16 PROF. B. STEWAKT ON AN INSTIWMENT VOR 



IV. On an Instrument for measuring the Direct Heat of the 

 Sun. Bv Professor Balfour Stewart, LL.D., F.R.S. 



Eeacl November i6th, 1875. 



The instrument generally employed for giving the radiant 

 energy of the sun's rays acts upon the following principle. 

 In the first jjlace, the instrument is sheltered from the sun, 

 but exposed to the clear sky, say for five minutes. Let 

 the heat so lost be termed r. Secondhj, the instrument is 

 turned to the sun for five minutes. Let the heat so gained 

 be termed E. Thirdly, the instrument, being now hotter 

 than it was in the first operation, is turned once more so 

 as to be exposed to the clear sky for five minutes while it 

 is shielded from the sun. Let the heat so lost be termed r'. 

 It thus appears that r denotes the heat lost by convection 

 and radiation united when the instrument, before being 

 heated by the sun, is exposed for five minutes to the clear 

 sky, while r' denotes the heat lost by these same two ope- 

 rations by a similar exposure after the instrument has been 

 heated by the sun ; and it is assumed that the heat lost 

 from these two causes during the time when the instrument 

 is being heated by the sun will be a mean between r and 

 r', and hence that the whole effect of the sun's rays will be 



r + r' 

 in reality II H . 



Now, although this assumption may in the average of a 

 great number of experiments represent the truth, yet in 

 many individual cases it may be far from being true. It 

 would therefore seem to be desirable to get rid of this un- 

 certainty by constructing an instrument in which we are 

 sure that the causes of variability are not allowed to operate. 



