degrees within a limited area. 
From the above we can see that on a day on which there is no 
wind, the temperature of the air in an exposed situation, and where 
the sun is shining, may he some degrees warmer than the air in a 
cup-shaped hollow, into which the sun’s rays do not penetrate, 
provided always that the sky overhead is clear and radiation into 
space strong. Take the case of the 18th November : on that day 
the temperature of the air was 45°, while the grass that was not 
exposed to the sunshine was under 32°. We can easily imagine 
conditions in which the air resting on grass at 32° in a hollow into 
which the sun does not shine might easily be cooled some degrees 
below the temperature of the air above. These observations are 
suggested by the peculiar condition of matters reported from 
different parts of London on the 24th November. In one part 
of London the maximum temperature was 40°, wdiile at another 
it did not rise above 32°, a difference of 8°, which might be 
accounted for in the way above stated, as there was no wind at the 
time. 
A record having been kept for some time of the maximum 
temperatures recorded by the three screens, the readings taken 
from the 23rd September to the 25th November are marked off at 
the top of PI. IY. The temperatures were not - taken every day, 
but every observation taken is marked. Most of these are from 
the readings of the maximum thermometers, but on those days on 
which the screens were under trial, the readings taken at short 
intervals were used, and the maximum recorded readings selected. 
As it was impossible within the limits of the plate to record the 
actual temperatures, only the differences between the readings are 
