-2 I 8 



NA TURE 



[June 27, 1901 



■occurred to me. The facts can be accounted for in a perfectly 

 ■simple manner if we suppose, as postulated by the theory of 

 'Hering, that there is an independent white sensation, and, 

 further, that the latent period for a colour sensation is very 

 much greater than that for white. For green, under the condi- 

 tions of my experiments, the latent period would be at least 

 1/40 second, while for white it can hardly exceed 1/500 second, 

 though the luminosity of the two may be as nearly as possible 

 equal. The latent period for red is probably not very different 

 from that for green under similar circumstances, while that for 

 Iblue is considerably greater. 



If in a darkened room a ray of green light is admitted to the 

 eye for 1/40 second, one sees a flash of green ; but, assuming 

 that the suppositions which have been put forward are correct, 

 the visible flash is not contemporaneous with the physical 

 illumination. One does not begin to experience the green sen- 

 sation until after the green ray which excited it has been shut 

 off. What is actually perceived is, in fact, a positive after- 

 image, the duration of which may be considerably longer than 

 that of the stimulus. But if a sufficiently luminous white sur- 

 face is presented to the eye immediately upon the expiration 

 of the brief period of stimulation by green light, the after-image 

 formed will not be positive, but negative, and the only colour 

 perceived will be purple. The fatigue to which the negative 

 image is due must have been set up during the latent period, 

 when no image at all was actually perceived. 



The formation of vivid pulsative images depends not only 

 upon the latent period, but also upon persistence, luminosity, 

 the duration of the primary impression and of the periods of 

 light and darkness, and upon other circumstances. And the con- 

 ditions which are best for some colours are not so for others. This 

 fact obviously suggests that the pulsative image might afford a 

 means of analysing compound colour-sensations, though so far 

 it has been found available only to a limited extent. If the 

 complete spectrum is projected upon the screen, it is seen at 

 once that the blue-green pulsative image of the red, and the 



than the red one before referred to, but if viewed in the manner 

 described in the paper, the appearance of the green border due 

 to pure spectral blue light of about ,\47oo is very striking. 

 Violet produces no coloured border of the kind, and its admix- 

 ture with other rays has no sensible effect upon the phenomenon. 

 It can hardly be doubted that effects which occur, sympatheti- 

 cally as is suggested, just outside the boundary of the physical 

 image, must also occur within the boundary ; and if that is so, 

 it follows that red, orange and yellow rays, nearly up to the 

 beginning of the greenish-yellow, excite a red sensation, while 

 green and blue excite a green sensation. There is at present no 

 evidence of the same kind as to the existence of any other 

 fundamental colour-sensation, though there must, of course, be 

 at least one more. 



The bearing of these border experiments upon theories of 

 colour-vision is indicated in the paper. The following is among 

 the most important points referred to. It is found that a com- 

 paratively small proportion of red mixed with other spectral rays 

 results in the formation of a red border. According to the 

 Young- Helmholtz theory, green spectral rays excite the funda- 

 mental red sensation to about the same extent as orange-red 

 rays ; yet no red border is forined by the green, though that 

 formed by the orange-red is very strong. The natural deduction 

 is that no red sensation is excited by green light. 



As regards a different point which has been much debated, 

 certain observations seem to Ije absolutely conclusive. Accord- 

 ing to Helmholtz the phenomena of simultaneous contrast are 

 due entirely to mental judgment ; according to Hering their 

 origin is a phj'siological one. Experiments with one of the eye- 

 piece methods, in which the apparent diameter of the pulsative 

 image is about one-fourth of that of the white-light disc or field 

 of view, seem to place the matter beyond dispute. If a purple 

 pulsative image is produced from a strongly illuminated green 

 colour-patch, the whole of the physically white field surrounding 

 the patch appears to be purple. It cannot possibly be that the 

 colour of the ground is a psychological effect resulting simply from 



purple pulsative image of the green, are far more intense than contrast with green, for no green whatever is consciously per- 



the pulsative images of the yellow, the blue and the violet 

 portions of the spectrum. Accordingly, if we make an orange 

 colour-patch by combining red and yellow rays, it is not sur- 

 prising to find that its pulsative image is blue-green, hardly 

 differing at all from that of red, instead of sky-blue, which is 

 the colour of the ordinary after-image of orange. Now the 

 pulsative image of a patch illuminated by the simple orange 

 rays of the spectrum is also found to be blue-green ; hence the 

 inference is clearly suggested that the spectral orange rays 

 excite a red sensation. This particular fact will probably be 

 regarded as one which needs no demonstration, but it is 

 mentioned as an illustration of the proposed method of analysis. 

 Several others in which the conclusions can be verified by trial 

 might be given. Now it is noticed that under most ordinary 

 conditions the purple pulsative image of green is more easily 

 produced than that of any other colour. Under the same con- 

 ditions we find that the pulsative images of yellow, of blue and 

 of white are purple, and, assuming that the test referred to is a 

 sound one, we conclude that yellow, blue and white all excite a 

 green sensation. The proposed method of analysis may prob- 

 ably be carried much further than has yet been done. 



This paper also contains an account of some new observations 

 upon a class of phenomena to which I have drawn attention in 

 a former communication {Proc. Roy. Sol., vol. Ix. p. 368 ; 

 Nature, vol. Iv. , p. 367). If the image of a white object is 

 suddenly formed upon the retina after 



ceived ; the cause must necessarily be of a physiological nature. .\ 

 similar effect is produced in an even more striking degree by 

 blue and by violet colour-patches, the whole field appearing to 

 be of the same hue as the pulsative images, namely orange and 

 yellow. Phenomena of simultaneous contrast, as they are 

 called, are therefore certainly not in all cases to be explained 

 solely on psychological grounds. 



The experiments which have been discussed establish 

 nothing decisively in favour of either of the two principal 

 theories of colour-vision. Some of the observations seem to 

 support the Young-Helmholtz theory, others that of Hering ; 

 others, again, appear to indicate that neither theory in its 

 present form is tenable. I venture to think that our knowledge 

 of the subject might be materially increased by further experi- 

 ments on the lines of those described. 



Shelford Bidwei.l. 



TBE SECOND INTERNATIONAL CONFER- 

 ENCE FOR THE EXPLORATION OF THE 

 SEA. 

 A FTER the International Conference which met at Stock- 

 holm in June iSgg for the consideration of a scheme for 

 the systematic scientific study of fishery questions, it was pro- 

 period of darkness, the I posed to meet again to complete the programmes at Christiania 

 object generally appears to be surrounded for about one-tenth ] in the autumn of 1900. Various circumstances made it necessary 

 of a second by a narrow red border. It was noticed that when I to postpone the meeting, which eventually took place in the 

 the bright object producing the image was looked at through second week of May, when representatives of Germany, Belgium, 

 variously coloured glasses the red border did not appear unless 1 Denmark, Finland, Great Britain, Norway, Holland, Russia and 

 the glass used was capable of transmitting red light, and it was | Sweden (the order is that adopted in the official conipte-rendu — 

 suggested that the phenomenon was due to sympathetic ex- I alphabetically in the French language), to the number of twenty- 

 citation of the "red nerve-fibres" lying immediately outside ' five, assembled in Christiania. The delegates included Dr. 

 the portion of the retina directly affected by the radiation. The Herwig, president of the German Society for Promoting Sea- 

 orange and yellow glasses employed for the observations referred 1 fisheries; Profs. Kriimmel and Brandt^ of Kiel, and Profs 



to of course transmitted red light. Using the pulsative image 

 apparatus with the eyepiece method, I now find that the simple 

 red, orange and yellow rays of the spectrum, whether alone or 

 in conjunction with any others, are competent to produce the 

 red borders. The effect stops short at the beginning of the 



Heincke and Henking from Germany ; Prof Gilson of Lo 

 from Belgium ; Captain Drechsel, Dr. Martin Knudsen and Dr. 

 C. G. J. Petersen from Denmark ; Dr. Xordqvist from Finland ; 

 Sir Colin Scott Moncrieff, Prof. D'Arcy Thompson, Dr. H, R. 

 Mill and Mr. W. Garstang from Great Britain ; Prof. Nansen 



greenish-yellow. When blue and green rays are employed to and Dr. Iljort from Norway ; Dr. P. P. C. Hoek from Holland ; 



illuminate the patch, either separately or in combination with ! Dr. Knipovich from Russia; and Profs. Pettersson and Cleve, 



each other, a blue-green border is produced. This is less intense | Dr. Trybom, Captain Maechel and Messrs. Wijkander and 



NO. 1652, VOL. 64] 



