216 RECENT PROGRESS OF THE THEORY OF VISION. 



tion), and difference of amount of illumination (brightness). It 

 is in this way that we describe the system of colours in ordinary 

 language. But we are able to express this threefold difference 

 in another way. 



I said above that a properly constructed colour-disc ap- 

 proaches a triangle in its outline. Let us suppose for a moment 

 that it is an exact rectilinear triangle, as made by the dotted line 

 in Fig. 34 ; how far this differs from the actual condition we 

 shall have afterwards to point out. Let the colours red, green, 

 and violet be placed at the corners, then we see the law which 

 was mentioned above : namely that all the colours in the in- 

 terior and on the sides of the triangle are compounds of the 

 three at its corners. It follows that all differences of hue depend 

 upon combinations in different proportions of the three primary 

 colours. It is best to consider the three just named as primary ; 

 the old ones red, yellow, and blue are inconvenient, and were 

 only chosen from experience of painters' colours. It is impossible 

 to make a green out of blue and yellow light. 



We shall better understand the remarkable fact that we are 

 able to refer all the varieties in the composition of external light 

 to mixtures of three primitive colours, if in this respect we 

 compare the eye with the ear. 



Sound, as I mentioned before, is, like light, an undulating 

 movement, spreading by waves. In the case of sound also, we 

 have to distinguish waves of various length which produce upon 

 our ear impressions of different quality. We recognise the long 

 waves as low notes, the short as high-pitched, and the ear may 

 receive at once many waves of sound that is to say, many 

 notes. But here these do not melt into compound notes in the 

 same way that colours, when perceived at the same time and 

 place, melt into compound colours. The eye cannot tell the 

 difference, if we substitute orange for red and yellow ; but if we 

 hear the notes C and E sounded at the same time, we cannot 

 put D instead of them, without entirely changing the impres- 

 sion upon the ear. The most complicated harmony of a full 

 orchestra becomes changed to our perception if we alter any one 

 of its notes. No accord (or consonance of several tones) is, at 



