300 Proceedings of Royal Society of Edinburgh. [sess. 
These numbers are shown graphically in the Plate, fig. 2 (upper 
curve). 
Multiplying the numbers in the second column by twice the 
value of the solar constant, we get in absolute units the amount of 
heat supplied daily by the sun to unit area of the earth’s horizontal 
surface. According to Langley’s elaborate researches the value of 
the solar constant may be taken as 3 calories per square centh 
metre per minute. Hence, multiplying by 6 we find that there 
fall on each square centimetre of the earth’s surface, in the lati- 
tude of the Mediterranean, 950 units of heat during the mid- 
summer day. 
To compare with the data furnished by the Pola observations, 
which were made during the months of July, August and Sep- 
tember, we should however take, not the midsummer value, but 
the average value during these months. This average is less than 
850 units per day. But, further, the temperature observations 
were made in the morning and afternoon, say, at 8 a.m. and 
4 p.m., an interval of only eight hours. Evaluating the areas of 
the curves through an interval of four hours from culmination 
instead of through the half day, we get in place of the first four 
numbers in the small table above the values 136, 131, 120, 97. 
The mean of these is 121, giving a total supply during the eight 
hottest hours of the day of only 730 units of heat to each square 
centimetre of surface. 
Let us now consider the data which Dr Buchan has extracted 
from the Pola observations. They are contained in the following 
table, in which the first row gives the depths in metres, and the 
second the excess in Fahrenheit degrees of the afternoon tem- 
perature over the morning temperature. 
Depth, ... 0 1 2 5 10 20 30 50 75 
Temp. Diff. Fahr., 1°*5 1°*4 1°*3 1°'3 0°'9 0°*5 0°*3 -0°‘l 0° 
Constructing with these a curve, and estimating the area con- 
tained within the curve and the co-ordinate axes, we find, on 
reducing to Centigrade degrees, that the afternoon excess of 
temperature means an accumulation during the eight hours of 1460 
units of heat under each square centimetre of surface. And yet 
direct pyrheliometric measurements give us only 730 units of heat 
