August ii, 1892] 



NA TURE 



347 



after-sensation results from the restitution of a substance decom- 

 posed by the red light. But if we reverse the experiment by 

 looking at a bright green light and then closing our eyes, the 

 after-sensation changes to red. The theory in question would 

 require us to suppose that the green light builds up a visual 

 substance which spontaneously decomposes when the eyes are 

 closed, and so produces the red after-image. I confess that such 

 a hypothesis seems to me incredible. Another remarkable 

 feature of Hering's theory is that colours termed coviplementary 

 ought to be termed antagonistic, ■ because they are capable of 

 producing a colourless sensation when mingled in due pro- 

 portions. If the complementary colours yellow and blue 

 could, when mixed, produce black, they might well be named 

 ' ' antagonistic ; " but since their combined effect is a sense 

 >f whiteness, and since the addition of them to white light 

 ncreases its luminosity, it seems very difficult to comprehend on 

 vhat ground the term antagonistic should be substituted for 

 omplementary. I confess I am quite unable to follow Hering 

 when he supposes that three pairs of mutually antagonistic 

 chemical processes are produced in the retina when white light 

 falls on it, that these processes are all continued on through the 

 optic nerve into the vision centre, and there give rise to our 

 different light and colour sensations. 



In 1881 Prof. Preyer- advanced a theory of colour sensation, 

 in which he supposes that in the retina there are four sets of 

 cones arranged in pairs — one pair being excitable by the waves 

 in the blue and yellow parts of the spectrum, the other pair 

 being excitable by the red and green. He supposes that each 

 pair of cones is connected with a ganglionic cell in the retina, 

 and through that with one fibre in the optic nerve, which trans- 

 mits the impulse to at least two cells in the vision centre, in 

 which two different qualities of sensation, red and green, yellow 

 and blue, are severally produced. I confess, however, that I 

 am not able to understand how nerve impulses received, say, 

 from the red terminal of a pair, can specially affect one of the 

 cells in the nerve centre to produce a red sensation. But if the 

 red or green sensation were supposed to arise in the same central 

 cell according to the frequency of the impulses transmitted from 

 either terminal of the pair at the periphery, I should feel that 

 an important difficulty had been removed from Prof. Preyer's 

 theory. 



It must be admitted that the production of nerve impulses 

 within the terminals in the retina is almost as obscure as ever. 

 It is still the old question, Does light stimulate the optic ter- 

 minals by inducing vibration, or by setting up chemical change? 

 Whichever view we adopt, it seems to me necessary to suppose 

 that all the processes for the production of nerve impulses can 

 take place in one and the same visual ceil, and are transmitted 

 to the brain through the same nerve fibre ; because the image 

 of a coloured star small enough to fall upon only one cone is 

 seen of a fixed and definite colour which does not alter when 

 the position of the eye is changed. It seems to me that if 

 there are special cones for red, green, yellow, and blue, the 

 colour of the star should change when its image falls on differ- 

 ent terminals, but I am assured by Mr. Lockyer that such is 

 not the case. 



I referred to the sense of smell because it seems to me that we 

 cannot in that case escape from the conclusion that the different 

 sensations arise from different molecular stimulations of the 

 same olfactory terminals. 



From Lippmann's recent researches on the photography of 

 colour ■* it appears that all parts of the spectrum can now be 

 photographed on films of albumino-bromide of silver, to which 

 two aniline substances, azaline and cyanine, have been added. 

 It seems, therefore, reasonable to suppose that a relatively small 

 number of substances could enable all the rays of the visible 

 spectrum to affect the retina. Helmhoiz believes that three 

 visual substances would suffice ; but if the primary sensations 

 are to be regarded as four — red, green, yellow, and blue — at least 

 four visual substances appear to be necessary ; and I think we 

 must assume that all of them are to be found in the same 

 visual cell in the retina, and that the nerve impulses which their 

 decompositions give rise to are all transmitted through the 

 same optic fibres to the brain cells, there to produce a sense of 

 uncoloured or coloured light. Evidently such a hypothesis is 



' E. HerinK, Zur Lehre vom Lichtsinne, 2nd ed. p. 121. 



- W. Preyer, " fiber den Fabren und Temperatur Sinn," &c., Archiv 

 fiir Physiotogie, j88i. Band xxv., p. 31 



' G. Lippmann, " On the Photography of Colour," Comftes Rendits, 

 1892, tome 114, p. 961. 



NO. II 89, VOL. 46] 



not altogether novel ; it is essentially a return to that long ago 

 suggested by Newton. The only difference is that light is sup- 

 posed to induce photo-chemical changes in (he retina, as Von 

 Helmhoiz suggested, instead of mere mechanical vibration, as 

 Newton supposed. But if in the sense of smell nerve undula- 

 tions are induced by mechanical vibrations of molecules acting 

 on delicate hairs at the ends of cells, would it, after all, be un- 

 reasonable to suppose that within each visual cell there are 

 different kinds of molecules that vibrate in different mode> when 

 excited by ether-waves? Four or five sets of such molecules in 

 each terminal element in the retina would probably be sufficient 

 to project successively or simultaneously special forms of un- 

 dulations through the optic nerve, to induce colour sensations 

 differing according to the wave form of the incoming nerve un- 

 dulation. It seems to me that the question becomes narrowed 

 down to this : Do the nerve impulses arise from mere vibration 

 or from chemical change in the molecules of the nerve terminal ? 

 The photo-chemical hypothesis has much in its favour. We 

 know how rapidly light can induce chemical change in photo- 

 graphic films, and we know that light induces chemical change 

 in the vision-purple in the outer segments of the rod cells in the 

 retina. The fact that the cones contain no vision-purple is no 

 argument against the theory, for the inner segment of both rod 

 and cone is by many regarded as the true nerve terminal, and 

 there is no vision-purple in either of them. The visual substances 

 in the cones, at all events, are colourless, and the existence of 

 them as substances capable of producing nerve impulses by 

 chemical decomposition is as yet only a speculation awaiting 

 proof. The fatigue of the retina produced by bright light is 

 best explained on a chemical theory, but it could also be ex- 

 plained on a mechanical theory, for we must remember that, 

 even if the nerve impulses produced in the visual cells 

 were merely a translation of the energy of light into vibra- 

 tion of nerve molecules, the nerve impulse has to pass 

 through layers of ganglionic cells before reaching the fibres of 

 the optic nerve, and in these cells it probably always induces 

 chemical change. The phenomena of partial colour-blindness 

 could be explained on a photo-chemical theory iiy supposing 

 that it arises from the absence of the substances required to pro- 

 duce the wave forms necessary for the colour sensation which is 

 defective, but the total colour blindness at the anterior part of 

 the retina is evidently a difficulty. How could we have a sense 

 of light from that portion of the retina if all the visual substances 

 are absent ? That is one of the reasons why Hering supposed 

 that a special visual substance is present everywhere in the 

 retina, which by decomposition gives rise to a sense of light as 

 distinguished from colour. But even on the hypothesis we are 

 pursuing, it is not necessary to suppose that all visual substance 

 is absent, for colour-blindness in the front of the retina could be 

 explained by supposing that colour perception has not been 

 developed in the corresponding portion of the vision centre, and 

 consequently all nerve impulses coming from that part of the 

 retina produce scarcely anything more than a sense of light. 



If the photo-chemical theory is entertained, it seems neces- 

 sary to suppose that there is some singular relation between the 

 pairs of substances which respectively give rise to red and green, 

 and yellow and blue, seeing that both members of a pair fre- 

 quently, if not always, fail together. 



It seems to me that the great difficulty arises when we con- 

 sider the puzzling phenomena of contrast. If light of a parti- 

 cular wave length decomposes a special substance, and gives rise 

 to, say, a sense of red, why does the complementary bluish- 

 green sensation appear in the vision centre around the spot in 

 which the red sensation arises? If the induced colour were a 

 pure green, one might attempt to explain it by supposing that a 

 sympathetic change had been induced in a substance closely 

 related to that suffering decomposition by the objective light, 

 but no such simple explanation is admissible ; the comple- 

 mentary contrast of red is not green, but a mixture of 

 green and blue. The inadmissibility of such an explanation 

 becomes still more apparent if we take pure green as the inducing 

 colour — the complementary contrast that appears is purple, which 

 involves a blue or violet, as well as a red sensation. It matters not 

 what inducing colour sensation we adopt, the induced contrast is 

 always the complementary required to make a sense of white. 

 George Wilson' long ago suggested that the simultaneous con- 

 trast probably arises from a "polar manifestation of force;" 

 indeed, he regarded it as a " true, though unrecognized, mani- 

 festation of polarity." It is enough to mention that interesting 



' Wilson, " Researches on Colour-Blindness," Edinburgh, 1855, p. 179. 



