CASE OF COLOUR BLINDNESS. 503 



If less white than 110 was taken, the grey was too dark; if more than 130 was 

 taken, the grey was too light. Thus green can be matched with grey. Yet it should 

 be noted that Mr A. always said that he was not satisfied with this match because the 

 grey disc had a red tinge. This red tinge was quite obvious to the normal eye and was 

 apparently due to the contrast with the strong green colour. It appeared whether a 

 small green disc was placed on large black and white sectors or small black and white 

 sectors were placed on a large green disc, i.e., whether the green colour was at the centre, 

 or the circumference, of the compound disc. The conclusion is that green, though it 

 does not produce a distinct colour impression, yet can, by contrast, give rise to the 

 complementary colour sensation. This point is considered farther below. 



In the same way, with the blue disc, the equations 



360B = 132W+228Bk) n 



360B = 112W+248Bki 



were obtained — the match being quite satisfactory. 

 Again, a mixture of blue and green gave 



180B + 180G = 127W+233Bk) m 



180 B + 180 G = 120 W+240 Bk ) 



A mixture of red and green gave 



174E + 186G=:91W + 269Bk| IV 



174 E + 186 G = 87 W+273 Bk ) 



a sufficient amount of green being taken to entirely destroy the red. A similar experi- 

 ment with blue and red gave 



75E+285B = 114W + 246Bk) v 



75R+285B= 95W+265BkJ' 



And, when the blue in this match was replaced by the equivalent amounts of black and 

 white, the match was at once destroyed, the mixture being called red by Mr A., and 

 appearing distinctly red to the normal eye. This proves that the blue, which was matched 

 with grey, possesses, in common with green, though to a less extent, the property of 

 neutralising red. 



By repeating observations on these several colour equations, and taking averages for 

 each, a series of self-consistent equations might be found. But it is easy, without the 

 labour involved in that method, to deduce a self-consistent set of equations lying well 

 within the experimental limits. 



