1919-20.] 
Thermometer Screens. 
179 
the hour of maximum temperature for the day ; and up to November it 
had no influence on the maximum recorded during the heat of the day. A 
low midwinter sun might possibly have some effect. 
To test the point here raised a larger rectangular screen was set up. 
The top part measured 30 by 21 inches (76 by 53 cm.) and the under part 
24 by 15 inches (61 by 38 cm.). The screen was placed with its length E. 
and W. to shut out the morning and evening rays. The small under 
screens were the same as in the other case. So far very few observations 
have been made ; but there is no evidence of any difference in the behaviour 
of the large and small screens.* 
Before concluding I would like to refer again to night temperatures 
taken on grass. The present haphazard method is of little value if the 
measurements made at different stations are to be taken as true indications 
of the intensity of the night radiation at these stations. The variables 
are far too numerous. Thus a thermometer placed on a small grass surface 
surrounded by bare ground reads higher than one placed on a large grass 
surface, owing to the drifting in of the warmer air over the bare ground. 
A thermometer placed on thin turf reads higher than one placed on thick, 
mossy turf, because it receives more heat from underneath. When placed on 
turf protected from wind, the thermometer reads lower than one more freely 
exposed to air currents, which mix the higher warmer air with the colder 
air at the grass level. Since it is impossible to have the same conditions 
at all stations, some less objectionable method is desirable. 
I have previously advocated the use of the radiation box placed at 
the height of the screen thermometers. It is not affected by variations in 
the amount of heat received from beneath the grass thermometers. It 
minimises the effect of the variations due to the extent of surface under 
grass and the variations due to wind. At night the R.B. always reads 
much lower than the grass thermometer. It must however be standardised. 
All radiation boxes must be of equal size. The ones I use are 36 cm. to 
the side, for the simple reason that tin plates can be easily got of that size. 
Any other convenient size might, however, be adopted. Again, the weight 
and kind of metal used for the top surface and for the tubes holding the 
thermometers must also be standardised, so that the thermal inertia may 
be a definite quantity ; and similar precautions must be taken as regard^ 
the depth of the box, the amount and nature of the non-conducting packing, 
the size and shape of the thermometers, and the like. These precautions 
were taken in the construction of a number of boxes used in this district;,, 
and all gave the same radiation temperatures. 
[* Further work as planned was interrupted by Dr Aitken’s death.] 
