382 Mr. H. R. Droop on 



The portion with crossed vertical and horizontal lines repre- 

 sents the extent to which the same rays operate on both 

 the yellow and the blue colour-senses, and thus produce 

 white light, which (combining with the colour produced by 

 the colour-sense which is most affected) produces a -whitish 

 yellow or a whitish blue. Similarly the portion crossed by 

 diagonal lines in the lower half of the diagram represents the 

 extent to which the same rays operate on both the red and the 

 green colour-senses. 



I have represented the complementary colour-sensations 

 as thus overlapping*, because Professors Preyer and Hippel 

 have found that, for red-green blind eyes, the neutral point 

 in the spectrum which appears white or grey varies in posi- 

 tion according to the intensity of the light (Pfliiger's Archiv, 

 vol. xxv. ; Graf's Archiv fur Ophthalmologic, vol. xxvii. 

 pt. 3). Professor Preyer found that enlarging the aperture 

 through which the light was admitted from *250 millim. to 

 •370 millim. shifted the neutral point, where the spectrum 

 appeared grey, from where the wave-length was 512'8 

 million ths of a millimetre to where it was 506' 6 millionths. 

 This is readily intelligible, if the rays in this part of the spec- 

 trum affect both the yellow and the blue colour-senses, while 

 the intensity of the light alters the proportions in which they 

 are respectively affected by it. 



It will be observed that a narrow strip of yellow extends 

 nearly to the red end of the spectrum. Professor Holmgren 

 (' Proceedings of the Royal Society ') states that the red seen 

 by the violet-blind eye, in his case of one-sided violet-blindness, 

 is not quite identical with the common spectral red of the 

 normal eye, but rather a clearer red having a shade of carmine, 

 about the same as the red towards the end of the subjective 

 spectrum of the normal-eyed. This colour would obviously 

 require a slight admixture of yellow to reduce it to the common 

 spectral red of the normal-eyed. Moreover the extension of 

 the yellow nearly to the red end of the spectrum explains how 

 it is that a great many persons who are red-green colour-blind 

 can see almost to the red end of the spectrum. 



The strip of red at the violet end of the spectrum also re- 

 quires explanation. The colours seen by the red-green blind 

 eye in the case of one-sided red-green blindness are not yellow 

 and violet, but yellow and indigo with only a faint shade of 



* The extent to which the colour-sensations are represented as over- 

 lapping rests on conjecture. Observations on colour-blind persons, i. e. 

 persons with only two colour-sensations, give readily the neutral points 

 where the sensations counterbalance each other and produce white ; but 

 the extent to which they overlap could only be inferred from observing 

 when the colours cease to have anv admixture of white. 



