90 Mr. Powell on Solar Light and Heat. [Aug. 



aperture necessary for the two effects would give the proportion 

 of light reflected by the disks. 



In the same way also I tried the distances from the eye at 

 which the disks became invisible in a room partially darkened. 

 Such trials, however, can never be susceptible of any accuracy 

 from the difficulty of saying precisely when the object is visible 

 or not. I, therefore, conceive it unnecessary to detail them 

 further than to mention, that the results uniformly gave a ratio 

 not very different from that above given, as the ratio of the 

 heating effects produced respectively by the proportions of light 

 which we suppose absorbed by the surfaces. 



It became necessary to seek for some other method of ascer- 

 taining this point ; and in this I succeeded as follows : 



(40.) Assuming that within ordinary limits, the heating effect 

 is precisely as the number of rays impinging, we may proceed 

 to a simple and, perhaps, sufficiently accurate method of esti- 

 mating the relative proportions of light absorbed by the black 

 and white surfaces employed on the thermometers from observ- 

 ing the quantities reflected. These data I obtained by placing 

 the photometer in the sun's light having the bulb protected by 

 a small screen from the direct rays, and, therefore, affected only 

 by the light reflected from a surface of paper, painted in one 

 instance with Indian ink, and in another with chalk ; and fixed 

 in contact with the outside of the glass case of the instrument, 

 on the side opposite the sun, and extending round two-thirds 

 the circumference of the cylinder. 



The following are the results of a set of experiments conducted 

 in this way : 



Hence we may take the proportions of light absorbed by the 

 two surfaces in the inverse of this ratio. 



This ratio may, however, possibly be rather too small, from 

 the circumstance that a small portion of light would be reflected 

 upon the bulb from the inner surface of the glass, which would 

 be the same in both cases. 



If on this consideration we take it = ~^, this ratio, it will 



