CHAP, in.] 



SIGHT. 



899 



trum, with certain restrictions, might be chosen as correspond- 

 ing to these three primary sensations so far as concerns the 

 reproduction, by means of them, of all other colour sensations; 

 hence in determining the nature of the primary sensations 

 we must have recourse to other considerations. We may for 

 instance very naturally suppose that two of the three correspond 

 to the two ends of the spectrum, and may therefore be spoken 

 of as more or less closely corresponding to our recognized sensa- 

 tions of red, and of violet. If red and violet be thus two of 

 the sensations the third one must correspond to green, for only 

 a sensation corresponding to green would give white when 

 mixed with the other two sensations. Or again, choosing green 

 in the first instance as one of the primary sensations for tin- 

 reason that it stands apart from the others in its complement, 

 purple, not being a spectral colour, we may decide that the two 

 other primary sensations ought to differ as much as possible 

 from each other, and therefore choose red and blue rather than 

 red and violet since violet is obviously more allied to red than 

 is blue ; indeed we may perhaps regard violet, on account of 

 its relations to red, as the beginning of a second spectrum the 

 greater part of which is invisible. The decision between these 

 two forms of the same theory rests on a number of considera- 

 tions, into the discussion of which we cannot enter here. 

 Unless we specially call attention to the difference between 

 them, which acquires importance on certain occasions only, we 

 shall treat them as identical, and use the words blue and violet 

 in this connection indifferently. 



Such a view of three primary colour sensations is represented 

 in the diagram (Fig. 150). Thus the red primary sensation, 



O Y Gr. St. V 



FIG. 150. DIAGRAM OF THREE PRIMARY COLOUR SENSATIONS. 



1 is the so-called ' red,' 2 ' green,' and 3 ' violet ' primary colour sensation 

 JJ, 0, Y, &c. represent the red, orange, yellow, &c., colour of the spectrum. The 

 diagram illustrates, by the height of the curve in each case, how the se\vr.il 

 primary colour sensations are respectively excited to different extents by vibra- 

 tions of different wave-lengths. Rut. in this, and also in Fig. lol, the" curves 

 are to be understood not as careful curves of actual variations in the intensity 

 of the several changes, but as simply serving to illustrate roughly the nature of 

 the theory. 



