STUDIES IN SPECIAL SENSE PHYSIOLOGY 403 



SECTION V. THE THREE-COLOUR HYPOTHESIS OF YOUNG 

 AS MODIFIED BY HELMHOLTX 



As we saw, in an earlier section, Newton's researches demon- 

 strating the (conceptually) complex nature of white light and the 

 physical substratum of chromatic stimuli as co-ordinated by the 

 undulatory hypothesis enable us to frame a coherent, if not 

 necessarily final account of the physical elements involved in 

 retinal stimulation. On the other hand, the specificity of physio- 

 logical response, which finds a not quite accurate expression in 

 Miiller's law, releases us from some of the difficulties attendant 

 upon the primitive theories of vision, the propounders of which 

 were hampered by their acceptance of the dogma of physico- 

 physiological identity. 



On so broad a basis, it is of course possible to erect 

 many, more or less, substantial edifices. In a famous passage, 

 R. L. Stevenson relates that he once overheard two persons 

 arguing. " What I advance," said one, " is true." " Yes," 

 replied the other, " but not the whole truth." " Sir," was the 

 retort, "there is no such thing as the whole truth." It is well 

 to bear that saying in mind when one considers the theories of 

 colour vision. 



In attempting to arrive at an adequate interpretation of any 

 experimental facts, more than one route is generally open. As a 

 rule, it does not much matter which path we choose so long as we 

 keep count of any assumptions made and avoid the introduction 

 of unnecessary steps. Let us start, then, from the experimental 

 laws of colour mixing, and see where we arrive by following the 

 most obvious route. We saw that, for most experimental pur- 

 poses, we could say that a sensation " produced by " a colour 

 stimulus could be matched by a sensation due to a stimulus obtained 

 by mixing not more than three lights. We saw that these three 

 lights did not coincide accurately with any spectral colours, but 

 that, if we admitted negative values into our colour equations, our 

 vision could be regarded as definitely trichromatic, even in terms 

 of known stimuli. In order, however, to avoid this, let us so 

 choose our stimuli that only positive values of each are employed. 

 This means that they must be so chosen that the colour table is 



