4o8 



NATURE 



[November 25, 1920 



be {greater than in cold ones. An extreme case is that 

 in which the temperature is so high as to produce 

 i<oneral dissociation of the electrons from the nuclei, 

 as appears to occur in the interior of stars. Apart 

 from the disturbance produced by collisions, the mag- 

 netons are then free to set their axes parallel to the 

 axis of rotation, and a large proportion of them would 

 seem to do this in certain cases (vide infra). The 

 upper limit of magnetisation is reached when all the 

 magnetons are set parallel to the axis ; this limit can 

 be calculated in some cases, it being possible to 

 estimate the number of electrons in electrostatically 

 neutral matter of given kind and density. Other 

 things being equal, the maximum magnetisation will 

 be proportional to the density and independent of the 

 speed of rotation. 



The magnetic fields of the earth and sun can thus 

 be accounted for on the present theory because these 

 bodies are at a high temperature (in the case of the 

 earth this is so except for the parts near the .surface) ; 

 what has hitherto been an obstacle in framing theories 

 of their magnetic fields is here regarded as probably 

 the determining factor as regards magnitude. 



The same explanation applies to sun-spots. These 

 occur in pairs of opposite magnetic polarity, and it 

 has been pointed out that the polarities are such as 

 would be accounted for by negative electricity in 

 rotatory motion such as is observed (with opposite 

 directions) in the members of a sun-spot pair. The 

 sun-spot pair is thus regarded as the surface portion 

 of a " line "-vortex continuously connecting the ends 

 beneath the surface. The difficulty has been that no 

 Stark effect is observable, such as would accompany 

 rotating charges of the required amount ; it has there- 

 fore been suggested that the magnetism of sun-spots 

 is due to galvanic currents, i.e. electrons in transla- 

 torv motion through electrostatically neutral matter. 

 This hypothesis is unnecessary ; it seems possible to 

 account for the observed fields, of order 3000 gauss, 

 simply by the observed rotation of the matter com- 

 posing the vortex, on the above magneton hypothesis. 



The obliquity of the magnetic axes of the earth and 

 sun requires some additional hypothesis for its ex- 

 planation, since an unsymmetrical condition cannot 

 be accounted for by a cause, like rotation, symmetrical 

 about an axis. The obliquity being present, however, 

 it seems possible to explain the secular variations of 

 the solar and terrestrial magnetic fields as due to 

 precessional motion of the magnetons. A magnetic 

 field applied obliquelv to a magneton will tend to 

 cause it to precess round the lines of external mag- 

 netic force. The magnetic fields of the earth and 

 sun, due mainly to rotation, will exert this effect on 

 anv magnetons the axes of which are inclined to 

 them ; the speed of precession depends solely on the 

 strength of the applied field if the magnetons are free, 

 and the direction of precession agrees with the direc- 

 tion of rotation of the magnetic axes of the earth and 

 sun. S. Chapman. 



The I'niversity, Manchester, November 13. 



Physiological Effect^ of Alcohol. 



During recent years a good deal of work has been 

 done and a great many results have been published on 

 the above subject. Some of these have been, very 

 recently reported in the general Press. In most, if 

 not in all, of the researches which have been made 

 on the physiological effects of moderate doses of 

 alcohol, taken in the form of beer, spirits, or other 

 alcoholic beverages, but little attention has been paid 

 to the very important disturbing influence of what 

 may be called "secondary products "—whether extrac- 

 tive matters normal to the beverages in question or 



NO. 2665, VOL. 106] 



volatile by-products of the alcoholic fenpentation 

 process itself. Most people are aware that there is 

 no definite relationship tetween the intoxicating efiects 

 of certain wines and the amounts of alcohol which 

 they contain, and the marked difference in the physio- 

 logical effects of new whisky on one hand, and of 

 well-matured spirit of the same alcoholic strength on 

 the other, is a matter of common experience. 



In the case of beer the soporific effect depends pot 

 a little on the hop and other extractives, and con- 

 sequently there is no scientific justification for driw- 

 ing any definite conclusions of a quantitative character 

 as to the physiological effects of small doses of alcohol 

 unless the beer used in the whole series of experiments 

 had been of precisely the same character throughout, 

 and even then the "personal equation" would intro- 

 duce a further serious difficulty. Most beer-drinkers 

 of middle age are well aware that the physiological 

 effects of bitter beer on one hand, and of mild ale or 

 lager (which are less heavily hopped) on the other, 

 are not by any means the same, even for equal per- 

 centages of alcohol. The investigation of the effects 

 of these various "secondary " constituents undoubtedly 

 presents great experimental difficulties, but until their 

 physiological significance is better understood there 

 must always be an element of uncertainty in any con- 

 clusions arrived at in regard to the physiological 

 effects of small doses of alcohol taken in the form of 

 the usual alcoholic beverages. 



A. Chaston Chapman. 



Atomic Structure. 



The arrangement of the non-nuclear electrons in 

 the atom is being determined by work along two 

 lines : first, from the Bohr-Sommerfeld theory of 

 characteristic frequencies, and, secondly, from the 

 Lewis-Langmuir or Born-Lande theory of the struc- 

 ture of molecules, supported by Prof. VV. L. Bragg's 

 work on atomic radii in crystals. It has been stated 

 several times recently that the conclusions drawn 

 from the two sources are mutually inconsistent, in 

 that the first indicates that the electrons are revolving 

 in planetary orbits, and the second that they are fixed 

 in constant positions. The essential truth of the first 

 theory is now be\'ond doubt ; the second is extremely 

 plausible. If they are really inconsistent the position 

 would be intolerable. 



But they are not really inconsistent. The Bohr- 

 Sommerfeld theory does not make explicit use of the 

 assumption that the electrons in their stable states 

 are moving; it assumes only that in those states 

 thev have the energy, calculated by Hamiltonian 

 (relativity) dynamics, which they would have if they 

 were moving in certain orbits. It is not logically 

 impossible to maintain that they have that energy 

 and are j'et at rest. Nor is it physically impossible 

 if we accept Bohr's principle of "correspondence," 

 which has been so astoundinglv successful in explain- 

 ing the Stark effect and in predicting the number of 

 components in lines of the hydrogen and helium 

 spectra. According to that principle, the intensity 

 and polarisation of components can be predicted by 

 the application of classical dynamics to certain 

 assumed orbits, although it must be assumed at the 

 same time that the electrons are not moving in those 

 orbits. If intensity and polarisation can be predicted 

 from orbits that are wholly fictitious, why not energy? 



Of course, the adoption of Bohr's principle in this 

 extreme form would sever the last connection between 

 classical dynamics and the real structure of the atom. 

 But we are surely now all convinced that we must 

 abandon part of that connection. Is there any reason 



