275 
1895 - 96 .] Dr J. Halm on the Temperature of the Air. 
We learn from this curve that the maximum of da occurs 
exactly at the time of the summer solstice. We should have a 
corresponding minimum in December, but it would be unsafe to 
follow out our theory for the winter months in these northern 
latitudes without further careful investigations into the new con- 
ditions introduced by the presence of snow and ice on the surface 
of the ground. Nevertheless, we have found at all the various 
places considered very distinct indications of that minimum between 
December and January. There are two points in our curve, March 
15 and October 1, where da becomes zero; and it is indeed a very 
remarkable fact that the daily curves of temperature at both 
terms are fully represented by our first simple equation— 
t — t Q + a cos 4> cos 8 cos (£-v). 
Finally, we come to the most important part of our question, the 
constant of solar radiation a. Theory requires that a must be a 
constant at all seasons (except winter) and for all places at a (com- 
paratively small) distance from the sea under a given state of 
cloudiness, both the thermal capacity and the absorptive power of 
the soil being supposed to have the average values. The amount 
of heat radiated by the sun being, of course, a function of the 
absolute state of cloudiness, I considered, by means of a great 
number of observations selected over the whole scale of clouds 
ordinarily used, the question, by what function of the cloudiness 
the sun’s heat may be represented, arriving at the remarkable 
result that for all the stations considered the simple linear function 
a n = (i - Pn)*o ; P = 0-080 
