of Edinburgh, Session 1884 - 85 . 
201 
the relative position of the curves remains unaltered ; hut if any change 
takes place with respect to clouds, the effect is at once noticeable. 
In the curves for June 9 (Plate VIII.), we have in the morning the 
various instruments reading very close together, the sky being covered 
with large quantities of cumulus cloud. After 1 1 hrs. we have a clear 
sky, and temperature rising steadily till noon, the gradual divergence 
of the curves being very well shown. Shortly after noon a maxi- 
mum occurs — the sky remaining clear — and hence we find the 
positions of the curves, with respect to each other, unaltered. At 
13 hrs. a minimum occurs, followed at 14 hrs. by another maximum ; 
but in this case we have cumulus clouds passing overhead and 
checking the radiation of heat to the two Stevenson boxes. These 
boxes have now the effect of cooling the air circulating through 
them, and hence the curve for the fan apparatus crosses the others, 
and becomes the highest. The curves remain reversed till 16 hrs., 
when the sky was again clear, and then fall back to their original 
positions, the old form of Stevenson being the last to do so, as the 
double top keeps the improved form from being heated as easily as 
the other. At 17 hrs. the curves show a minimum point after a con- 
siderable fall of temperature. Although a small quantity of cloud 
was observed overhead at this hour, its effect is scarcely perceptible, 
as it was of a cirrus type, very high up. After 1 8 hrs. a steady fall 
of temperature sets in, continuing till 21 hrs., when observations 
were stopped. Whenever the radiation from the boxes may be 
considered to have fairly commenced, we find the curves for the old 
and improved Stevenson screens changing places, the former being 
now lower than the latter, as the double top and louvred bottom 
cause the improved form to lose heat less rapidly. Both, however, 
as the process of cooling goes on, gradually approach the curve for 
the standard apparatus. Throughout these observations the wind 
was north-westerly, varying in force from 0 to 2 (Beaufort Scale). 
The effect of varying wind force is only shown in the general form 
of the curves — their relative positions remaining almost unchanged. 
This shows that although in a dead calm the radiation effects are 
enormously increased, they still require to be taken into account in 
accurate measurements of temperature, even when the wind-force is 
considerable. 
In the curves shown for 8th June one point is of especial interest, 
