Oct. 30, 1879] 



NATURE 



627 



to your readers a recently discovered regular secular period in 

 one of the meteorological elements of Calcutta ; a period too, 

 (though this is at present a matter of secondary importance) which 

 decidedly favours the reverse hypothesis to that entertained by 

 Prof. Smyth regarding the variation of solar energy. The 

 following figures have been worked out, and communicated to 

 me, by Prof. S. A. Hill of Allahabad, and he has I believe 

 given his conclusions from them and similar results, in a recent 

 number of the Austrian Zntschrift fiir Meteorologie, which, 

 however, I liave not as yet seen. 



The tatle which follows, shows the annual range of mean 

 monthly barometric pressure at Calcutta, from 1840 to 1878 

 inclusive, bloxamed in a series of eleven years, the average 

 length of a sun-spot cycle, beginning with the year of sun-spot 

 minimum. 



Calcutta. 

 Years of Annual range of mean 



cycle. monthly pressure in 



""""^ decimals of an inch. 



II ) ^ , . . f "530 



J f Years of mmimum ) . c .« 



2) ^""-^p"' 1 .■.'.■ .;.' ^ 



3 "Sio 



4 '499 



5 ) Years of maximum I '502 



6 ( sun-s]X)t I '502 



7 ^500 



8 ^06 



9 •512 



10 •514 



11 '53° 



The figures for Roorkee, from 1864 up to the present time, 

 give a fimilar result. So far then as we have gone at present in 

 India, we find years of few sun-spots characterised by higher 

 temperatures, greater wind-velocity, and greater range of 

 barometric pressure than those of many spots. The terrestrial 

 effects of a " languid " sun are therefore strikingly like those of 

 an unusually hot sun. E. Douglas Archibald 



Grosvenor House, Tunbridge Wells, October 18 



Colour-Blindness 



AVhile the 'subject of colour-blindness is before your 

 readers, the present seems a favourable opportunity for calling 

 attention to a method of experimenting which I used some years 

 ago ' for testing normal vision, and which seems, if applied to 

 colonr-blind eyes, likely to be capable of telling us something of 

 the nature of that peculiar form of colour sensation. 



When I made my experiments on normal eyes I intended 

 extending the investigation to colour-blind eyes, but most 

 unfortunately I was quite unable to find a true case of colour- 

 blindness. All the cases reported to me proved, on examination, 

 not to be produced by colour-blind eyes at all, but to be the 

 result of want of observation and knowledge, as they all could 

 distinguish between different colours, when placed alongside each 

 other, and could also arrange the different colours, though when 

 shown colours separately they made dreadful mistakes in naming 

 them. 



'Hie method adopted in my experiments was as follows : — A 

 prismatic spectrum was produced by passing a beam of light 

 through a slit, a lens, and a bisulphide of carbon prism, in the 

 usual way. The spectrum was thrown on a large number of 

 rectangular reflectors, placed close to each other, and all capable 

 of being moved so as to throw the light reflected from them to 

 any point on a screen in front. With this apparatus we have 

 the means of testing what colours can be produced by mixing 

 others, and what colours cannot be so produced — by throwing 

 the light reflected by one of the reflectors on the screen and 

 trying if it is possible to match it by combinations of rays from 

 other parts of the spectrum. It is found that for the normal 

 eye the same sensation which is f reduced by the yellow part of 

 the spectrum can lie produced by mixtures of rays from the red 

 and green parts, and also by rays from parts lying between these 

 colours and yellow. And that the sensation which we call blue, 

 can be produced by the blue part of the spectrum or by mixing 

 rays from each side of the blue, that is by mixtures of violet and 

 green. The yellow and blue are, however, the only two parts 

 of the spectrum, the sensation of which can be imitated by 

 combining rays from other parts of the spectrum. We cannot, 

 • /Vw-wrf/ojfj of the Royal ScDtiish Society of Aru, 1871-a. 



for instance, produce green by any mixtures of rays firom 

 other parts of the spectrum. The red and the violet sensations 

 are also incapable of being produced by mixtures. 



These results are, to a certain extent^ a proof of the threefold 

 nature of our colour sensations. And they also show us that it 

 is a mistake to talk of colours as simple and compound, as all 

 the colours we find in nature are compounded of rays of many 

 different rates of vibration. The difference between different 

 colours is, those of one rate of vibration, say those of the D-line, 

 even though absolutely pure, are capable of exciting a com- 

 pound sensation, namely, the red and green, while mixtures of 

 rays from each side of the line B, are only capable]of giving rise 

 to a simple sensation — namely, the red. 



Supposing this three-sensation theory to be true, then there 

 are certain conceivable variations of it which would give rise to 

 colour-blindness. The blindness, for instance, might be pro- 

 duced by two of the three sensations being very similar. This 

 does not seem improbable when we consider that, to any two 

 persons, with normal eyes, the different colours will not neces- 

 sarily appear equally different, and that, in the same normal eye, 

 the different simple sensations are not separated by equal differ- 

 ences from each other. That is, supposing our sensations of the 

 three primary colours to be represented by the three angles of a 

 triangle, then the triangles, if drawn to the same scale, would be of 

 different sizes for the eyes of different persons, and for almost all 

 eyes the triangles would not be equilateral. The side between the 

 green and the violet would be shorter than the other two, because 

 the sensation of green is more similar to the sensation of violet, 

 than green is to red or red is to violet. Or we might conceive 

 the colour-blindness to be produced by the different sensations 

 being irregularly, or by being too widely, distributed over the 

 spectrum. If, for instance, the green sensation extended info 

 the red part of the spectrum and the red sensation into the green 

 part, that is, if the same rays excited both sensations in the same 

 proportion, not only in certain parts, but throughout their entire 

 range, then an eye, so constructed, would be incapable of dis- 

 tinguishing red from green. Another way in which colour- 

 blindness might result, is by an absence of one of the three 

 sensations. 



It is impossible, without experimenting on colour-blind eyes, 

 to say whether any of these, or some other, is the true cause of 

 colour-blindness, and it is very desirable that some one, accus- 

 tomed to make colour observations, would test colour-blind eyes 

 in the way suggested ; it would settle at once, for the particular 

 eyes experimented on, whether they are badly defined trichroic 

 eyes or are dichroic. If the eyes are dichroic, then, clearly, 

 there will be only one part of the spectrum, the sensation of 

 which can be produced by mixtures of rays from other parts, 

 and not two as in trichroic vision. 



Besides the apparatus described many others, more accurate, 

 might be constructed, but the great advantage of this arrange- 

 ment is, that it is suited for testing eyes not accustomed to make 

 accurate observations or to be trammelled with elaborate appara- 

 tus. If Prof. Pole was to undertake the investigation, he could 

 easily devise some simple apparatus to suit the experiments 

 which, in his hands, would probably give some valuable results. 



Darroch, Falkirk, October 7 John Aitkkn 



Subject-Indexes to the Royal Society Catalogue of 

 Scientific Papers 



As yon have opened your columns to Mr. Garnett's valuable 

 paper on " Subject-Indexes to Transactions of Learned .Societies," 

 you will perhaps allow me to make a suggestion in regard to the 

 proposal contained in it. The initial objection to Mr. Garnett's 

 scheme appears to me to be that the work he suggests will really 

 be as large as the original catalogue, and, in fact, the same work 

 in a new order. Even were it possible to get the money (pro- 

 bably little short of 10,000/.), the question would naturally arise 

 whether or no the result was likely to be worth this great outlay. 

 Moreover, the plan proposed by Mr. Garnett would not meet the 

 great difficulty of compilation, which consists in the getting 

 together of papers treating of identical subjects, but written with 

 various titles by different persons. This would make it necessary 

 to employ experts in each subjfct, and also a general practical 

 editor for the whole, under svhose eye all entries must pass. I 

 cannot help thinking, therefore, with Mr. J. B. Bailey (p. 580), 

 that the titles of the papers would have to be generally ignored. 



If the index were made as indexes to catalogues are usually 

 compiled, it might be got into at least a third of the space of the 



