August i8, 19 io] 



NATURE 



20: 



The lij^ht which the flower of that name sends to the 

 eye undoubtedly includes red rays. 



The apparatus employed is on the model of the first 

 described in an early paper,' the only difference worth 

 mentioning being that the side upon which the mov- 

 able slits are disposed is made oblique, to meet the 

 variation in focal length along the spectrum. By 

 this means any desired mi.xture of spectrum colours 

 can be exhibited in juxtaposition with any other. For 

 example, the violet can be shown alongside the blue, 

 and anv addition can be made to either. A few trials 

 in 1907 confirmed my anticipations, an approximate 

 match being easily attained by addition of red to the 

 blue or of green to the violet. The slits by which the 

 light entered were protected with suitable coloured 

 glasses, cobalt glass being used for the blue and violet 

 slits. In this way, as already mentioned, the danger 

 of false light is obviated. I do not affirm that the 

 mixture of blue and red looked exacth' the same as 

 the violet. I thought that I could recognise the violet 

 as being more saturated, but the difference, if real, 

 was very small and certainly a mere fraction of the 

 original difference between blue and violet. Needless 

 to say, the blue chosen was a full blue, showing no 

 approximation to green. 



The point of greatest interest lies in the contrast 

 between my observations and those of Mr. Gerald Bal- 

 four, who was with me at the time. Mr. G. Balfour 

 is one of the three brothers whom I found in 1881 

 (loc. cit.) to make anomalous matches of mixtures of 

 red and green with spectrum yellow. To effect the 

 match they use much smaller amounts of red than is 

 required by normal eyes. But their colour vision is as 

 acute as usual, and the abnormality is quite distinct 

 from what is called colour-blindness. To Mr. Bal- 

 four's vision the violet of the spectrum is not redder 

 than the blue, and such addition of red to blue as I 

 required to make the match gave, in his estimation, a 

 "reddish purple." Curiously enoueh, Mr. Knock, 

 who was mv assistant at that time, bore similar testi- 

 mony, no addition of red on either side improving the 

 match, which was indeed nearly complete as it stood. 

 It is probably not a coincidence that Mr. Knock is 

 also abnormal in his red plus green = yellow match, 

 coming perhaps about half-way between myself and 

 the Balfours. 



\\'hen a few months ago I commenced to write out 

 an account of these observations, it occurred to me 

 Ihnt it would sound strange if I described mv own 

 judgments as normal and those of two other male 

 observers as abnormal, and I sought to confirm mv 

 own judgment by that of others, especially of women. 

 As to this, there was no difficulty. I usually showed 

 first the simple blue and violet with about equal illu- 

 mination - and asked the observer to describe them. 

 In nearly every case the names blue and violet were 

 correctly civen. Can you describe one as redder than 

 the other? was the next question. In most cases the 

 answer came, " the violet is the redder " ; but in some 

 others all I could get at this stage was a negative. 

 When, however, the same addition of red light that 

 I require was made to the blue, everv female observer 

 that T have tried agreed that now the difference had 

 practicallv disappeared. I can say with confidence 

 that in this matter mv own vision is normal. 



Lately I have had another opoortunitv of repeating 

 the observation with Mr. G. Balfour. It is certain^ 

 that he sees no colour difference at all between the 

 blue and violet. When to the blue an addition of red 



1 Nature, vol. xxv., p. 64. iSSir Sci. Paoers. i.. p. 541. 



- This adjustment can be made bv partially cutting off the liaht on the 

 side required by means of strips of glass interposed. By varying the number 

 fup to 5 or 6) or inclination, the proportion of light transmitted can be regu- 

 lated. This procedure was found more convenient than altering the widths 

 of the slits. 



NO. 2129, VOL. 84] 



(less than I require for a match) is made, he describes 

 the mixture as a reddish-purple, strongly distin- 

 guished from the violet. Mr. A. J. Balfour also could 

 see no difference between the blue and violet, but he 

 seemed rather less sensitive to additions of red. A 

 determination of wave-lengths gave for the (mean) 

 violet 415 (above G), and for the blue 440. The red 

 was rather extreme. 



That ordinary normal vision is verv approximately 

 trichromic cannot be doubted ; but a question may 

 be raised as to the possible existence of a very subor- 

 dinate fourth element of colour. Thus Dr. Burch's 

 descriptions might suggest that in his vision the sensa- 

 tion of violet depended upon such a fourth element. 

 I am speaking here of fundamental sensations, not 

 of such judgments as make yellow appear a distinct 

 sensation to normal eyes, although certainly resolvable 

 into red and green. The only wav to get a final 

 answer to such questions is by making matches with 

 superposed colours ; but to this method some workers 

 seem singularly averse. In mv own case I am certain 

 that there is no fourth element of colour practicallv 

 operative. 



The character of the three primary sensations in 

 normal vision is another and a much more difficult 

 question. Perhaps in recent years we have rather 

 lost sight of the argument which weiErhed with Max- 

 well in the passage above quoted. The better to see 

 its significance, let us suppose that the spectrum is 

 accurately represented on Newton's diagram bv two 

 sides of a triangle, and incuire into the sis'nificance 

 of this disposition. The only explanation which does 

 not involve highly improbable coincidences seems to be 

 that in each spectrum colour only two of the three 

 elements are involved. If the third is involved at all, 

 how comes it to be involved in such a wav as to make 

 the spectrum straight? .And the fact that near the 

 red end variation of wave-length entails no variation 

 of colour, makes in the same direction. That the 

 green corner is rounded off and that (if it be so) thf" 

 sides, are not quite straight, may diminish, but cannot 

 destroy, the cogency of the artruiTient, while the less 

 precise character of the conclusion is not without 

 advantages. Rayleigh. 



MORE ANTARCTIC NATURAL HISTORY.' 

 (i) 'TTHK fifth, probably the penultimate, volume of 

 -•- the natural history results of the voyage of 

 the s.s. Discovery has followed its predecessors without 

 loss of time, and it resembles them in quality and 

 interest, reflecting great credit on all concerned. The 

 first memoir, by Dr. H. W. Marett Tims, deals with 

 the embryos of Weddell's seal. The author finds in 

 the musculature some additional support for Mivart's 

 suggestion of a lutrine origin for the Phocidae, and 

 he has discovered in a very early embryo what seems 

 to be the vanishing point of a vestigial external ear. 

 Prof.Herdman deals like an old hand with the small 

 but interesting collection of tunicates, comprising 

 twenty-two species, of which ten are new to science. 

 None of them are very remarkable forms in any way, 

 but they confirm the impression which other collections 



1 (i) National Antarctic Expedition, 1Q01-4. Natii'al Hi~torv. vol v. 

 Z00I02V .in t Botany. (London: British Museum [N. H.], ino.) Price 30c. 



(2) British Antnrct'c Expedition, iqo7-Q, unlet the conmand of Sir 

 E. H. Shacliletnn. C.V.O. Reports on the S-ientific Inveuigations. Vol. !., 

 Kioligy, parts i.-iv. Pn, i-jg, 13 pis., t fiirs. (London : Published for 

 the Expedition by W. Heinemann, .010.) Price 121. 6./. net. 



(3) Expedition Antarctique Beige. Riiultats du Vovage du S.V. Belncz 

 en 1897-8-9. Sous le Commandement de A. de O-rlache de r.om-ry. 

 Rapports scientifiques. Botanique— Diatomees. By H. Van Heurck. 

 Pp. 128 + 13 plates (loop), Geoloeie— Petrograohische Untersuchung der 

 Gesteinsproben, i Theil. By A. Pelilcan. Quelqnes Plantes Fo^^iles des 

 Terres Magellaniqnes. By Professor A. Gilltinet. Pp. 50+2 pis. +6 (igogl. 

 Oceanographie — Les Glaces— Glace de Mer et Banqiiis^s. Bv H. 

 Arctowski. Pp. 55+7 pis. (iqoS). Z-ioIo^i-— Schizoooda and Cuma-ea. 

 By H. J. Hansen. Pp. 23+3 pU. (tQoS). (Anveri : J. E. Bus:hmann.) 



