70 
Proceedings of Royal Society of Edinburgh, [jan. 17 , 
shown on PL III., another thermometer was hung up at the hack of 
each screen, with its bulb in the place where the bulb of the maximum 
thermometer is usually situated, and readings of these thermometers 
were taken at the same time as the others. These readings showed 
that the thermometers placed at the back of the screens indicated on 
this occasion a greater difference than those placed at the sides. The 
one placed at the back of the screen with open bottom was as much 
as 1°*4 higher than the one at the back of the closed screen. 
The readings given in the middle of PI. TIL, taken on the 25th 
September, show that the open bottom affects the readings not only 
on fine sunshiny days, but also on dull ones. On the 25th, the 
sky was uniformly covered all over with a dense mass of clouds, 
through which the sun was never visible. But though clouded, 
there was a good deal of heat reflected and radiated from the sky, 
the surfaces of all exposed bodies were hotter than the air, and the 
temperature of the grass rose as much as twelve degrees above the 
temperature of the atmosphere. At the beginning of this trial screen 
A was open, while B was closed. At 11 o’clock, when the first read- 
ings were taken, there was but little difference in the temperatures 
given by the two screens, but at 11.35 the open screen read 0°*5 
higher than the closed one. After the 11.35 reading, the bottom 
was taken out of B, and put into A. After which it will be seen 
that the lines connecting the temperatures cross each other, and B, 
which at first was closed and read lowest, now that it is open reads 
higher than the closed one by 0 o, 6. These readings show that even 
on a dull day there may be a considerable difference in the readings 
given by the two screens. 
The third set of observations at the foot of PL III. show a series 
made on the 7th October with the two screens A and B and with 
the new screen C. In these curves, as well as in those taken on the 
25th September, there is a curve marked G. This curve represents 
the temperature of the grass, and was taken by means of a thermo- 
meter placed with its bulb on the grass underneath the open screen. 
This curve is not drawn to the same scale as the others, as there 
would not be room for it ; the temperature rises so high it could 
not be represented within the limits of the plate. For this curve 
each space between the lines represents 1 degree, instead of 0 o, 2 of a 
degree, and as the disturbance produced by the hot grass will be in 
