64 
Proceedings of Royal Society of Edinburgh, [jan. 17 , 
cold. It will be seen from this that the radiant heat which will 
affect the thermometer will be that absorbed by the two horizontal 
surfaces of F, and the heat absorbed at the outer surfaces of F, and 
conducted through the wood into the inner chamber surrounding 
the thermometer bulb. The amount of heat received from the first 
of these sources will be very small, as the horizontal surfaces of F 
receive most of their heat at second hand, that is, after being 
radiated and reflected from C and D. The amount received by 
direct radiation from without is very small, and is represented 
by the angle KIH in the figure ; from which it will be seen that 
these surfaces have a verv limited exposure to outside objects, and 
the amount conducted through F is probably very small. It will be 
seen from the sketch that the annular piece F has its outer surface 
groove-shaped. The object of this groove is to prevent the air heated 
on it from flowing into the inner chamber, the groove conducting it 
round the outside. How far this groove is, necessary, or to what 
extent it improves the readings, I cannot say, nor can I say 
whether the double bottom is necessary, or whether the passage 
between F and D might not be abolished, and the screen and 
sunshade thus made smaller. It was thought advisable to take all 
these precautions, as there was not time this season to work 
upwards from the simpler to the more complicated ; these points 
were therefore left for future consideration. 
It may be mentioned that this screen has been tried without the 
annular piece F, and it was found to work very well, but did not 
give quite such low readings as with it in, and its inertia was also 
much less without the ring. When the piece F is out, the bulb is 
freely exposed to radiation from all surrounding objects ; but as the 
space between C and D can then be reduced, the bulb does not get 
a very wide view of the outside. Its readings without the annular 
piece were very much more correct than the Stevenson screen. This 
at first may seem strange, as the bulb of the thermometer in it is 
much mere freely exposed to radiation than the one in the Steven- 
son screen. The reason for its lower readings would appear to be 
that it is exposed to the radiation from trees and other objects 
high up, freely exposed to the wind, and therefore cooler ; whereas 
the bulb in the Stevenson screen is exposed to the highly heated 
grass. This new screen without the ring has not been tried in a 
