894 CHARACTERS OF COLOUES. [BOOK in. 



trum itself. Now the latter is produced by rays of certain 

 wave-lengths, whereas the rays of red and of yellow are respec- 

 tively of quite different wave-lengths. The orange of the spec- 

 trum cannot be made up by any mixture of the red and the 

 yellow of the spectrum in the sense that the red and yellow rays 

 can unite together to form T&ys of the same wave-lengths as the 

 orange rays; the three things are absolutely different. It 

 is simply the mixed sensation of the red and yellow which is 

 indistinguishable from the sensation of orange ; the mixture is 

 entirely and absolutely a subjective one. In the same way we 

 may by appropriate mixtures produce the sensations correspond- 

 ing to other parts of the spectrum. Now we must suppose that 

 rays of different wave-lengths affect the retina in different ways 

 and so give rise to different visual impulses, that, for instance, 

 the visual impulses generated by orange rays are different from 

 those generated by red rays or by yellow rays. Hence we are 

 led by the fact of mixed sensations being identical with other 

 apparently simple sensations to infer that the visual impulses 

 and hence the visual sensations which any ray originates are 

 of a complex character. We conclude, for instance, that the 

 impulses which a ray in the middle of the orange gives rise to 

 are not simple impulses answering exclusively to the colour of 

 that ray, but complex impulses, parts of which may be excited 

 by rays other than the particular orange ray in question. In 

 saying this we must bear in mind that we possess no direct 

 information of the nature of visual impulses, our knowledge of 

 these being limited to what we learn through the sensations to 

 which they give rise ; the complexity of the sensation may be, 

 and indeed probably is, of a different order from that of the 

 visual impulse ; to this point we shall return. 



The view that our ordinary colour sensations are mixtures 

 of simpler sensations is further confirmed by an examination of 

 the colours of external nature. For, though we see around us 

 very many colours besides those present in the spectrum, yet 

 we find that the sensations of all these colours may be repro- 

 duced by mixtures of sensations excited by various parts of the 

 spectrum. Thus the colour purple, which is so abundant in 

 the external world and yet so conspicuous by its absence from 

 the spectrum, may be at once reproduced by fusing in proper 

 proportions the sensations of red and of blue. And very many 

 other colours present in the external world but not seen in the 

 spectrum itself may be produced by mixing various spectral 

 colours in various proportions. 



Other colours in nature may be reproduced by mixing spec- 

 tral colours with white or with black. When by means of a 

 slit we allow a certain limited part of the spectrum, say in the 

 green, to fall on a certain area of the retina, the rays exciting 

 that area have certain wave-lengths, lying within certain limits, 



