KADI ANT HEAT AND ITS RELATIONS. 93 



conclusion, our duty is to bring that conclusion to an 

 experimental test. In this way we fortify our science. 

 For the purpose of testing our conclusion regarding 

 the influence of the gum, I take two powders presenting 

 the same physical appearance ; one of them is a com- 

 pound of mercury, and the other a compound of lead. 

 On two surfaces of a cube are spread these bright red 

 powders, without varnish of any kind. Filling the 

 cube with boiling water, and determining the radiation 

 from the two surfaces, one of them is found to emit 

 thirty-nine units of heat, while the other emits seventy- 

 four. This, surely, is a great difference. Here, how- 

 ever, is a second cube, having two of its surfaces coated 

 with the same powders, the only difference being that 

 the powders are laid on by means of a transparent gum. 

 Both surfaces are now absolutely alike in radiative 

 power. Both of them emit somewhat more than was 

 emitted by either of the unvarnished powders, simply 

 because the gum employed is a better radiator than 

 either of them. Excluding all varnish, and comparing 

 white with white, vast differences are found ; comparing 

 black with black, they are also different ; and when 

 black and white are compared, in some cases the black 

 radiates far more than the white, while in other cases 

 the white radiates far more than the black. Deter- 

 mining, moreover, the absorptive power of those 

 powders, it is found to go hand-in-hand with their 

 radiative power. The good radiator is a good absorber, 

 and the bad radiator is a bad absorber. From all this 

 it is evident that as regards the radiation and absorp- 

 tion of non-luminous heat, colour teaches us nothing ; 

 and that even as regards the radiation of the sun, 

 consisting as it does mainly of non-luminous rays, 

 conclusions as to the influence of colour may be alto- 

 gether delusive. This is the strict scientific upshot of 



