Heat radiated by Rough and by Bright Surfaces. 91 



now radiates less heat than before. And if a strip, already pla- 

 tinized on one side is platinized on the other side also, the radia- 

 tion of the first side is diminished. This can only be explained 

 by supposing that the strip, notwithstanding that it is entirely 

 surrounded by the flame, has, in consequence of its increased 

 radiation, a somewhat lower temperature at its outer surface. 

 There thus occurs a transformation of light into heat, in so far 

 as, in consequence of the greater radiation, the temperature of 

 the radiating surface falls, and thereby the intensity of the light 

 is diminished. 



If we fasten to a thin strip of platinum, which is bright on 

 both sides and heated to redness in a flame, a second thin strip 

 platinized on one side, in such a manner that the platinized side 

 is outwards, the double strip appears darker than before : the 

 change is chiefly apparent on the platinized side of the second 

 strip, but it is perceptible also on the outer bright surface of the 

 first strip. But if the second strip is now turned round with its 

 platinized side inwards, both sides of the double strip appear 

 luminous again. 



Comparison of the Spectra of Luminous and Non-luminous Flames. 



The great illuminating power which a non-luminous flame 

 acquires when a little soda or some other salt is put into it, sug- 

 gested the supposition that, in proportion as the intensity of the 

 light increases, so might also the heat be augmented. It is true 

 that I have already mentioned on a previous occasion*, that a 

 flame rendered luminous by soda does not radiate more heat 

 than when it is not luminous. This, however, might be caused 

 by the intensity of the heat-rays being diminished in the same 

 proportion as that of the luminous rays is increased. Whether 

 it is the case that the heat augments uniformly with the light, 

 must be ascertained by examining the heat-spectra of the same 

 flame in its luminous and non-luminous conditions. 



The flame of a Bunsen's burner radiates, however, in any case 

 not a large quantity of heat, and of this only a small proportion 

 is contained in the coloured part of the spectrum ; so that the 

 heating effect in the luminous portion further than the red was 

 so small that it could not be perceived. This, however, does not 

 prevent the soda-flame being compared with the non-lumi- 

 nous flame ; for if the radiation of heat increases at the same rate 

 as the radiation of light, or at anything near the same rate, the 

 heating effect of the soda-flame must increase to such an extent 

 in the yellow as to become perceptible. But, as I have found, 

 this is not the case. In order further to ascertain the relation 



* Monatsbericht d. Berlin. Akad. 1864, p. 594. [Phil. Mag. S. 4. vol. 

 xxix. p. 59.] 



