entitled *^A Theory of the Composition of Colours" 423 



the substance being given, only on the value of X/^ would cease 

 to be true. 



As to the second, the principal phsenomenon consists in this : 

 that when a beam of sunlight, condensed by a lens, is admitted 

 into certain perfectly clear {i. e. not muddy) media, the path of 

 the rays is marked by light, of different colours in different cases, 

 which emanates in all directions. As the real nature of this 

 remarkable phsenomenon was not at the time understood, and 

 the phsenomenon itself was confounded with the effects of mere 

 suspended particles, it is needless to discuss its possible bearing 

 on any theory of the sensation of colour under this head. 



As to the third, the new light emanating from the media 

 which possess the property in question is just like any other 

 light of the same prismatic composition. In its physical pro- 

 perties it retains no traces of its parentage, and its colour depends 

 simply upon its new refrangibility, having nothing to do with 

 that of the producing rays, nor to the circumstance of their 

 belonging to the visible or the invisible part of the spectrum. 

 Hence, in speculating on the sensation of colour, this phseno- 

 menon may be set aside as not bearing upon the question. I may 

 remark, however, that with regard to the sensation of colour, an 

 analogy has often struck me between the retina and a fluorescent 

 substance, or rather a mixture of three or more fluorescent sub- 

 stances : but this is only an analogy. 



It is not true, as Professor Challis seems to suppose (p. 332), 

 that absorption is always, or even generally, accompanied by 

 epipolic dispersion. Among the great variety of coloured me- 

 tallic solutions, I have hitherto found that property only in solu- 

 tions of salts of sesquioxide of uranium. I make this remark 

 merely by the way, to prevent misconception : I perfectly agree 

 with Professor Challis in believing that a ray of definite refran- 

 gibility is uncompounded ; in fact, it was my firm belief in that 

 doctrine which led me to make out the phsenomenon of the change 

 of refrangibihty of light. 



The superposition of two coloured glasses or ribbons by no 

 means gives the effect of the mixture of the two colours. Various 

 methods of mixing colours are enumerated by Mr. Maxwell at 

 the end of his paper, entitled " Experiments on Colour,^' &c., in 

 the twenty-first volume of the Edinburgh Transactions, p. 275. 

 The production of white by a mixture of blue and yellow is by 

 no means confined to prismatic blue and yellow, but takes place 

 just as well with the colours of coloured bodies. In making ex- 

 periments with the spectrum, in order to neutralize, when pos- 

 sible, a prismatic colour of given intensity by another prismatic 

 colour, so as to produce white, two points must be attended to : 

 the place of the second colour in the spectrum must be properly 



