318 FOURTH REPORT — 1834. 



When from the simple fact of absorption we proceed to con- 

 sider its law, as depending on the nature of the light, the diffi- 

 culties increase at every step. The intensity of the transmitted 

 light considered as a function of its refrangibility appears to be 

 subject to no law, or to a law so complicated as completely to 

 baffle all attempts to embrace it in an empirical rule. The max- 

 ima and minima are often actually numberless ; and the vari- 

 able does not reach them gradually, but by what seems to be an 

 abrupt violation of the law of continuity. These apparently ca- 

 pricious changes were observed long since by Dr. Young, in the 

 light transmitted through the common smalt-blue glass. Sir 

 David Brewster has recently directed his attention to the same 

 subject, and examined a great number of coloured bodies with 

 reference to their absorptive properties. He has found, in par- 

 ticular, that a very remarkable definite action is exercised upon 

 the rays of the spectrum by the green liquids obtained by extract- 

 ing the colouring-matter of the leaves of plants in alcohol; and 

 this action does not cease altogether even when the liquid has 

 become perfectly colourless *. But the absorbing properties of 

 nitrous acid gas, observed by the same author, are by far the 

 most remarkable ever noticed. When the light transmitted 

 through this gas is analysed by a prism, it is found that about 

 two thousand portions of the beam are stopped, and two 

 thousand dark spaces, or abrupt deficiencies of light, appear in 

 the spectrum. These increase in number and magnitude with 

 the temperature of the gas, until, by a sufficient elevation of 

 temperature, this rare body becomes perfectly opaque, and re- 

 fuses to transmit a single ray of the brightest sunshine f. 

 Prof. Miller and Prof. Daniell have found some analogous pro- 

 perties in other gases. In the spectrum produced by the light 

 transmitted through the vapours of bromine and iodine, more 

 than one hundred dark lines are visible, disposed at equal di- 

 stances J. 



To account for the selection of certain classes of rays by 

 coloured media, in the theory of emission, it seems necessary 

 to suppose that an attractive force is exerted at a distance be- 

 tween the molecules of the body and those of light, and that 

 the absolute value of this force varies with the colour. It does 

 not seem easy to reconcile these suppositions to the Newtonian 

 account of refraction ; and the difficulty is still further increased 

 when we proceed to apply the same considerations to the ab- 

 sorption of definite rays ; and introduce the hypothesis of 



• " On the Colours of Natural Bodies," Edin. Trans., vol. xii. 



f " On the Lines of the Solar Spectrum," &'c., Edin. Tran.t., vol. xii. 



J French transkttion of Herschel's Essay on Light, Supplement, p. 45.'). 



