THE MIXTURE OF PURE-COLOUR STIMULI 33 



CHAPTER III 



THE MIXTURE OF PURE-COLOUR STIMULI 



Sir Isaac Newton (1704) first scientifically investigated the pheno- 

 mena of colour mixtures. The fundamental laws were first enunciated 

 by Grassmann (1853). They may be stated in the following form : 



(1) Unlike lights mixed with like lights produce unlike mixtures : 

 or, if in a mixture one component is continuously altered the appearance 

 of the mixture will also continuously alter. 



(2) Like lights mixed with like lights produce like mixtures : or, 

 if two lights that look the same are each mixed with a third light the 

 resultant mixtures will look alike. 



Special Case : Proportional increase of intensity of each component 

 does not destroy the match. This corollary will be found on further 

 investigation to require some reservation. 



(3) Every mixture of lights can be matched by a definite spectral 

 light or a definite purple mixture which is mixed with a definite amount 

 of white light : or, if we take any fixed homogeneous or composite 

 light and mix it with the whole series of pure spectral lights, completed 

 by purple, varying the proportions in the mixture from zero of one to 

 zero of the other, we obtain every known variety of stimulus. 



The importance of these laws is that matches of optical mixtures 

 resemble algebraical equations and can be treated as such, the match 

 holding good if any addition or subtraction is made to both optical 

 mixtures. 



Owing to the facts already stated on p. 27 it is clear that we can 

 pass continuously from red to violet by two paths, one via the spectrum, 

 the other via purple. This is represented graphically by a closed curve, 

 and Newton naturally chose the circle as the simplest. It is not, 

 however, the most comprehensive, as will be seen by the results of 

 observing various colour mixtures. It is at once obvious, for instance, 

 that the progress from red to violet via purple must be represented 

 by a straight line, for experiment shows that purple can only be obtained 

 by mixing red and violet, and therefore a given purple must be 

 represented as a point on the chord joining the points which represent 

 these colours. 



The same applies to all hues of wave-length greater than 540 ///x 

 (in the yellow-green). For example, any mixture of red light of 670 /x/x 



p. c. V. 3 



