448 



Mr, II. C. Sorby on 



[June 19, 



having many properties in common, but differing in the position of their 

 absorption-bands, are decomposed when exposed to the light in the order 

 in which the bands lie towards the red end, that I cannot but think there 

 is some connexion between these facts. When the spectra are quite 

 unlike (for example when one has two and the other three bands, or 

 two at a different interval, or no bands at all) there is no such connexion. 

 Those which seem to approach most closely and uniformly to the law 

 belong to simple groups, and give spectra more or less closely analogous, 

 as it were, to the same musical chord in different keys ; and this leads me 

 to think it probable that, if we had such a perfect series of substances, 

 the rate at which they would be decomposed by the action of light would 

 vary directly as the length of the waves absorbed, or as would express 

 the facts equally well in many cases, as though the power of decom- 

 position varied directly as the width of the spectrum over which the 

 absorption extends, i.e. as the number of different rays absorbed. This 

 would also, in great measure, correspond to what occurs when the same 

 substance is dissolved in liquids which cause the absorption to extend 

 more and more towards the red end. Whether this be a general law or 

 only a common fact, the results are of very great importance in connexion 

 with my present subject. The effect of this difference in the rate of 

 change will be better understood by means of the following rough illus- 

 tration. Suppose that there is a mixture of three analogous colouring- 

 matters, a, b, and c, the absorption-bands of a being nearest to the red and 

 those of c to the blue end of the spectrum, and suppose that on exposure 

 to the sun, in the time required to destroy one part of c, two parts of b 

 and three of a are decomposed, it is obvious that the original proportion 

 between the different substances would soon be very materially changed. 

 This would alter the ratio in which they would be decomposed by further 

 action until a sort of equilibrium was established ; so that, whatever 

 might have been the original proportion, there would at length be a sort 

 of ultimate vanishing ratio, depending on the nature of the substance 

 and of the solvent and on the* kind of light and its intensity. I find this 

 fluxional manner of viewing such facts often very useful, and cannot but 

 think that it might be introduced with advantage into similar subjects. 

 However, for the sake of simplicity, supposing that the assumed rate of 

 decomposition remained constant, and adopting arbitrary units, which 

 only express the ratios for each substance, we should have some such 

 result as the following : — ■ 



Original state 10a -f 10b + 10c 



f la + 86 + 9c 



Successive states, after more and ^ a ^ c 

 more exposure to the sun. j 2b + 6c 



^ 5c 



If instead of an analogous substance («) we had an entirely different 



