1 66 



NATURE 



[August 4, 1923 



molecule in space, and hence, in finding the total 

 components in these directions, we have to add the 

 intensities, not the amplitudes. A fluid consisting 

 of anisotropic molecules oriented at randorn must 

 therefore necessarily scatter unpolarised light in 

 proportion to its density, and as remarked in my letter 

 m Nature of March 31, p. 428, considerations similar to 

 those which enter into the Lorentz refraction formula 

 introduce a further factor (m* + 2)V9. which increases 

 the unpolarised scattering to be expected. The whole 

 question will be found elaborately discussed in a paper 

 by Mr. Ramanathan in the Proc. Indian Association 

 for the Cultivation of Science, vol. viii., Part I., just 

 published. 



I think I should make it clear that the suggestion 

 made in my letter in Nature, March 31, and endorsed 

 with some modifications by Sir William Bragg, re- 

 garding the relations between the liquid and the 

 crystalline states, is very different from that put 

 forward by Prof. King. In my opinion, neither the 

 facts regarding the scattering of light nor the X-ray 

 data require the assumption of the existence of crystal- 

 line aggregates in liquids. All that the experimental 

 facts suggest is that the molecules in a liquid influence 

 the orientations of their nearest neighbours to a 

 sensible extent, and that this results in the amount of 

 unpolarised light scattered being somewhat smaller 

 than on the hypothesis of random orientations of the 

 molecules. 



C. V. Raman^ 



210 Bowfcazaar Street, Calcutta, 

 June 15. 



The Doublet Separations of Balmer Lines. 



In his theory of the structure of the lines of the 

 Balmer Series based on the principle of relativity, 

 Sommerfeld shows that each of the members of the 

 series should consist of a doublet and that each of 

 the components of these doublets should possess a 

 fine structure. The calculations show besides that 

 the frequency difference for these doublets should be 

 constant over the whole of the Balmer Series and 

 should be equal to 0-36 cm"^. For Ha the separation 

 should perhaps be slightly less. As the theory 

 applies equally well to the doublets of the corre- 

 sponding series in the spectrum of positively charged 

 helium, these were investigated by Paschen and were 

 found to have separations that lead to a value of 

 0-3645 ±0-0045 for the frequency difference of the 

 doublets of the Balmer Series. 



Since the publication of Paschen's work on helium 

 a number of investigators, including the writer, have 

 attempted from the measurements on the separations 

 of Ha and H;3 and in some cases of Hy and Hs to look 

 for evidence that would lead to a confirmation or rejec- 

 tion of Sommerfeld's theory. Up to the present the 

 results obtained could not be considered as satis- 

 factory. There was a lack of agreement in the values 

 obtained for the separations by different investigators, 

 and on the whole the values obtained were less than 

 that demanded by the theory. In the case of the 

 observations made by myself and Mr. Lowe on the 

 separations of Hy and Hg, values were obtained that 

 seemed to point in the direction of a steady decrease 

 in the frequency differences as one passed to the 

 higher members of the series. 



At my suggestion the matter was re-investigated 

 recently by one of the research workers in the 

 Physical Laboratory of the University of Toronto, Mr. 

 G. M. Shrum. In his experiments the tubes were of 

 a special design and were cooled with liquid air. 



His method of operating these tubes, which will be 



NO. 2805, VOL. 1 12] 



described later in his own paper, enabled him to 

 eliminate practically the whole of the secondary 

 spectrum and thus permitted him to include in the 

 measurements of the doublet separations that of H. 

 as well as those of Ha Hs Hy and H«. 

 The results are the following : 



It will be seen that as far as the doublet separations 

 are concerned, they afford a striking confirmation of 

 Sommerfeld's theory. J. C. McLennan. 



The Athenaeum, 

 July 2. 



"Guide to the Mollusca." 



With reference to the review of the " Guide to the 

 Mollusca" in Nature of July 21, p. 93, may I be 

 allowed to point out that our rather cautious state- 

 ment, " A species of Helix has been said to tolerate 

 a temperature of -i2o°C.," was based on Pictet's 

 paper " De I'emploi methodique des basses tempera- 

 tures en biologic " (Arch. Sci. Phys. et Nat. Geneve 

 (3) XXX., 1893, pp. 293-314). The reviewer's remark 

 about the scientific names of the pearl mussel and 

 the pearl oyster scarcely makes it clear that we are 

 simply keeping to the names used by the late Mr. 

 E. A. Smith in 1908, Margaritana margaritifera for 

 the mussel and Margaritifera margaritifera for the 

 oyster. I hope that the other errors he has dis- 

 covered are not more serious than these. 



C. Tate Regan 

 (Keeper of Zoology). 



British Museum (Natural Histor>'), 



Cromwell Road, S.W\, 



July 23. 



Mr. Regan is quite right to direct attention to the 

 fact, which I should have noted, that the confident 

 statement in the text-book concerning the survival of 

 a species of Helix submitted to a temperature of 

 - 120° C. had been altered from " has been known " 

 to " has been said," but I still think it would have 

 been better to have omitted it altogether. Pictet in 

 his paper does not say whether the degrees he cites 

 were registered by any one of the more usual ther- 

 mometers or by a scale of his own (the " C" is an 

 addition in the text-book), and his paper altogether 

 does not suggest that amount of accuracy which the 

 subject demanded. The admission that a system of 

 nomenclature nearly a quarter of a century old has 

 been deliberately adhered to in a work supposedly 

 brought up-to-date, speaks for itself. Much progress 

 has been made in this section of systematic zoolog>^ 

 since 1908, and according to all the Rules the pearl 

 oyster (Pinctada) has no right to the name Mar- 

 garitifera, which belongs to the pearl mussel. There 

 are other examples in the " Guide " of what a 

 malacologist of to-day would call misnaming. 



The Reviewer. 



