898 THEORIES OF COLOUR VISION. [BOOK m. 



have no choice. If however we are allowed three colours 

 instead of two, we have a much greater range. If we take any 

 three colours, provided only that they lie a certain distance 

 apart along the spectrum, we can produce white by mixing 

 them in certain proportions. If we take any red, green and 

 blue, we can by adjusting the amount of each, that is the in- 

 tensity of each, produce white. 



We may go further than this. By adjusting the amounts of 

 each of the three colours we can reproduce all the colours of 

 the spectrum. If we take, for instance, a red of a certain 

 wave-length, a green of a certain wave-length, and a blue of a 

 certain wave-length, we can, without calling to our aid any 

 other wave-lengths, by varying the relative intensities of the 

 three, produce not only white light, but also orange, yellow, 

 and violet, with all the intermediate tints, that is to say, pro- 

 duce all the colours of the spectrum ; and we may in the same 

 way produce the non-spectral purple. Our choice however is 

 to a certain extent limited ; the three colours which we choose 

 must be spread over the spectrum, for we cannot obtain these 

 results with three colours taken from the red and yellow alone, 

 or from the green and blue alone. Moreover, the result is not 

 a complete one ; the colour which we thus produce by combin- 

 ing three spectral colours differs from a true spectral colour in 

 not being saturated ; it is " mixed with white," more so in some 

 cases than in others ; in relation to this deficiency of satura- 

 tion, the green region of the spectrum behaves differently from 

 the red end and the blue end. 



564. These results shew that the primary colour sensa- 

 tions out of which our recognized colour sensations originate, 

 may be reduced to three in number. If we suppose that we 

 possess three primary sensations so disposed in reference to the 

 spectrum, so arranged so to speak along the spectrum, that a 

 ray of light affects each of the three differently according to 

 its wave-length, we can understand how all our multitudinous 

 colour sensations may arise from the varied excitation of these 

 primary sensations. There may be more than three of these 

 primary sensations, but if so they must behave as if they were 

 three ; they cannot be less, since as we have seen the results 

 of mixing two sensations only are extremely limited. We 

 may therefore speak of our vision as trichromic, as based 011 

 three, or the equivalent of three, primary sensations. 



When we attempt to inquire further into the nature of these 

 primary sensations, we find ourselves in the face of two rival 

 theories. 



The one, propounded by Young but more fully elaborated 

 by Helmholtz and Maxwell, and known as the Young-Helmholtz 

 theory, teaches that there are three and only three such primary 

 sensations. As we have just seen, any three parts of the spec- 



