A.— MATHEMATICAL AND PHYSICAL SCIENCES. 41 



monovalent, and its spectrum contains relatively few arc lines which have 

 not been classified in the system of doublets. The spectrum of copper, 

 on the other hand, is more complex than that of silver ; in addition to a 

 doublet system similar to the doublets of the alkalis, it shows many lines 

 of a quartet system. The underlying shell cannot therefore be always 

 completed, and it must be assumed that, in addition to the valence 

 electron, one or more of the underlying electrons is easily displaced, giving 

 an uncompleted group resembling that of nickel, which has a *F^ term for 

 ground term. The two valencies are probably to be accounted for in this 

 way. A similar argument applies to gold, which is monovalent and 

 trivalent, since the preceding element, platinum, has a ^D term for ground 

 term, and the spectrum also is of greater complexity than that of silver. 



An outstanding difficulty in connection with the scheme of atomic 

 structures embodied in Table IX is that the azimuthal quantum numbers 

 of Sommerfeld's theory of the regular X-ray doublets correspond with 

 inner quantum numbers, and not with the azimuthal quantum numbers of 

 the levels indicated. The difficulty is further emphasised by Millikan and 

 Bowen's^* important discovery that the regular doublet law is valid also 

 in the optical doublet and triplet spectra, which they have especially 

 investigated in the extreme ultra-violet, and by the further work of 

 Land^^' on the same subject. 



An important step towards the removal of this and other theoretical 

 difficulties, however, appears to have been taken by Uhlenbleck and 

 Goudsmit^^ in a consideration of the possible effects of a rotation, or spin, 

 of each electron. In further developments of the theory by Heisenberg 

 and Jordan,^' it has, in fact, been found that the ' spinning electron,' 

 combined with the ' new quantum mechanics ' previously initiated by 

 Heisenberg, is competent to explain why the relativity doublet occurs 

 between two levels that differ in their inner quantum numbers, and not, 

 as in the original theory, in their azimuthal quantum numbers. At the 

 same time, it may be noted, the conception of a spinning electron suggests 

 a modified view of the fine structure of the hydrogen lines, which has been 

 further developed by Sommerfeld and Unsold*" and appears also to be 

 capable of giving an explanation of anomalous Zeeman effects. 



The extraordinary theoretical developments in recent years, leading to 

 the prediction of certain features of the spectra of elements and the 

 structure of atoms, have possibly overshadowed the progress in experimental 

 spectroscopy. Nevertheless, much experimental work of immediate 

 importance to theory has been carried on, and much more is urgently 

 called for. It must not be forgotten that, notwithstanding their general 

 probability, the adopted electron configurations and the spectroscopic 

 terms which are deduced therefrom are by no means all finally established. 

 The theory is at present largely empirical, and important modifications 

 may be demanded when the structures of other spectra have been deter- 

 mined. Indeed, there are relatively few spectra for which the analysis 



3« Phys. Rev., vol. 24, p. 209 (1924). 



»' Zeit. f. Phys., vol. 25, p. 46 (1924). 



38 Nature, February 20, 1926 ; see also Thomas, Nature, April 10, 1926. 



»» Zeit.f. Phys., vol. 37, p. 263 (1926). 



*" Zeit.f. Phys., vol. 36, p. 259 (1926). 



