J.— PSYCHOLOGY 217 



The test makes use of the fact that blue and violet and yellow appear 

 much brighter to the red-green colour blind than the colours red and 

 green. Pitt^^ questions this contention of Ishihara and asserts that 

 ' blue is not bright compared with red and green : from the hue dis- 

 crimination curves, however, it can be seen that blue is a colour having 

 a noticeably different hue from red and green, and it is for this reason 

 that blue is distinguishable for dichromates.' In some of the plates 

 only part of a numeral appears to a person who is colour blind, this being 

 dependent on the colours used. For example, an 8 may be the numeral 

 read by one with normal colour vision, but only part of it may be seen 

 by the colour blind, and he reads it as 3. Or a totally different numeral 

 may be seen by the colour blind from the arrangements of the spots 

 before him, and instead of seeing a 5 as the majority would do, he sees 

 a 2 standing out distinctly from the background. Further, in some 

 plates, the position is reversed as it were, and although the person with 

 normal colour vision sees nothing but a blur of colour, the colour blind 

 sees a numeral quite clearly. If one with normal colour vision looks at 

 these plates through a blue glass, the ' hidden ' numbers become visible. 



The resuhs from this test given to 42 colour blinds appear in Table I. 



The figures 12 are seen by all, normal and colour blind alike. The 

 8 appears as a 3 to the colour blind, the 6 as a 5, the 5 as a 2, and the 74 

 as 21 . The next four plates, containing the figures 2, 6, 5, 7, are generally 

 not visible to the colour blind, this depending on the degree of the defect. 

 The two next plates contain figures which are hidden to the normal eye 

 but are seen by the colour blind, a 5 and a 2 respectively. The two last 

 plates are intended to diagnose the defect as one of deuteranopia or of 

 protanopia. The deuteranopes read the 2 but fail to read the 6, the 

 protanopes read the 6, but fail to read the 2. Similarly with Plate XIII. 

 These plates seem to be diagnostic as will be seen from the table. 



The individuals forming the control group who were tested were found 

 to vary markedly in their colour discrimination. It seemed advisable 

 to divide them into two groups termed respectively N (normal) and N — 

 (showing greater deviation from the normal). The classification is of 

 necessity an arbitrary one because it is difficult to know where to draw 

 the dividing line ; in fact, the line of demarcation between colour blinds 

 and non-colour blinds may itself vary according to the purpose in hand. 

 In the report of the Colour Vision Requirements in the Royal Navy, 

 already referred to, it is stated of a certain test that if used alone it causes 

 extravagant rejection of candidates who may be fit even for the Seaman 

 Branch. This merely indicates that the standard for rejection can be 

 changed in accordance with the post to be filled. It is interesting to note 

 that three grades, so far as colour discrimination is concerned, are 

 suggested in this report. 



The colour blinds in the above table have been classified thus because 

 of their responses not to any single test, but to at least six tests, and in 

 some cases a spectrometric examination was also made. Similarly, in 

 the N and N — groups, their allocation to these classes is based on the 

 results obtained from the same battery of tests. 



*' Collins : Colour Blindness, p. 44. 



