43-4 Scientific Intelligence. 



The Purkinjo effect may also be accounted for along the same 

 general lines. This effect consists in the horizontal contraction 

 of the visibility curve and in the shifting of its maximum toward 

 the limiting wave-length (V50/v, as the intensity of the light 

 decreases to a very small value. Instead of supposing that the 

 rods in the retina are chiefly responsible for vision at low inten- 

 sities and the cones for vision at high intensities, it is only neces- 

 sary to assume that the vibrators have different free periods and 

 that the number O-bo/A (which was substituted for a symbol a in 

 calculating the first approximation to Ives' curve) is a mean value. 

 In other words, the visibility curve is the resultant of a very 

 large number of elementary curves of the same type and not of 

 three curves which would correspond to the hypothesis of three 

 independent primary sensations. 



(b) Since it is not possible to give briefly an adequate account 

 of the general theory of the cerebral process, a few sentences 

 from the original paper, which contain the kernel of the matter, 

 will now be quoted. " We suppose the vibrators to set up waves 

 in the nerves and that the nerves carry these waves to the brain. 

 The analogue of the telephone may be useful in this respect ; the 

 vibrators may be likened to the diaphragm and the nerve to the 

 telephone wire. There is, however, one important difference. 

 When a musical note is sung into the telephone it is transmitted 

 correctly along the wire and reproduced at the other end, because 

 the diaphragm reproduces accurately the sound wave. It is funda- 

 mental to the theory given here that the vibrator does not repro- 

 duce the light wave accurately, owing to its being subject to too 

 many disturbing influences. No matter how monochromatic the 

 incident wave is, the wave transmitted along the nerve is not 

 monochromatic. It is as if the diaphragm in the transmitter 

 were subject to disturbing influences of such a nature that the 

 person at the other end hears a medley of musical notes over a 

 range of half an octave on each side of the notes originally sung 

 into the transmitter." 



Houstoun takes a slightly asymmetric curve of the same gen- 

 eral shape as a probability curve as typical of the distribution of 

 energy over the range of wave-lengths set up in the nerve by 

 monochromatic stimulation. The area under the curve gives the 

 luminosity of the impression, the position of the maximum, the 

 hue, and the narrowness of the curve the degree of saturation. 

 The phenomena of color mixing are readily reproduced theoret- 

 ically by drawing energy curves corresponding to the objective 

 constituents of the incident light and then constructing the 

 resultant curve. For example, we may plot one sensation curve 

 for lithium red and another for thallium green and then add the 

 ordinates, the abscissas being proportional to the wave-lengths. 

 Finally, following Dr. Edridge-Green, the author explains the 

 apparent trichromatism of our ordinary visual sensations on the 

 ground that the color-perceiving center in the brain is not suf- 

 ficiently developed to discriminate between the character of 



