434 



REPORT 1886. 



them like an engraving or a photograph to us. Such persons, whose 

 number is very small, have one elementary sensation only . I have met 

 one such a person, and the curve of elementary sensation Tirhich we 

 obtained from him is the curve H in fig. 3. Professor Donders has made 



„ o the same curve for 



another similar ia- 

 dividual and has ob- 

 tained almost iden- 

 tical results. So that 

 we may regard this 

 as a typical form for 

 monocnromatic co- 

 lour-systems. 



(b) There is an- 

 other very numerous 

 class of persons, 

 generally called co- 

 lour-bUnd, in whose 

 case we can divide 

 the whole spectrum into three parts. The parts near the ends we shall 

 call ' boundary regions ' and the parts between them the ' interval' For 

 these persons each boundary region has its own light, varying in intensity 

 only but not in colour, and the colour of any part of the interval they can 

 produce by the mixture of the light of two parts, one from each boundary 

 region. Here we must assume two elementary sensations, and the simplest 

 way to analyse the colour-system of such persons is to take the sensations 

 of the boundary regions as elementary sensations. On these assumptions 

 we have determined the curves of elementary sensations for this class of 

 persons. 



We have obtained three different curves. The curve K in fig. 3 was 

 obtained from every person but the other curve was diflerent with dif- 

 ferent persons. Some had the curve W,, others had the curve W,. So 

 that as far as our own observations go we must distinguish all the colour- 

 blind into two classes, and two classes only. A third very difierent class 

 of the colour-blind was found by Professors Holmgren and Donders. But 

 their analysis was a qualitative one only. For the purpose of simplifying 

 calculations, which I shall mention later on, the curves were drawn in 

 such a manner that the area bounded by them and the axis of abscissae 

 should be the same for each curve. 



So far we have two large classes of persons, and we have seen that the 

 small number of persons belonging to the first class possess one elemen- 

 tary sensation only, whereas the large number of persons belonging to the 

 second class possess two elementary sensations, and we have also seen 

 that this class must be subdivided into two divisions or types. 



(c) Now we pass over to the third very large class, which includes all 

 persons not belonging to any of the two preceding classes. We shall 

 presently see that if in the case of these persons we assume three elemen- 

 tary sensations, we shall be able to explain all colour-equations made by 

 them. 



It has been found by Lord Rayleigh and Professor Donders that the 

 persons of this class differ from each other considerably, and that this 

 class, too, must be divided into two subdivisions at least ; the persons 

 belonging to the first subdivision forming the great majority of this 



