496 



Dr. G. J. Burch. On Light- [Apr. 19, 



system. They represent the four colour-sensations described in my paper on 

 " Artificial Temporary Colour-Blindness."* 



Inasmuch as according to the theory these curves must be symmetrical, 

 I assumed the apex of the green sensation to lie midway between the two 

 ends of it, and similarly with the blue sensation. Then that part of the 

 yellow where neither red nor green predominates must be where the tangents 

 to the two curves are equal and opposite. In this way I got a position for 

 the apex of the red sensation, and in like manner for the violet sensation by 

 means of the blue. It may serve to fix our ideas if I refer to the illustration 

 of the musical scale. Taking E as the apex of red, green would be 

 approximately at G, blue at A, and violet at B. 



It is noteworthy that green, blue, and violet are practically equidistant 

 and much closer together than red and green — red in fact might almost be 

 taken to correspond with Eb. 



In order to assign a value to the coefficient of friction I was guided by the 

 keyboard interval over which the forced vibration must extend. After 

 a number of trials I decided to take 4k 2 = - 16 as a trial value, giving each 

 of the four colour-sensations the same coefficient of friction. 



It will be observed that the transition from blue to violet occurs at G-, 

 that from green to blue at F, and that from red to green a little above the 

 D line — or exactly at the D line if the centre of the red is placed at Eb. In 

 this case, too, the red would meet the blue about the b line, as it does 

 during artificial green-blindness. 



Eig. 4 shows the effect of increasing 4& 2 to 0'60. Each one of the curves 

 is lower, but it extends over a greater range of wave-lengths in proportion to 

 its height. This agrees perfectly with the well-known fact that " by feeble 

 light all colours tend to grey." 



ABC D b F GHK 



Fig. 4. 1 



Eesonance Curves of the Colour Sensations Keel, Green, Blue, and Violet, with the 

 coefficient of friction, k = 0'4 nearly, i.e. 4P = 0*60. 



In my paper on " Colour- Vision by Very Weak Light "f I have noted the 

 changes which occur in such cases between the boundaries of the colour- 



* ' Phil. Trans.,' B, vol. 191, pp. 1-34. 



+ ' Boy. Soc. Proc.,' 1905, B, vol. 76, p. 214. 



