A THEORY OF COLOUR VISION 



38? 



developed to discriminate between the characters of adjacent 

 curves. Two curves must be widely different in shape and 

 position before the colour-perceiving centre can detect the 

 difference. 



We can consequently regard each energy curve as equiva- 

 lent, as far as the colour-perceiving centre in the brain is 

 concerned, to one, two, or three rectangles. For example, 

 figs. 8, 9, and io represent a sodium yellow, a thallium green, 

 and a white light of equal luminosity. The areas of the rect- 

 angles may be regarded as the amount of stimulation of the 

 three primary sensations of the Young-Helmholtz theory. 

 Indeed I took the value of the ordinates from one of Sir Wm. 

 Abney's curves, merely exaggerating the third component in 

 fig. io to make it visible. So all the data of the three-colour 

 theory may be utilised. But figs. 8, 9, and 10 do more than 



represent the results of the three-colour theory. They are 

 energy curves, crude ones it is true, but sufficiently repre- 

 sentative for the discriminating power of the colour-perceiving 

 centre of the brain. The sides of the rectangles have been 

 placed at '4, *5, '6 and "j in the diagram, but this is not essential. 



Colour-blind persons may resolve their curves only into 

 two rectangles, and it is possible that under certain circum- 

 stances specially gifted persons may resolve theirs into four 

 rectangles, in which case the ordinary representation of the 

 results of colour-mixing by the positions of points on a plane 

 would fail to hold. It would be difficult to establish a failure 

 of this construction, however, since the experimental results 

 are never very accurate. It is interesting to note that Burch x 

 is inclined to predicate four visual components. 



Instead of accounting for colour-blindness by a want of 

 discriminating power in the colour-perceiving centre of the 



1 Parson's Colour Vision, p. 115. 



