394 Report of the Committee on Colour- Vision. 



In any examination it is essential that both the examiner and the 

 examinee should see the test colours at the same time, or at least 

 that the former should by some means know exactly what is 

 being shown. For a simultaneous view it is advisable that the 

 spectrum should be formed by a source of light at least as 

 bright as the lime-light, when it has to be thrown upon some white 

 reflecting surface. If the apparatus be so made that a patch 

 of monochromatic light from any part of the spectrum can 

 be thrown upon some one spot of the white receiving surface 

 the examination will become easy, more particularly if a patch of 

 white light can also be thrown separately or together with the 

 monochromatic light on the same spot, as this enables any dilution 

 of the pure spectrum colour to be effected, and gives a means of 

 detecting imposition. As already pointed out, every decidedly 

 colour-blind person sees some one part of the spectrum as — what 

 he calls — white. If the spectrum colours alone were thrown on 

 the screen, it is quite possible that the examinee might be taught 

 that when a colour which formed the patch appeared white to 

 him he ought to call it green or bluish-green, and thus detection 

 of the imposture would be difficult. 



But if the mode of testing be arranged as follows, the difficulty 

 would be overcome : — 



A patch of any coloured light should be thrown on the screen, 

 and the candidate asked to indicate if it was white. The colour 

 might then be diluted with white, and the question again asked. 

 A pure white patch might then be put on the screen and the 

 question repeated. The colours should be gradually changed until 

 his neutral point was approached. At this place the colour seen as 

 white would be mistaken for white, as the changes would be 

 made by dilution or by omitting the colour altogether. This 

 test involves no naming of a colour, but only a knowledge of 

 white. The discovery of a neutral point would infallibly indicate 

 that the candidate was colour-blind. 



Another simple test is to mix three spectrum colours to 

 form white, one of the rays being situated near the neutral 

 point of the red-green blind. The white would be the same to 

 the colour-blind as to the normal eye, and it would still remain 

 white to the colour-blind whether the colour at the neutral point 

 were increased or diminished. No amount of coaching would 

 enable the examinee to make constantly correct answers. 



By placing a bull's-eye of a lantern in this patch, and by 

 arranging that the three colours, the blue-green of the neutral 

 point, a green closer to the red, and a red, and also the white 

 should all have about the same luminosity, a further test in 

 imitation of signal lights could be carried out. 



The question as to the character of the colour-blindness need 

 not be investigated ; but if a patch of light from the extreme red 

 of the spectrum were thrown on the screen, and diluted slightly 

 with white light, the green-blind would see it coloured red or 

 yellow, whilst the red-blind would see only the white. 



