VISUAL SENSATIONS 643 



so far as the stimuli are concerned, are compound so far as the sensa- 

 tion is concerned. Thus most people would say at once that orange 

 is a mixture of red and yellow, and, as a matter of fact, we find that 

 on mixing rays from the red with others from the yellow part of the 

 spectrum we do obtain a sensation of orange. The stimulus obtained 

 by mixing the red and yellow rays is not the same as a stimulus 

 caused by rays from the orange part of the spectrum. In the former 

 case compound waves made up of the two wave-lengths, 656 /x/x and 

 564 yu/jt, are falling on the retina, in the latter case a simple wave 

 with a length of 608 JULJUL, and yet the sensations produced are identical. 

 Experience shows that there are relations between the physiological 

 effects produced by different parts of the spectrum which have no 

 physical analogue in the stimuli themselves. Such, for instance, are 

 the phenomena known as the coloured after-image. We have seen 

 that, as the result of fatigue, stimulation of any part of the retina 

 by a bright object produces, when stimulation is removed, a dark 

 after-image, which has its seat in the previously stimulated portion 

 of the retina. If we look steadfastly for a minute in a bright light 

 at a red disc on a white ground and then look away at a uniform 

 white surface, we see an after-image of the disc on the surface. This 

 after-image is, however, green, and the white background takes on 

 a reddish tinge. If the disc in the first instance has been a greenish 

 blue the after-image is red, and to every colour in the spectrum we 

 find there corresponds another which represents the after-image 

 evoked by stimulation with the first colour. If the first disc has 

 been green the after-image will be purple, and vice versa. We 

 can therefore arrange the spectrum into a series of pairs of colours 

 which are known as complementary. The following is a list of 

 such pairs, with wave-lengths of the rays involved, as determined 

 by Helmholtz : 



Colours Wave-lengths 



Red greenish blue .... 656 492 



Orange blue 608490 



Bright yellow blue 574482 



Yellow indigo 567465 



Greenish yellow violet .... 564 433 



If the retina be stimulated by any of these pairs of colours simul- 

 taneously the effect is not that of colour, but of white. White light, 

 or the sensation of white, can thus be due either to simultaneous 

 stimulation of the retina by all the rays of the spectrum, or to stimu- 

 lation of the retina by pairs of colours which are known as comple- 

 mentary. If these pairs be taken rather nearer in the spectrum 

 a colour is obtained representing a part of the spectrum situated 

 between the two constituents of the pair. If the rays are further 



