126 Mr. E. C. Kemble on the Probable 



addition a rediscussion of the existing experimental material 

 bearing on the nature of the helium atom, with special 

 reference to the recent study of the ultra-violet arc spectrum 

 -of helium by Fricke and Lyman*. 



Objections to the Model of Franck and Reiche. 



The first objection to the model of the normal helium atom 

 proposed by Franck and Reiche is that, like that of Lande, 

 it is difficult to reconcile with the chemical behaviour of 

 helium. Both of these models require that one of the 

 -electrons revolve about • an outer orbit, while the other 

 revolves about an inner one. The recent discussion of 

 atomic structure from the chemical point of view by 

 Langmuir | and others can leave little doubt, however, that 

 all the electrons in the normal atoms of the inert gases are 

 arranged in peculiarly stable and symmetrical groups of 

 orbits. The atoms o£ the alkali metals differ from those of 

 the inert gases which precede them in the periodic table in 

 having one loosely-attached valence electron which does not 

 filter into the stable group system. It is not therefore sur- 

 prising to discover that the orbit of the valence electron in 

 the normal sodium atom is of the same general character 

 as the outer orbits through which the electron passes when 

 emitting radiation. On the other hand, the assumption that 

 one of the electrons in the helium atom normally rotates 

 about an outer orbit enclosing that of its companion puts the 

 former in the valence electron class, and leaves the chemical 

 behaviour of helium a mystery. 



The second objection to the model of Franck and Reiche 

 is that it gives the wrong energy and ionization potential. 

 Or, stating the matter somewhat differently, the energy of 

 the actunl helium atom as measured by its ionization potential, 

 and more recently by the spectroscopic observations of Fricke 

 and Lyman, does not fit with the energy* values of the other 

 stationary states in the series to which it is supposed by 

 Franck and Reiche to belong. 



In testing this point it is necessary that we compare the 

 energy of the normal state with the energy obtained by 

 extrapolation from the sharp series crossed orbit terms of 

 higher quantum number. We make use of the scheme 

 adopted by Lande and Sommerfeld, who have employed the 

 approximate formula 



(m, S) = N/(m+S)» (1) 



* Phil. Mag. May 1921, p. 814. 



t I. Langmuir, Am. Chem. Soc. Journ. xli. pp. 868-934, June 1919. 



