828 



NATURE 



[DECEMlJtK 



Colour Vision and Colour VUlon Theory. 



In his letter on this subject which appeared in 

 Nature of September 29. p. 473. I>r. KdridRc-CJreen 

 apparently promised to deal with my explanations 

 when these were given. In reply I selected two m 

 which the full reasoning had been given, and I 

 invitetl him to implement his undertaking. In his 

 letter in Nature of November 10, p. 687. he ap- 

 parently declines to do so, and passes his burden on 

 to the reader. ^^ . . ^ 



In none of his letters has Dr. Edndge-Green 

 attempted to discuss my proofs, or indications of the 

 proofs ; but on the contrary he has in each letter 

 merely made fresh a.ssertions of other cases in which 

 he presumes similarly that the trichromatic theory 

 is incompetent. Nevertheless, ignoring the ab.sence 

 of discussion on his part. I took up each fresh 

 assertion as it came, and dealt with it as with those 

 which preceded. In his present letter he repeats 

 the process, bringing forward three new cases. 

 Since, in one of these, he deals with a statement of 

 mine (" Colour Vision," p. 151). I shall, as before, 

 discuss these new examples also, though that pro- 

 cedure cannot be continued indefinitely. 



He seems to agree that my explanation of the 

 absence of shortening of the spectrum at the red 

 end after fatigue by yellow light is sound if the 

 presumption made be true. But he asserts that the 

 presumption is inconsistent with the work of Konig, 

 Abney, and others. The presumption is that all 

 three sensations are stimulated by visible light of 

 any wave-length. Now Dr. Edridge-Green is wrong 

 in asserting that this presumption is inconsistent 

 with the work of Konig, Abney, and the others. It 

 is certainly inconsistent with the presumption made 

 alternatively by these investigators, which implied 

 that the red sensation alone was stimulated by light 

 near the red end of the spectrum. But the work 

 never proved the correctness of that presumption, 

 which, like the other, was quite a legitimate one 

 nevertheless. For all conclusions regarding colour 

 mixture, obtainable by one set of fundamentals, 

 are necessarily obtainable by any other set linearly 

 related thereto. The only type of work, by which 

 discrimination between two otherwise suitable sets 

 of fundamentals is possible, is work which deals 

 with phenomena related to the sensations by a 

 non-linear law. That condition was apparently 

 unknown to Abney, and so he adopted the not 

 necessarily true view that his selected fundamentals 

 were absolute. They are certainly very convenient 

 for observational work. 



Dr. Edridge-Green then says that my explanation 

 does not explain why there is shortening of the red 

 end of the spectrum after fatigue by red light. Now 

 that is an entirely separate point to which my ex- 

 planation was not directed. But the trichromatic 

 explanation thereof is quite simple and straight- 

 forward. The theory never asserts that lights of 

 two different wave-lengths (yellow and red lights in 

 the present case) will necessarily produce the same 

 fatigue effects in any given region of the spectrum. 

 The law of fatigue is not yet known, but the fatigue 

 parameters used in the trichromatic theory are 

 adequate fully to express it whenever they are 

 formulated as functions of the various conditions 

 which can affect fatigue. It is the absence of re- 

 cognition of these and similar features of the moulding 

 of the trichromatic theory which has led Dr. Edridge- 

 Green into the erroneous statements unfortunately 

 made by him so frequently regarding the powers of 

 the theory. 



Dr. Edridge-Green next raises another new point 



NO. 2823, VOL. I 12] 



in saving tliat 1 do not "explain SIk' 

 crucial experiment, namely, that hi.«» 

 not seen with spectral yellow light but 

 mixed yellow made up of re<l and j.'i'< n n 

 it." Now Hidwell's experiment is in im ?>• 

 The trichromatic theory has no a priori 



that a pure yellow light and a mixed \«- 



shall have the same effect with regard to 

 colours any more than it has with regard to • 

 That is entirely dependent on the nature of 

 and interactions which are involved, a ■ 

 which the theor>' makes no fixed for. 

 nouncement. It is a matter for pli 

 physiological investigation. When that inquiry is 

 settled, the theory will incorporate the result as an 

 aid to the formulation of the parameters n 

 of known quantities, just as in the case of 

 discussed above. The theorist welcomes 1' 

 observations, and he is willing to give a 

 welcome to those made by Dr. Edridge-Green. 



The third and last new case is in the same positi- 

 Dr. Edridge-Green cites the gradun' '■ - - - 

 of the positive after - image of a 



Sroceeds succes-sively from the red to i... w.,.^;. < .. . 

 [e says that the trichromatic theory states that tht- 

 positive effect of the red sensation disappears before- 

 that of the green. The statement applies if wr 

 adopt (say) Abney's fundamentals, but might require 

 modification otherwise. Yet that is not of anv 

 essential importance. But he proceeds to say that. 

 " in an absolutely dark room, if pure spectral yellow 

 light be thrown on a white screen and a flicker 

 apparatus rotated slowly in front of it. the yellow \' 

 not change its hue ; on the trichromatic theor> 

 should become green. The results are quite different 

 when stray light is allowed to fall on the screen as 

 well." Now, while the trichromatic theorist will 

 welcome any such verified data, he cannot admit 

 any compulsion towards the expectation that the 

 colour should become green. The conditions of the 

 retinal illumination are entirely different in the two 

 cases. So the results of the observations can only 

 furnish information regarding the manner of variation 

 of the decay parameters and of the threshold values 

 as functions of the illumination and its duration 

 and its quality, of the length of the rest interval, 

 and also of the areal distribution of the illumination. 



I am glad that Dr. Edridge-Green has brought 

 forward these three examples, for they are typical 

 of many cases in which the views of writers on the 

 subject have been adversely affected by the stereo- 

 typing of ideas which, while being appropriate 

 enough to the strong restrictions properly imposed 

 in the earlier stages of the theoretical development, 

 have long since been removed. W. Peddie. 



University of St. Andrews, 

 November 10. 





Late Fertilisation and Sex-Ratio in Trout. 



Mrsic ^ has shown that in rainbow trout late 

 fertilisation — i.e. the retention of ova within the 

 body of the female after thev are fully ripev— results, 

 as it does in frogs (Hertwig,* Kuschakew^tsch ») m 

 an increased percentage of males in the offspring. 

 This is due (as also in frogs — Hertwig *) to the trans- 

 formation into males of some of the young animals 

 which had started to develop as females. The only 

 difference between the frog and the trout is that, 

 whereas in the former case the short period of 4 days' 

 delay will cause all females to become transformed 



1 Mrsic, 1923, Arch. Eniw. Afeck., 98. 129. 



* Hertwig, R., 1912, BM. CentraM. 32. 



* Kuschakewitsch, 1910, " Festschr. f. R. Hertwig, 1910. 



* Hertwig, R., Sitz. Bayr. Ak. Wiss., 1921. 



