COLOUR VISION AND COLOUR-VISION THEORIES 473 



The theory of vision and colour vision which I adopted as a 

 working hypothesis is as follows : 



A ray of light impinging on the retina liberates the visual purple 

 from the rods, and a photograph is formed. The rods are concerned 

 only with the formation and distribution of the visual purple, not 

 with the conveyance of light impulses to the brain. The ends of 

 the cones are stimulated through the photo-chemical decompo- 

 sition of the visual purple by light, and a visual impulse is set 

 up which is conveyed through the optic-nerve fibres to the brain. 

 The character of the stimulus and impulse differs according to 

 the wave-length of the light causing it. In the impulse itself, we 

 have the physiological basis of the sensation of light, and in the 

 quality of the impulse the physiological basis of the sensation of 

 colour. The impulse being conveyed along the optic-nerve to 

 the brain, stimulates the visual centre, causing a sensation of 

 light, and then, passing on to the colour-perceiving centre, causes 

 a sensation of colour. But though the impulses vary in character 

 according to the wave-length of the light causing them, the 

 retino-cerebral apparatus is not able to distinguish between the 

 character of adjacent stimuli, not being sufficiently developed for 

 the purpose. At most, seven distinct colours are seen, whilst 

 others see, in proportion to the development of their colour- 

 perceiving centres, only six, five, four, three, two, or none. 

 This causes colour-blindness, the person seeing only two or 

 three colours instead of the normal six, putting colours together 

 as alike which are seen by the normal-sighted to be different. In 

 the degree of colour-blindness just preceding total, only the 

 colours at the extremes of the spectrum are recognised as 

 different, the remainder of the spectrum appearing grey. 



It is obvious that this theory could not be true if the facts of 

 colour vision were as stated in the books of twenty-five years 

 ago. Apart from the relative functions of the rods and cones 

 if colour vision were a secondarily developed power of discrimin- 

 ation, then the following should be facts : 



1. There should be innumerable varieties of colour discrimin- 

 ation, which could be arranged in a series from total colour- 

 blindness to super-normal colour vision. 



2. The number of colours seen in the spectrum should depend 

 upon the development of colour discrimination; those colours 

 presenting the greatest physiological difference being the first 

 to be discriminated. 



