ON THE THEORY OF EXCHANGES. 101 



Presuming therefore that the radiation of a particle is independent of its 

 distance from the surface, let us endeavour to realize what takes place in the 

 interior of a substance of indefinite thickness in all directions, and kept at a 

 constant temperature. Let us suppose that a stream of radiant heat is con- 

 stantly ilowing past a particle A in the direction of the next particle B. Now 

 since the radiation of B is by hypothesis equal to that of A, the absorption 

 of B must be equal to that of A. But Itt us notice what has happened to 

 the stream of heat in passing A. Part of it has been absorbed by A, but on 

 the other hand it has been recruited by the radiation of A, and this being 

 equal to the absorption, the stream of heat when it has passed A will be 

 found unaltered by its passage with regard to quantity. But it must also 

 remain unaltered with respect to quality, otherwise when it falls on B, the 

 amount absorbed by B will be different from that absorbed by A ; and hence 

 the radiation of B will be different from that of A, which is contrary to 

 hypothesis. The absorption of A is therefore equal to its radiation in quality 

 as well as in quantity ; or in other words, we have a separate equilibrium for 

 every description of heat. We have thus an explanation of the experimental 

 fact already alluded to, that a body is particularly opake with regard to that 

 heat which it radiates, since we see that a substance is predisposed to radiate 

 that description of heat which it absorbs. 



It is easy also to perceive why heat from a thick plate may be more easily 

 transmitted through a screen of the same nature as the source of heat, than 

 that from a thin plate, the reason being that the rays from the furthest por- 

 tion of the heated plate have already been sifted in their passage through the 

 plate, and hence that that portion of them which escapes is more easily able 

 to penetrate a screen of the same material. 



I have before alluded to a conclusion derived by Provostaye and Desains 

 from the theory of exchanges, that in an enclosure of constant temperature 

 the sum of the radiated and reflected heat from any portion of the walls is 

 equal to the lampblack radiation of that temperature. This is a case which 

 evidently comes under the scope of the law, which provides for a separate 

 equilibrium for every description of heat ; hence we may assert that the sum 

 of the radiated and reflected heat is in this case equal to the lampblack radi- 

 ation in quality as well as in quantity ; and we are thus also led to perceive 

 why opake bodies heated up to the same temperature always radiate the same 

 description of heat. 



We come now to the subject of light ; and since radiant light and heat 

 have been shown by Melloni, Forbes, and others to possess very many pro- 

 perties in common, it was of course only natural to suppose that facts analo- 

 gous to those mentioned should hold also with regard to light. One instance 

 will at once occuff in which this analogy is perfect. For, as all opake bodies 

 heated up to the same temperature radiate the same description of heat, so 

 also when their common temperature is still further increased, they acquire 

 a red heat, or a yellow heat, or a white heat simultaneously. 



The idea of applying these views to light had occurred independently to 

 Professor Kirchhoft' and myself; but, although Kirchhoff slightly preceded 

 me in publication, it will be convenient to defer the mention of his researches 

 till I come to the subject of lines in the spectrum. 



In February 1860, I communicated to the Royal Society of London a 

 paper in which certain properties of radiant light were investigated, similar 

 to those already treated of with respect to heat. 



In this paper it was mentioned that the amount of light radiated by 

 coloured glasses is in propojtion to their depth of colour, transparent glass 

 giving out very little light; also that the radiation from red glass has a 

 greenish tint, while that from green glass has a reddish tint. 



