208 



NATURE 



[October 13, 1921 



Letters to the Editor. 



\Th& Editor does not hold himself responsible for 

 opinions expressed by his correspondents. Neither 

 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.] 



Occurrence of the Aurora Line in the Spectrum of the 

 Night Sky. 



I HAVE pursued the line of work outlined in my 

 letter to Nature of March 31, 192 1 (vol. 107, p. 137). 

 The result has been to show that at Terling, Essex, 

 the aurora line can be photographed on two nights 

 out of three. Exposures were made on 150 nights, 

 irrespective of weather. 



The intensity on ordinary occasions appears to have 

 little or no connection with magnetic disturbance or 

 the distribution of spots on the sun. The most in- 

 teresting point that has come out, however, is that 

 the aurora line is much stronger at Terling than at 

 Beaufront Castle, near Hexham, Northumberland. I 

 have, made exposures on twenty-six different single 

 nights at the latter place, and have never found a trace 

 of the aurora line on any of them, though the same 

 instrument and the same kind of plates were used as 

 at Terling. 



Positive results at Terling were sandwiched in 

 between the negative results at Beaufront ; thus the 

 latter cannot be attributed to seasonal variation. 



Five nights' cumulative exposure have been tried on 

 two occasions at Beaufront, and on each plate the 

 aurora line was obtained. 



I have been very much astonished at this diminished 

 intensity of the aurora line as one goes north. The 

 difference of latitude is about 3°. It would seem 

 that the aurora line as photographed In the south of 

 England will not fit into the scheme of distribution 

 of the polar aurora. I hope to pursue this line of 

 work to the north and to the south as opportunity 

 mav offer. Rayleigh. 



October g. 



Atomic Structure. 



In connection with the problem of the constitution 

 of the atom discussed in my letter to Nature of 

 March 24 last (vol. 107, p. 104), I should like to add 

 a few complementary remarks about the manner in , 

 which the orbits of the electrons in the atom are 

 characterised. 



According to this view of atomic constitution, the 

 electrons in the atom are arranged in groups in such 

 a w;ay that the orbit of every e'ectron within one 

 and the same group is characterised by the same 

 total number of quanta. Since, however, for orbits 

 characterised by more than one quantum there exist 

 several types of orbits possessing the same total 

 number of quanta, the electrons within each group do 

 not in general play equivalent parts, but are divided 

 into a number of sub-groups corresponding to different : 

 types of possible orbits. Now it is a salient feature 

 of this picture that the atom cannot be said to be 

 composed of a number of well-defined spherical shells ' 

 of electrons moving in sharply separated regions of 

 the atom. In fact, although the electrons of a given 

 group mainly move within one and the same shell- 

 shaped region of the atom, the electrons, at any rate • 

 of certain sub-groups, will in their revolution pene- \ 

 trate into the region of the orbits of the electrons . 

 of inner groups. This gives rise to a coupling 

 between the various groups, which is an essential 



NO. 27 II, VOL. 108] 



feature of the interpretation of the stability of the 

 atom. As a consequence of this, the orbit of an 

 electron may be considered from different points of 

 view, according as attention is mainly paid (i) to the 

 larger part of the orbit which lies outside the region 

 of inner groups, and which nearly coincides with an 

 almost closed Keplerian ellipse, or (2) to the 

 mechanical properties of the whole orbit, regarded as 

 a type of central orbit composed of loops which only 

 in their outer part possess an approximately Kep- 

 lerian character. 



Now in the classification described in my former 

 letter the orbits were regarded from the first, and 

 more superficial, point of view. The numbers of 

 quanta characterising the orbits of the electrons in 

 the different groups correspond to Keplerian ellipses, 

 which coincide approximately with the outer parts of 

 the orbits of the electrons in question. It has since 

 been possible, by a detailed examination of the parts 

 of the orbital loops situated within the region of 

 inner groups, to classify the orbits from the second, 

 and more fundamental, point of view, leading to a 

 simple and unambiguous result. In fact, we are led 

 to a classification in which, when we proceed out- 

 wards from the nucleus, the number of quanta charac- 

 terising a certain group of orbits is always larger by 

 one unit than that of the preceding group. For the 

 groups in the inner region of the atom, where the 

 attraction of the nucleus preponderates, this new 

 rigorous classification coincides with the old one of 

 my former letter. But it departs from the old _ for 

 groups in which the orbits of the electrons mainly 

 fall in the outer region of the atom, where the attrac- 

 tion of the nucleus is largely compensated for by the 

 repulsion of the electrons in the inner groups. For 

 these groups the quantum numbers of the orbits given 

 in my former letter were equal to, or even smaller 

 than, those of inner groups. 



Notwithstanding the essential progress made by this 

 modification in the classification of the orbits, the 

 main features of this model of the atom remain the 

 same. For instance, my former statements of the 

 numbers of electrons in the various groups and sub- 

 groups in the atom hold unaltered for all groups. In 

 fact, in fixing these numbers by the correspondence 

 principle we find them to depend on the harmony of 

 the motion of the electrons within each single group. 

 They depend, therefore, primarily on the relative 

 dimensions of the approximately Keplerian loops, and 

 only secondarily on the way in which these loops are 

 joined together to form complete central orbits. Thus 

 the previous model of the atoms of the inert gases 

 holds unaltered also as regards the outer groups, 

 provided that the numbers stated as defining the 

 number of quanta of the orbits in the various groups 

 are considered instead as defining the number of sub- 

 groups within the corresponding groups. Moreover, 

 the numbers in question offer an approximate esti- 

 mate of the spatial extension of the. regions of the 

 orbits of the electrons in the different groups in the 

 atom. For instance, the orbits in the outermost 

 " shell '* in the Niton atom must be characterised as 

 six-quanta orbits instead of as two-quanta orbits ; but 

 the dimensions of the orbital loops will by no means 

 be of the same order of magnitude as those of the 

 orbit of an electron revolving in a Keplerian orbit 

 characterised bv six quanta in the region outside that 

 of the orbits of the electrons in the five inner groups ; 

 they will rather be of the same order as those of a 

 similar Keplerian orbit characterised by only two 

 quanta. 



From these remarks it will be seen that my former 

 applications of the theory to the interpretation of 

 the physical and chemical' properties of the elements 



