402 Dr. R. A. Houstoun on a 



variable ; so that for practical work the elastic support 

 under the motor should be adjusted until the critical speeds 

 of the motor are nowhere near the normal speed of rotation. 



(7) The way in which the critical speed of a motor is 

 related to the deflexion of its support is given in fig. 2. 



(8) If the elastic support of the motor is made up of pads 

 or layers of cork, rubber, or felt, then n a is reduced if the 

 thickness of the layer is increased ; n a is increased if the 

 area is increased. 



XXXVI. A Theory of Colour Vision. By Dr. R. A. 

 Houstoun, Lecturer on Physical Optics in the University 

 of Glasgoic *. 



§ 1. TTNDER the above title I contributed a short 

 KJ article to the Proceedings of the Royal Society t 

 two years ago, in which I showed that it was not necessary 

 to assume the existence of three fundamental sets of nerves 

 or mechanisms in the retina in order to explain the facts of 

 colour vision. This article has not been understood, possibly 

 owing to defects of exposition on my part, but more probably 

 to the unfamiliarity of the ideas involved. In the present 

 paper I develop my theory somewhat further ; on account 

 of its nature I cannot hope to make it fully intelligible to all 

 the psychologists, physiologists, and non-mathematical physi- 

 cists interested in colour vision, but I have included some 

 numerical examples, and hope at least to show that my theory 

 is as capable of giving correct numerical results as the 

 Young-Helmholtz theory is. 



To put my theory as shortly as possible : 



(1) The eye is sensitive only to a limited range of wave- 

 lengths, from \ = 4 x 10" 5 cm. to A, = 7*6 x 10" 5 cm. 



(2) To explain this we must assume resonators or vibrators 

 in the retina. 



(3) No matter how regular the incident light is, the 

 vibration set up in the retina will always be more or less 

 irregular; there will be stoppages and changes of amplitude 

 and phase owing to molecular disturbances. 



(4) In Optics an irregular vibration of this kind is equi- 

 valent to a very great number of regular vibrations of 

 different periods. This follows from Fourier's integral 

 theorem ; it is also the basis of the modern work on the 



* Communicated by the Author. 

 t Vol. xcii. A, p. 424 (1916). 



