PHILOSOPHICAL SOCIETY OF WASHINGTON. 93 



edge, the value of p would be ■>. In the case of such a disk, in 

 which the radiation is from one side only, which would be approxi- 

 mately so in the case of such a disk with the opposite side of pol- 

 ished silver, the value of p would be unity. 



The amount of heat, K, received from the sun through the atmos- 

 phere at the earth's surface is usually expressed by 



(2) K = Ape 



In which 

 A = the heat received from the sun on a unit of surface at the top 



of the atmosphere ; 

 s == the secant of the zenith distance of the sun ; 

 p = a constant for all zenith distances, but differing in different 



states of the atmosphere, but always less than unity. 



In the case of a static equilibrium of temperature, which was 

 the only case considered, 8h vanishes, and the preceding equations, 

 (1) and (2), give 



(3) pApe = B(j* — p/) 



This equation expresses the condition which determines the static 

 temperature, t, of a body, and it is seen that this depends.upon the 

 solar constant A; the form of the body,' upon which the value of r 

 depends; upon the value of p, or the state of the atmosphere; 

 upon the zenith distance, which determines s ; upon the radiating 

 constant, B ; and upon the shade temperature, t\ 



Putting for the unit of heat the amount required to raise the 

 temperature of a cubic centimetre or gram of water one degree 

 centigrade, and the square centimetre, second, and degree centi- 

 grade, for the units of surface, time, and temperature, respectively, 

 the value of B was determined by the author, from the experiments 

 of Mr. J. P. Nichol on the rate of cooling of a blackened copper 

 ball in vacuum, surrounded by an enclosure of blackened surface, 

 (Proc. Royal Soc. Edin., 1869-70, p. 207,) to be .01808. This 

 value was considered more reliable than that of Pouillet from the 

 experiments of Dulong and Petit, since the latter were made on the 

 rate of cooling of mercury in a glass bulb, and the results had to 

 be reduced to those which would have been obtained with a black- 

 ened surface ; and the value of the radiating power, /, for glass, 

 which was used in this reduction, Pouillet states, was somewhat 

 hypothetical, and so it left some doubt with regard to the true value 



