Survey of the Colour Vision of 1000 Students. 199 



When we come to sum up the results of our investigation, 

 we are faced at the outset with the difficulty that there is 

 apparently no satisfactory definition of a homogeneous 

 population. Perhaps the best definition is a distribution 

 with a frequency curve fitting one of Pearson's formulas. 

 If we take this definition and reject the fourteen men on the 

 left of fig. 5, all our distributions may be regarded as homo- 

 geneous, as variations about a mean: all the blue-green 

 colour-blindness and a part of the red-green colour-blindness 

 is thus an outlying part of a homogeneous distribution. But 

 we cannot think of a simple analogue for the fourteen rejected 

 observers. 



The striking fact brought out by the investigation is that 

 fig. 5 is a very good Gaussian distribution, if the tail of four- 

 teen observers on the left is rejected. This seems to suggest a 

 difference in kind between the colour blindness of the rejected 

 observers and the colour blindness of those in the outlying- 

 portion of the main curve. 



We have not yet followed out all the consequences of the 

 investigation, but from the standpoint of Hering's theory 

 figs. 6 and 7 show that none of the men were wanting in the 

 blue-yellow or the black-white processes. The fourteen 

 rejected observers must consequently be deficient in the red- 

 green process. From the standpoint of the Young-Helmholtz 

 theory fig. 6 shows that none of the men lacked the blue or 

 green sensations. The fourteen rejected observers must con- 

 sequently lack the red sensation. But 14 out of 835 is a 

 smaller percentage than these theories require, and the distri- 

 bution of the 14 does not suggest a well-defined class. Also 

 the bases of these theories deserve reconsideration in view of 

 the large role normal variation has now been shown to play 

 in the matter. 



Perhaps it is not superfluous to say a word here on the 

 position of those who do not accept the Young-Helmholtz 

 theory, as their views are often misunderstood. The fact 

 that normal colour vision is trichromatic was really discovered 

 by Newton (' Opticks,' Book I. Prop. VI. Problem II.) ^ It 

 is implied in the somewhat fanciful colour diagram he gives, 

 though not expressly stated in words. His diagram makes 

 it clear, that almost all colours can be produced by mixing a 

 red, a green, and a blue. This fact is accepted by all theorists, 

 and forms the basis of the three-colour process of photography. 

 It was not until two centuries later that Helmholtz took the 

 additional step of assuming that there were three elementary 

 colour sensations, and that colour blindness was accounted 

 for by the absence of one or more of these sensations. In 



