6 3 2 



NA TURE 



[November i i , 1922 



Letters to the Editor. 



[The Editor docs not hold himself responsible for 

 opinions expressed by his correspondents. Neillier 

 can he undertake to return, or to correspond with 

 the writers of rejected manuscripts intended for 

 this or any other part of NATURE. No notice is 

 taken of anonymous communications.] 



The Structure of the Red Lithium Line. 



In a recent number of the Proceedings of the Royal 

 Society Prof. McLennan and Mr. Ainslie have 

 announced the interesting discovery of a new com- 

 pi ment of the line X =6708 in the spectrum of lithium, 

 the line appearing, under the conditions of excitation 

 employed by them, as a quartet. They proceed to 

 discuss the possibility of this structure being due to 

 two pairs of lines, each pair being assigned to one of 

 the isotopes of lithium. To the present writer it 

 appears that the new components cannot be accounted 

 for in this manner. 



The structure of the line in question has been 

 investigated by Kent (Astrophys. Journ. vol. 40, 

 P- 337. T 9'4). Takamine and Yamada (Proc. Tokio 

 Math. Phys. Soc, vol. 7, No. 18, p. 339, 1914), Zeeman 

 (Proc. Roy. Acad. Amsterdam, p. 1130, Feb. 1913; 

 p. 155, Sept. 1913), King (Astrophys. Journ., vol. 44, 

 p. 169, 1916), and the writer (Proc. Roy. Soc. A, vol. 

 99, p. 101, E921). Kent, and Takamine and Yamada, 

 observed it as a single pair of emission lines, and 

 Zeeman, who investigated the absorption spectrum, 

 also recorded a single pair of lines, with the reserva- 

 tion that with a high density of the absorbing vapour 

 other lines made their appearance. Zeeman con- 

 sidered that these lines which appeared at high 

 vapour densities were analogous to lines observed 111 

 the sodium spectrum by R. W. Wood. King, who 

 investigated the structure of the line in the arc and 

 in the tube-furnace, found that with a low vapour 

 density the line appeared as a simple pair, and that 

 at higher vapour densities a third component 

 appeared ; with a still greater amount of vapour the 

 phenomena were complicated by reversal. King has 

 published one photograph in which, owing to reversal, 

 the line has the appearance of a quartet. McLennan 

 and Ainslie used a vacuum arc under conditions in 

 which it would appear that the density of the lithium 

 vapour must have been very great, and one may 

 surmise that this condition is essential for the appear- 

 ance of the fourth component. 



It seems, however, that under appropriate condi- 

 tions the line appears as a simple pair, and our ideas 

 as to the nature of isotopes would have to be pro- 

 foundly modified if the pairs due to the two isotopes 

 were found to require different conditions for their 

 excitation. If the four components were really two 

 pairs due to the two isotopes they should always 

 appear together with an invariable intensity ratio of 

 1 : 16. The line can be seen easily as a simple pair 

 in a carbon arc in air if the poles are brushed over 

 with an exceedingly dilute solution of a lithium salt. 

 The components are then less sharp than when the 

 vacuum arc is used, and the main difficulty is to have 

 little enough lithium in the arc, so as to avoid the 

 appearance of the third component and complex 

 structures dm to reversal. In the vacuum arc the 

 third component appears very readily unless the 

 amount of lithium vapour is small. It may further 

 be mentioned that the relative intensities of the 

 components are not in good accordance with the view 

 that they are due to the two isotopes. 



From a theoretical point of view also, there are 

 grave difficulties. The calculated separation, on 



NO. 2767, VOL. I IO] 



Bohr's theory, of corresponding lines in the pair, is 

 about 0.087A., the observed separation being between 

 three and four times as great. McLennan and Ainslie 

 put forward the suggestion that the separation may 

 in fact be the product of the " calculated separation " 

 and the atomic number ; but the correctness of the 

 calculated separation has been verified by the 

 observed differences between the lines of the Balmer 

 series of hydrogen and alternate members of the 

 f Puppis series of helium, and in this case the agree- 

 ment is exact and the " calculated separation " does 

 not require to be multiplied by a factor of 2, the 

 atomic number of helium. T. R. Mertox. 



The Clarendon Laboratory, Oxford, 

 October 19. 



The Mechanism of the Cochlea. 



In view of the discussion in these columns towards 

 the end of 1918, and the letters which followed it at 

 various times, the model designed by Mr. George 

 Wilkinson, of Sheffield, and described in Nature of 

 October 21, p. 559, is of much interest and importance. 

 It is obvious that the construction of such a model 

 presented many mechanical difficulties, and great 

 credit is due to Mr. C. E. Stewart, the mechanician of 

 Prof. Leathes's laboratory, for the successful result. 

 It may, therefore, be useful to mention that a full 

 description was published in the journal of Lai yn- 

 and Otology, of September last, a short account 

 having been given in the Proceedings of the Physio- 

 logical Society (Journal of Physiology, vol. 56, p. ii). 

 The apparatus was demonstrated to the Physiological 

 Society in December 1921, as also to the British 

 Association in September 1922. 



I take it that others besides medical students have 

 been dissatisfied with most of the theories put forward 

 to avoid the difficulties thought to be involved in the 

 Helmholtz view of the resonance of the basilar 

 membrane. Those theories in which this membrane 

 is supposed to act as a whole, like a telephone dia- 

 phragm, or by " pressure patterns," are inconsistent 

 with the progressive differentiation of structure along 

 the membrane, in addition to being in conflict with 

 what is known of the conducting properties of nerve 

 fibres. Thus the views suggested by Ewald, Ruther- 

 ford, Waller, and Wrightson are unacceptable. It 

 appears that although Helmholtz had referred 

 incidentally to " loading " of the vibrating elements 

 of the membrane by the liquid in which it lies, the 

 i< it importance of this factor was first realised by 

 Mr. Wilkinson and investigated experimentally by 

 him. His model is doubtless capable of still further 

 improvement, but even in its present form many 

 problems would have light thrown upon them by Us 

 behaviour. The degree of damping and the spread 

 of resonance to neighbouring elements may be 

 mentioned. The number of waves required to excite 

 sympathetic resonance of a tuned element may 

 perhaps be determined. Some degree of spread is 

 not inconsistent with the Helmholtz theory, since the 

 amplitude of vibration of other elements than those 

 in tune with the vibrations received might well be 

 too small to stimulate the nerve endings. Dr. Gray 

 has shown that a similar cutting out of small stimuli 

 takes place in the localisation of a point of pressure 

 in the skin. 



It is of interest to note that the model responds to 

 a tuning-fork held in contact with the brass case, just 

 as the cochlea does to conduction through bone. 

 This indicates that the impulses given by the move- 

 ments of the stapes are the same as those of sound 

 waves directly transmitted through water, as would 

 be expected from theoretical considerations. 



