362 



NA TURE 



[September 8. 1923 



inferioiii\, th good character of Eskdalemuir seems 

 difficult to I \j)lam, as the value there makes a much 

 closer approach to the Kew than to the Ebro value, 

 the latter being notably below what Dr. Bauer puts 

 forward as the normal. Again, if a high value of 

 c,/Ci is a sign of inferiority, is it not strange that c,/c, 

 is highest at Kew in summer when the potential 

 gradient is lowest ? Ebro and Eskdalemuir have a 

 variable number of monthly quiet days, while at 

 Kew with rare exceptions the number is uniform. 

 Weather conditions usually reduce the number of 

 quiet days used at Eskdalemuir below the Kew 

 number lo. Thus a priori we should have expected 

 Kew to be the station least affected by accidental 

 irregularities. According to Dr. Bauer the sign of 

 c' (his t) " may depend upon whether the sun-spot 

 cycle ... is below or above average development." 

 Apparently he expects a revolutionary change from 

 a steady fall to a steady rise and conversely 1 It is 

 obvious that if a steady fall did go on at the Ebro at 

 the rate obtained by Dr. Bauer we should before long 

 have the potential gradient negative. 



The fact that Dr. Bauer finds negative values for c' 

 at all three stations, Ebro, Eskdalemuir and Kew, may 

 possess some physical significance unrelated to sun- 

 spots. In my Physical Society paper I referred to 

 volcanic dust as a possible natural agency influencing 

 potential gradient over wide areas. Even the agency 

 of man may influence a considerable area. Thus I had 

 myself regarded the value for 192 1 at Kew as excep- 

 tionally low, and attributed this at least in part to the 

 abnormal purity of the English atmosphere brought 

 about by the coal strike. At all events the mean 

 value for 1922, unlike that at the Ebro, shows a 

 substantial rise. 



In view of Dr. Bauer's concluding remarks it may 

 not be amiss to point out that the earth's atmo- 

 sphere is generally believed to contain an equal and 

 opposite charge to the earth's surface. Thus the total 

 charge on the earth as a planet would seem to be nil 

 whether a sun-spot influence exists or not. 



C. Chree. 



August 17. 



Colour Vision and Colour Vision Theories. 



In his letter published in Nature of August 25, 

 Dr. Edridge-Green seems to admit the accuracy of 

 the deductions from the trichromatic theory which I 

 made in the issue of August 4. But, in making these, 

 I used no other postulate than that of the fact of 

 normal trichromasy. Iji the sense in which the word 

 is used, trichromasy is now a qualitatively and 

 quantitatively proved fact, although at the time of 

 its first assertion it was in considerable part hypo- 

 thetical. Strict logical development (which may be 

 mathematical when necessary, since mathematics is 

 merely symbolised logic from this point of view) leads 

 directly to the explanation of certain phenomena 

 which Dr. Edridge-Green had thought to be un- 

 explainable on the basis of trichromasy. If the 

 logical developments are sound, the conclusions are 

 inevitable. But he brings forward three other facts 

 which he still considers to be inexplicable on the 

 theory. 



First ; a man, stated to be completely " red-blind," 

 can recognise red as easily as a normal-sighted person. 

 From the trichromatic point of view one might 

 say. Why not ? No doubt the term " red -blind " 

 might preferably be avoided, seeing that it is a reUc 

 of the " hard-atom " stage of the theory- ; but the 

 theory' does not give the result that a dichromat of 

 that type cannot distinguish red light from other 

 lights. The notion that it must do so is a survival 



of ideas held under the restrictions of the eaI^ 

 applications of the theory. 



Second ; 50 per cent, of the dangerously colour 

 blind get through the wool test. Again. Why not ? 

 The theory would only use the fact, if g^nted, to aid 

 in further elaboration of the details of the visual 

 peculiarities. 



Third : the theory is said to fail to explain the 

 class of colour vision which Dr. Edridge-Green denotes 

 as trichromatic, in which yellow is not recognised, 

 the region of the spectrum occupied by yellow hues 

 being called red-green. I cannot occupy space here 

 in showing how this is directly predictable as a 

 possibility on the trichromatic basis. I have discussed 

 it, and other such cases, in my book on colour vision. 

 Dr. Edridge-Green says that, in this case, the inter- 

 section of the dichromatic curves should be shifted 

 towards the red on the trichromatic theory, and they 

 are not so shifted. The statement is mistaken. 

 There is no such compulsion on the theory. 



The statements in Dr. Edridge-Green 's last two 

 sentences are in complete agreement with the theory. 

 He says also that the theory is burdened with self- 

 inconsistent subsidiary hypotheses. Actually the 

 theory is based, and based alone, on two postulates ; 

 the qualitative postulate of trichromasy, and the 

 quantitative postulate of the intensity law. All 

 further development is straightforward, any definite 

 constructive presumption being used in illustration 

 only, and being clearly stated by Helmholtz to be 

 quite inessential. In fact, he left the theory totally 

 unburdened with fixed presumptions regarding 

 structure and function. The fixation was to come 

 later, probably by way of many supplemental theories 

 consonant with it. AH, including the views of Dr. 

 Edridge-Green, may possibly help. 



I would appeal to Dr. Edridge-Green not to pit 

 his views against the trichromatic theory, but rather 

 to consider wherein they may supplement it. Multi- 

 chromasy higher than triple is without evidence. 

 If he accepts Dr. Houston's work as the mathematical 

 expression of his views, he thereby makes them 

 trichromatic in the usual sense of the term. His 

 views may supplement the theory on the side of 

 functional physiology or psychology ; they cannot 

 refute it on the formal side. W. Peddie. 



August 25. 



NO. 2810, VOL. 112] 



The Phosphate^Deposit of Ocean Island. 



On p. 787 of Nature of June 9, which has just 

 reached me, a notice appears, under the heading of 

 " Mineral Fertilizers," of my paper on " The Phos- m 

 phate Deposit of Ocean Island " (Quart. Joum. S 

 Geol. Soc, vol. Ixxix., p. i, 1923.) 



As this notice misinterprets certain of the state- 

 ments made in the paper, I beg the courtesy of your 

 space for the necessary corrections. 



(i) One of the points emphasised in the paper is the 

 gradual and uniform change which occurs in the 

 composition of the deposit as one passes from peri- 

 meter to centre. This change is so regular that it 

 can be expressed by a simple formula. 



There is no normal 88 per cent, and no " level . . . 

 where the phosphate sinks from its normal 88 to 

 79 per cent.," the change being gradual and without 

 break from 79 to 92 per cent. 



(2) The deposit cannot be truly described as 

 having " a depth of fully fifty feet." As stated in 

 the paper, it is sometimes as much as 80 feet thick, 

 but usually less than 50 feet. 



(3) The excess lime shown by analysis (i.e. the Ume 

 over and above that required for the phosphoric, 

 carbonic, fluoric, and sulphuric acid radicals) varies 



