entitled "A Theory of the Composition of Colours," 425 



clear coloured fluids or solids, and the light transmitted through 

 or reflected from dyed or other coloured substances, must be 

 forced to admit, that, setting aside a comparatively small number 

 of cases in which the colour observed is referribJe to other causes, 

 the colours of natural bodies are due to absorption. The excep- 

 tions are colours due to fluorescence, as in the case of solutions 

 of quinine, or to regular chromatic reflexion, as in the case of 

 gold, copper, platino-cyanide of magnesium, murexide, &c., not 

 to mention such colours as those of the rainbow, &c., which 

 result from the general properties of bodies with regard to their 

 action on light, not from any speciality of the substance by which 

 the colours happen to be produced. The mode in which I con- 

 ceive absorption to operate in occasioning the colours observed 

 in dyed ribbons, flowers, coloured powders, &c., I have more fully 

 explained elsewhere*. Now absorption is best studied in clear 

 solids or solutions,. where it is not complicated by irregular re- 

 flexions or refractions. But when such media are studied by the 

 aid of a pure spectrum, there cannot be a moment's hesitation that 

 the colour of the transmitted light is due to the abstraction from 

 the incident white light of some of the component rays, as ex- 

 plained by Newton. The colour results, not from the light acted 

 on by the medium, but precisely from the portion left unafi*ected. 

 Hence its origin is celestial (supposing the sun to be the source 

 of the light employed), not terrestrial. But if the colours of 

 natural bodies arise from absorption, the origin of those colours 

 must be deemed celestial too. To make the origin of the green 

 colour of a leaf terrestrial, but that of the green colour of the 

 light transmitted through an alcoholic solution of the colouring 

 matter celestial, notwithstanding that the two greens agree in 

 their very remarkable prismatic composition, would be needlessly 

 and most capriciously to multiply the causes of natural phseno- 

 mena. The light which gives us the sensation of greenness 

 when we look at a leaf is, I conceive, no more terrestrial in its 

 origin than the sun's light reflected from a mirror is terrestrial, 

 as not retaining the direction which it had in travelling to us 

 from the sun. It is only in the phsenomenon of fluorescence, 

 and the closely allied phsenomenon of phosphorescence, that the 

 light emitted can be considered as new light having a terrestrial 

 origin. 



* Philosophical Transactions for 1852, p, 527. 



Phih Mag. S. 4 Vol. 13. No. 81. Dec. 1856. 2 F 



