550 REPOKT — 1902. 



nerve could convey all waves of light. The facts of colour vision can only be 

 satisfactorily explained on the assumption that each optic nerve fibre does convey 

 impulses corresponding to all waves of light. It occurred to me that if there 

 were a transforming apparatus in the eye we could explain the facts. The tele- 

 phone shows how this may be accomplished in the case of sound. I saw that the 

 retina was constructed in a manner theoretically perfect from this point of view. 

 The percipient layer of the retina is made up of two kinds of elements, the rods 

 and the cones. The portion of the retina corresponding to the central portion of 

 the field of vision contains only cones. External to this tspot the cones are 

 arranged with one or more rings of rods round them, the single ring being round 

 those cones which are nearest to the central portion. In the rods there is a rose- 

 coloured suKstance, the visual purple, which is very seiisitive to liglit. This photo- 

 chemical substance is found exclusively in the rods. I assumed that light falling 

 upon the eye liberated the visual purple from the rods, just as heat would an 

 ointment, and a photograph is formed. The decomposition of the visual purple 

 by light chemically stimulates the ends of the cones, and a visual impulse is set 

 up which is conveyed through the optic nerve tibres to the brain. I have examined 

 the retinas of several monkeys after they had been kept in a dark room, and 

 found that the visual purple was to be seen in the yellow spot, but situated 

 between, and not in, the cones. This view gives a reason for a great many facts 

 which were previously inexplicable. For instance, a bright light may fall upon 

 the fovea (the centre of the yellow spot) without producing any sensation, and a 

 perceptible interval elapses before we are able to see with the yellow spot, after 

 the remainder of the retina, the fovea being the last point to convey a sensation 

 of light. The first fact we should expect, the cones being insensitive to light ; 

 the second corresponds to the diffusion into the jellow spot of the visual purple. 

 All the facts of colour mixing, contrast, and after-images can be explained by the 

 hypothesis that the visual purple is the visual substance. A positive rose-coloured 

 after-image can be obtained after white light or any spectral colour. The ordinary 

 explanation of this — namely, that the action of the hypothetical red and violet 

 fibres persists longer than those for green — cannot be true, because it is exceedingly 

 difficult to obtain this after-image after spectral red, and very easy to see it after 

 green. It would be against the whole principle of the theory that the red fibres 

 should be excited most efficiently by green. ]>ut if we assume that the visual 

 purple is the visual substance, then we have an easy explanation of the facts. 



The fibres of the optic nerve pass to the visual centre. I have assumed that 

 the visual centre transmits to the mind impressions of white light, and that by it 

 objects are seen monochromatically, as in a photograph. The visual centre is, 

 therefore, acted upon by impulses caused by all rays of light, the colour-perceiving 

 centre being concerned with the quality of the impulse within the power of 

 perceiving ditlerences possessed by that centre, or portions of tliat centre. 



I will now apply this theory to colour-blindness, and it will be seen that it 

 gives a simple explanation of the facts. 



Cases of colour-blindness may be divided into two classes, which are quite 

 separate and distinct from each other, though both may be present in the same 

 person. In the first class there is light as well as colour loss. In the second 

 class the perception of light is the same as the normal sighted but there is a 

 defect in the perception of colour. In the first class certain rays are either not 

 perceived at all or very imperfectly. Both these classes are represented by 

 analogous conditions in the perception of sounds. The first class of the colour- 

 blind is represented by those who are unable to hear very high or very low notes. 

 The second class of the colour-blind is represented by those who possess what is 

 commonly called a defective musical ear. Colour-blind individuals belonging to 

 this class can be arranged in a series. At one end of this series are the normal 

 sighted, and at the other tlie totally colour-blind. The colours appear at the 

 points of greatest difierence, and I have classified the colour-blind in accordance 

 with the number of colours which they see in the spectrum. If tlie normal 

 sighted be designated hexachromic, those who see five colours may be called 

 pentachromic ; those who see four, tetrachromic ; those who see three, trichromic ; 



