NATURE 



485 



THURSDAY, SEPTEMBER 22, 1892. 



A ^-SENSATION THEORY OF VISION. 

 'Color Vision. By E. Hunt. (Glasgow : John Smith and 



Son, 1892.) 

 "T^HE author entitles his work, " An essay discussing 

 existing theories, explaining views hitherto incom- 

 pletely published, and comprising illustrated descriptions 

 of important neW experiments." We shall now proceed 

 to see how the promise conveyed by the title is fulfilled. 

 In the early pages of the book he makes the statement 

 that there are five colours which are distinct sensations, 

 viz., red, yellow, green, blue, and purple. The last, how- 

 ever, he is rather less certain about as conveying his 

 meaning, but finally adopts the name after explaining 

 what he designates by purple. These are the five 

 colours of a 5-colour theory which he propounds, all 

 other colours being mixtures. He has felt, however, that 

 it is no use to bring forward a theory unless he de- 

 molishes those other theories which block the way. His 

 examination of these last is chiefly confined to that of 

 Young, which accounts for colour-vision on the assump- 

 tion that there are only three colour sensations — a num- 

 ber which is a minimum when the fact is remembered 

 that all colours can be produced by a simple colour, or by 

 a mixture of two or three of the colours which are con- 

 sidered to be primary colours. His criticism of the 

 theory is mostly confined to a paper published by Clerk 

 Maxwell thirty years ago, and he scarcely refers to any 

 ■evidence in its support which has been brought to light 

 in more recent years. Mr. Hunt has entangled himself in 

 mixing up colour and colour sensations together, and has 

 forgotten that in Clerk Maxwell's papers three colours 

 were chosen empirically as approaching the colours 

 which are perceived when the three fundamental 

 sensations are stimulated. Later work has shown 

 that the colours thus shown are not representative 

 of the fundamental sensations. No one, for instance, 

 would say that any green in the spectrum was the colour 

 evoked by the stimulation of the green fundamental 

 sensation, for it is well known that, according to the 

 theory, at every part of this region of the spectrum all 

 three sensations are stimulated, and the nearest approach 

 that a retina possessing normal sensations could make 

 to perceiving this one sensation, would be when the 

 colour evoked was mixed with a percentage of white, 

 rendering the colour impure. We may here parenthetic- 

 ally remark that it is too late for Mr. Hunt to quarrel 

 with the designations of the colour constants, for they 

 are accepted terms. " Impure," for instance, may be an 

 objectionable term to apply to a colour when mixed with 

 white, but as what is meant by it is understood, it can 

 only be used in that sense. The position of the colour 

 which stimulates only the red fundamental sensation is 

 fairly well known, being near to the red lithium line. 

 The position of the colour which stimulates the violet 

 fundamental sensation is still not absolutely settled, but 

 it cannot be very far from the G line of the solar spec- 

 trum. Moreover, recent researches show that at the ex- 

 tremities of the spectrum only a red or a violet sensation 

 is stimulated, any change in colour observed being due 

 to a slight admixture of white light, which is derived 

 NO. I 195, VOL. 46] 



from the imperfect transparency of the prisms or re- 

 flecting surface of the grating. The colour, for instance 1 

 near G, when mixed with a small percentage of white 

 light, in excess of that already mixed with it, takes a 

 violet hue, a colour which is associated with the most 

 refrangible part of the spectrum. As the luminosity of 

 this part is very much less than that near G, the extra 

 percentage of white light required to form this hue is 

 always present. The colour of any ray of the spectrum 

 can be almost entirely freed from the white light derived 

 from the prism by placing another prism in the path 

 of such rays, after passing through a second slit. In 

 an eye-piece or on a screen, the ray will be seen as 

 a well-marked line lying in a faint continuous spectrum. 

 Again, the references to the sensations stimulated in the 

 various types of colour-blind people are not described in 

 any detail, though the evidence which is derived from an 

 examination of their vision is of the greatest importance 

 for the Young or any other theory. The author gives 

 most undue weight to colour diagrams. Colour triangles o r 

 circles are not intended to be the basis of a theory, but 

 simply as illustration of it. It is quite possible that Clerk 

 Maxwell's diagrams would not tally with those based on 

 Kcenig's observations, nor should they do so. In fact 

 a diagram may be drawn to illustrate any theory, as the 

 author himself has done to illustrate his own. 



In animadverting on Clerk- Maxwell's colour equations, 

 the author, it may be remarked, has himself made a mis- 

 take as regards certain reductions to be given to the 

 intensities of different colours. The equations are right 

 as they stand, when it is remembered that Maxwell chose 

 to adopt certain arbitrary units which he carried through- 

 out them all. 



The author in one place endeavours to prove the 

 superiority of a 5-colour theory over a 3- or 4-colour 

 theory, by narrating what is seen when a spectrum is 

 formed with — what comes to — a very wide slit to the 

 collimator. Practically he shows that this wide slit may 

 be supposed to be made up with narrow slits, and that 

 the spectrum formed when the wide slit is used is made 

 up by the overlapping of the spectrum formed from the 

 narrow slits. He then adds up the colours (or colour sen- 

 sations) as follows : — 



R— G— P 

 R— G— P 

 R— G— P 

 R— G— P 



R— y— W— W— b— P 

 With the 4-colour theory he has five rows for addition, and 

 with the 5-colour theory he has six rows. The last gives 

 as the result R— 2O— 3Y— 4b— W— W— 4pe— 3B— 2v— P, 

 where red, orange, yellow, lemon, white, peacock, blue, 

 violet, and purple are denoted by the letters used as the sum 

 of the additions. He remarks " that the results obtained 

 for the 3-colour and 4-colour theories do not agree with 

 what is actually seen with a prism." We can well believe 

 it ! But why he confines the colours to be added up to 

 the number of colours in the theory we are at a loss to 

 understand. It may be taken for granted that in employ- 

 ing this method of proving a theory, that theory which 

 annexes the greatest number of primary colours will give 

 results which are closer to what is seen than even the 

 5-colour theory. 



Y 



