proceedings of the canadian institute. i07 



'* Recent Views on Colour." By Albert H. Abbott, B.A. 



(Read January 29, 1898.) 



The colour problem has three aspects : 



I. The physical problem, which investigates that energy in nature which is 

 especially connected with our sensation of light and colour. 



II. The physiological problem, which investigates the processes in the eye and 

 its accessories as the organ of vision. 



III. The psychological problem, which investigates our sensations of colour, or 

 colour as it is experienced. The question here is: What are the mental facts of light 

 and colour, and on what conditions do they depend ? 



The first " recent " view on colour discussed was the emphasis which has been 

 laid upon this psychological colour problem with the rise of scientific or experimental 

 psychology. Both of the other aspects, the physical and physiological, must refer 

 continually to the facts of colour which scientific psychology discovers or estab- 

 lishes, as the final test of the adequacy of their theories. The facts of all sciences are 

 (^riniarily facts for psychology (i.e., psychic or mental facts), and secondarily, facts for 

 these sciences, and hence, the conclusions and theories of all sciences must be judged 

 by their faithfulness to the facts of experience. 



The second view o« colour discussed was a modification to the ordinarily accepted 

 physical theory of colour, suggested by Dr. Kirschmann. The ordinary theory 

 contends that colour is an explicit function of the wave length. There is a difficulty, 

 however, in this view which is raised from the fact that no one has ever seen light or 

 a colour of only one ivave length, and, therefore, that, could we get light of one wave 

 length, there is no guarantee at all that we should see it coloured. Colour of one 

 wave length is a purely hypothetical conception; at every point on a spectrum there is 

 always a superposition or interaction of wave lengths. A slit inAnitcly small would, 

 so far as mathematics are concerned, give the pure spectral colours which advocates 

 of this theory demand : but, on the other hand, a plate bearing a slit which is 

 infinitely narrow would be for us an opaque dbject. Colour as seen in the spectrum 

 must actually be projected by use of a slit of finite width, and, therefore, it must 

 always be produced by the superposition or interaction of wave lengths. 



This contention is based directly on psychical considerations, viz., whether we 

 see colour or not. To contend that that alone would be a pure colour which is to 

 be produced under circumstances which would prevent us seeing either light or 

 colour seems to overlook the fact that it is our sensations of colour which make any 

 science of optics possible, and surely they must be the deciding factor in such a 

 matter to the last. 



A second line of objection to the theory that colour is an explicit function of 

 the wave length arises in connection with the discussion regarding purple, i.e.. the 

 colour which would form the transition from violet to red. This colour is not 

 present in the ordinary spectrum, and from this it has been concluded that purple 

 is not a pure but a mixed colour, and as such it is not a constituent of white 

 light at all. 



An experiment was shown which seems to have some bearing on the question. 

 By very simple means two spectra were thrown upon a screen together, parallel 

 and in close juxtaposition to each other. The one was the ordinary spectrum, con- 

 sisting of red. orange, yellow, green, blue, violet, and the second was an "inverted" 

 spectrum, consisting of blue, violet, purple, red, orange, yellow. (Note. — Purple 

 is absent from the first, green is absent from the second.)* This "inverted" spectrum 



* The "inverted " spectrum was first shown in this connection in a lecture given by Dr. Kirschmann before 

 the Mathematical and Physical Society of the University of Toronto. The objection may be raised that the 

 colours in the inverted spectrum are not as " pure " as those in the ordinary spectrum, but this is met by the fact 

 that, as sensations, the colours are quite as cure and brilliant as the ordinary spectral colours. The right of these 

 colours to rank physically the same as the latter colours was further demonstrated in the above-mentioned lec- 

 ture by the fact that both spectra show interference bands equally well. 



