July 14, 1893.] 



SCIENCE. 



19 



of three different parts of the spectrum respectively), and they 

 severally cause the sensations of red, green and blue. But when 

 all three of these substances are present at once they recoinbine 

 to produce the exciter of the grey sensation, and thus it happens 

 that the objective mixing of three colors, in proper proportions, 

 gives a sensation of no color at all, but only grey. 



This theory is found, upon working it out in detail, to avoid 

 the difficulties of the theories of Helniholtz and of Hering. 



Its assumption of a separate chemical process for the produc- 

 tion of the sensation of grey gives it the same great advantage 

 over the Young-Helmholtz theory that is possessed by the theory 

 of Hering ; it enables it, namely, to account for the remarkable 

 fact that the sensation of grey exists unaccompanied by any sen- 

 sation whatever of color under the five following sets of circum- 

 stances — when the portion of the retina affected is very small, 

 ■when it is very far from the fovea, when the illumination is very 

 faint, when it is very intense, and when the retina is that of a 

 person who is totally color-blind. This advantage my thecry at- 

 tains by the perfectly natural and simple assumption of s. partial 

 decomposition of chemical molecules ; that of Hering requires us 

 to suppose that sensations so closely related as that of red and 

 green are the accompaniments of chemical processes so dissimilar 

 as the building up and the tearing down of photo-chemical sub- 

 stances, and farther that two complementary colors call forth 

 photo chemical processes which destroy each other, instead of 

 combining to produce the process which underlies the sensation 

 •of grey. 



Of the first four of the above enumerated cases the explanation 

 will readily suggest itself ; in the case of the totally color-blind it is 

 simply that that differentiation of the primitive molecules by 

 which they have become capable of losing only a part of their 

 exciting substance at one time has not taken place; the condition, 

 in other words, is a condition of atavism. In partial color-blind- 

 ness and in the intermediate zones of the retina in normal vision 

 the only colors perceived are yellow and blue. This would indicate 

 that the substance which in its primitive condition excites the 

 sensation of grey becomes in the first place diflferentiated into 

 two substances, the exciters of yellow and blue respectively, and 

 that at a later stage of development the exciter of the sensation 

 of yellow becomes again separated into two substances which 

 produce respectively the sensations of red and of green. In this 

 way the unitary (non-mixed) character of the sensation yellow 

 is accounted for by a three-color theory as completely as by a 

 four-color theory. A three-color theory is rendered a necessity 

 by the fact that it a'one is reconcilable with the results of Konig's 

 experiments for the determination of the color-equatinns of 

 color-blind and of normal eyes,' experiments which far exceed in 

 accuracy any which have yet been made in color-vision, but 

 which, owing to the intricate character of the theoretical deduc- 

 lions made from them, have not hitherto been allowed their due 

 weight in the estimation of color theories. 



The explaniition which the theory of Hering gives of after- 

 images and of simultaneous contrast are not explanations at all, 

 but merely translations of the facts into the language of his 

 theory. My theory is able to deal with them more satisfactorily ; 

 when red light, say, hf s been acting upon the retina for son e 

 time, many of the photo-chemical molecules have lost that one of 

 their constituents which is the exciter of the red sensation; butin 

 this mutilated condition they are exceedingly unstable, and their 

 other two constituents (the exciters of the sensations of blue and 

 of green) are gradually set free ; the effect of this is that, while 

 the eyes are still open, a blue-green sensation is added to the red 

 sensation with the result of making it gradually fade out into 

 white, and, if the eyes are closed, the cause of the blue-green 

 sensation persists until all the molecules affected are totally de- 

 composed. Thus the actual course of the sensation produced by 

 Icoking at a red object, — its gradual fading out. in case of care- 

 ful fixation, and the appearance of the complementary color if 

 the illumination is diminished or if the eyes are closed, is exactly 

 what the original assumption of a partial decomposition of 

 molecules would require us to predict. The well-known ex- 

 treme i-apidity of the circulation in the retina would make it im- 



»A Kon'g und C. DislBriQl. Sitzungberlchteder Berl. Jioff voni39 Jull, 1ES6. 



possible that the partly decomposed molecules just referred to 

 should remain within the boundaries of the portion of the retina 

 in which they are first produced; and their completed decomposi- 

 tion after they have passed beyond these boundaries is the cause 

 of the complementary color- sensation which we call simultaneous 

 con'rast. The spreading of the actual color which succeeds it 

 would then be accounted for, as Helniholtz suggests, by a 

 diffusion of Ihe colored light in the various media of the eye. 



No effort has hitherto been made to explain a very remarl able 

 feature in the structure of the retina, — the fact that the retinal 

 elements are of two different kinds, which we distinguish as rods 

 and cones. But this structure becomes qui'e what one might 

 expect, if we suppose that the rods contain the undeveloped 

 molecules which give us the sensation of grey only, while the 

 cones contain the color molecules, which cause sensations of grey 

 and of color both. The distribution of the rods and cones cor- 

 responds exactly with the distribution of sensitiveness to just per- 

 ceptible light and color excitations as determined by the very care- 

 ful experiments of Eugen Fick." 



Two other theories of light sensation have been proposed 

 besides the one which I have here outlined, either one of which 

 meets the requirements of a possible theory far better than that 

 of Hering or of Helmholtz; they are those of Goller^ and 

 Bonders.' The former is a physical theory. That of Bonders 

 is a chemical theory, and very similar to the one which I here 

 propose. Every chemical theory supposes a tearing down of 

 highly complex molecules; Donders's theory supposes, in addition, 

 that the tearing down in question can take place in two succes- 

 sive stages. But Donders"s theory is necessarily a four color 

 theory ; and Bonders himself, although the experiments of 

 Konig above referred to had not at that time been made, was 

 so strongly convinced of the necessity of a three-color theory for 

 the explanation of some of the facts of color-vision that he sup- 

 plemented his four-process theory in the retina with a three- 

 process theory in the higher centres. The desirableness, therefore, 

 of devising a partial decomposition of molecules of such a nature 

 that the fundamental color processes assumed can be three in 

 number instead of four is apparent. 



But the theory of Bonders is open to a still graver objection. 

 The molecules assumed by him must, in order to be capable of 

 four different semi-dissociations, consist of at least eight different 

 atoms or groups of atoms. The red green dissociations and the 

 yellow-blue dissociations we may then represent symbolically by 

 these two diagrams respectively : 



But it will be observed ihat the two completed dissociations end 

 by having set free different combinations; in the one ca-e 1 is 

 combined with 2 and in the ether case 1 is combined with 8, etc. 

 If, now, the partial dissociations are so unlike as to cause sensa- 

 tions of yellow and blue (or of red and green) it is not probatde 

 that completed dissociations which end in setting free different 

 chemical combinations should produce the same sensation, grey. 

 The difficulty introduced by Bonders's theory is theiefore (as in 

 the case of Hering's theory) as great as the difficulty sought to 

 be removed. It is the desire to secure the advantages of a partial 

 d ssociation theory, without the disadvantages of the theory of 

 Bonder's, that has led me to devise a partial dissociation of mole- 

 cules of a different kind. The theory will be found more 

 explicitly set forth in the next number of the Zeitschr,ft fiir 

 Psychologie. 



2 Studlen uber L'cht uad FarbeQempaadung. Pfiiiger's Archiv, Bd. 

 XLIV., s. 441, 1888. 



3 Die Analyse der Ll'ihtwellen darcli das Auge. Oil Bois-Reymond^s Arcfiiv, 

 1889. 



* Kocli elnmal die Farbec-systeme. Grd/e's Archiv fiir OpMhalmologie, 

 Bd.SOd), :SSi4. 



