114 HA KIT AND INTELLIGENCE. [cHAP. 



We might separated by an arc of 360° stand to each other in the ratio of 

 expect J ^Q 2^ the wave-frequencies of any pair of colours which are 

 mentaries opposite to each other in the circle must he in ratio of 1 to the 

 to be 180° soTjare root of 2. It might, I say, be expected, that such pairs 

 of opposites should he pairs of complementanes. 



But this does not appear to be the case. According to Pro- 

 fessor Clark Maxwell, the wave-frequencies of some of the 

 pairs of complementary colours are in the ratios of the numbers 

 here given with them ; ^ each colour being the complementary 

 of that opposite to it in the other column. 



Eed . . . 36-40 Bluish green . 4765 

 Orange, . . 39'80 Blue . . . 51-80 

 Yellow. . . 41-40 Indigo. . . 54-70 



If the wave- frequencies of the colours in the right-hand 

 column were to those of their complementaries in the ratio of 

 the square root of 2 to 1, they would be — 



Bluish green 51-47 



Blue 56-28 



Indigo 58-54 



Disere- It will be noticed that the observed wave-frequencies of these 



pancy of i]xTee colours are considerably smaller than those calculated 

 tion and by the hypothesis from the wave-frequencies of their comple- 

 theory. mentaries ; and the differences, besides being all on the same 

 side, are too great to be accounted for by any accident or error. 

 Thus in such a circle as I have imagined (in which equal dis- 

 tances along the circumference correspond to equal ratios of 

 frequency), any two complementaries are not precisely opposite, 

 but approach each other on the green side of the circle. 



Prom a mathematico-physical point of view this may appear 

 How anomalous, but it is easily explained from a physiological one. 



acconntecl ij^ite is not a mathematical expression ; it is the name we give 

 for, 



to the colour- sensation produced by sunshine. We say that the 



rays of which the sunbeam is composed are of very unequal 



1 See the paper already referred to. The determinations Tvere made by 

 means of an interference-spectrum, obtained by an application of the prin- 

 ciple of Newton's rings, and the numbers given are the numbers of wave- 

 lengths in tlip retnrdations. The determinations are for the positions of 

 certain of Fi ounhofer's lines. In the cases of bluish-green, blue, and indigo, 

 I have taken the mean of the determinations for two or three lines in the 

 same colour. 



