Heat radiated by Rough and by Bright Surfaces. 89 



curves and their axis of abscissae, express the total quantities of 

 heat contained in the spectra of the bright and of the platinized 

 platinum. If these areas are calculated, we find that these 

 quantities of heat are in the proportion of 10 : 18. 



If we calculate the quantity of heat in the coloured portion 

 of the two spectra, as far as the limit of the red, as given in 

 fig. 6, we find that for the bright plate it amounts to 0*1 of 

 the total quantity ; but for the platinized plate it amounts to 

 only 0-068. 



The heat- spectra of both the strips are probably continuous. 

 It is nevertheless possible that they are broken at particular 

 points, and that the missing portions, on account of their small 

 width, elude detection by the thermo-pile. If, however, we pro- 

 ceed on the assumption that they are continuous (that is, that 

 throughout the whole space within which a rise of temperature 

 can be detected, rays of all wave-lengths are present without 

 any breaks), it follows that the heat emitted by the platinized 

 platinum contains no rays of different wave-lengths from those 

 contained in the heat radiated by the bright platinum ; for if 

 it were otherwise, the distance to which the spectra of the two 

 sources extend would be different ; and this is not the case. 



Since the absolute quantities of heat from both sources, which 

 a plate of alum allows to pass, are very nearly the same, it seems 

 probable that the wave-lengths transmitted by this substance 

 are those which lie towards the blue side of the spectrum, since 

 these are given out in nearly equal quantities by the bright and 

 by the platinized platinum. 



Hitherto I have not succeeded in directly proving that this is 

 the case — namely, that it is the rays belonging to the coloured 

 portion of the spectrum, or those lying nearest to it, which are 

 transmitted in the greatest proportion by alum ; for when a 

 plate of alum was interposed in the experiments with the rock- 

 salt prism, the deflections were so slight, in consequence of the 

 small quantity of heat which passed through, that no certain 

 conclusion could be drawn. 



It is, however, rendered probable, by Jamin and Masson's* 

 investigations into the distribution of heat in the solar spectrum, 

 that the rays from incandescent platinum, which would by pre- 

 ference be transmitted by a plate of alum, belong to the coloured 

 portion of the spectrum, just as in the case of solar heat. These 

 gentlemen in fact caused the rays of the solar spectrum to tra- 

 verse plates of rock-salt, glass, and alum, each of the same thick- 

 ness, and found that the rays belonging to the dark part of the 

 spectrum passed through glass in much smaller proportion than 

 they do through rock-salt, and that the quantity which passes 

 * Jamin's Cours de Physique, vol. li. p. 236. 



