1910.] Light Perception, to Colour Perception. 



463 



which has caused the non-perception of certain red rays has not caused the 

 dichromatism. A case of this kind will put a bright rose with a dark blue, 

 a bright red with a dark sjreen, designate a darlc red leatlier as black, and 

 find difficulty in seeing in a red light. On examining such a case it will be 

 found that the shortening of the red end of the spectrum may be abrupt 

 or more or less diffuse, that is, there may lie defective perception of red 

 over a nnich larger area. It will, however, be found that the light percep- 

 tion is absolutely normal immediately adjacent to the portion in which there 

 is defective light perception. For instance, the red may be shortened to 

 A. 680 ; at \ 070 the perception of red may be defective to about half the 

 normal, and at A, 660 it may be quite normal. If we now test such a case 

 with spectral colours from A 660 onwards to the violet end of the spectrum, 

 we find that their luminosity, i.e., differences of light and shade, is identical 

 with the normal. Thus the subtraction of the element which causes the non- 

 perception of certain rays cannot be responsible for the dichromic vision 

 which extends from X 660 to X 385. 



In estimating defects of light perception, colours should be directly com- 

 pared in order to ascertain their comparative luminosity. For instance, light 

 of A, 589 can be compared with light of A. 570 and A 535. A comparison 

 with white light gives rise to results which are very fallacious. Not only 

 have the missing red rays to be subtracted, but if the individual have a 

 spectrum which is lengthened at the violet end or is more sensitive to any 

 other rays than the normal, a co-efficient corresponding to these rays must 

 be added. Therefore if « represent the missing red rays and /3 the added 

 violet rays, the formula of white light as seen by the individual in question, 

 as compared with the normal, will be white —u + j3. 



Still more difficult to explain on any light perception theory are the cases 

 of so-called green-blindness. These, as I have shown, are simply cases of 

 dichromic vision without shortening of the spectrum, and, indeed, with no 

 defect in light perception in any part of the spectrum. If to such cases we 

 give colours to compare which differ only slightly in shade but are absolutely 

 different in hue, we find that their selection of the lightest or darkest corre- 

 sponds to that which would be made by a normal individual. A red and a 

 green, or an orange and a green, will be selected which to the normal eye 

 match exactly in shade.* 



During two years I examined every case of colour-blindness which came 



* These observations are in agreement with those of A. Konig, who also found that 

 there were considei^able variations in the luminosity curve of persons of normal colour 

 perception, whilst a so-called green-blind had almost exact!}- the same curve as himself 

 and a normal sighted woman. ' Beitrage zur Psychologie und Physiologic der Siunes- 

 organe,' 1891, p. 311. 



