448 BELL SYSTEM TECHNICAL JOLRSAL 



(3) In the same formula, the magnetic quantum-number m depends 

 on the numeral J for the level in question; it assumes 27 ^•alues 

 altogetlicr, commencing at the maximum \alue {J—\) and going 

 downwards across zero to { — J+\). 



These rules form a beautiful little problem for the designer of 

 atom-models. They have often been tested and \erified (it is not easy 

 to find out just how far), and at present are widely used in the decipher- 

 ing of spectra. It appears, however, that some spectra — particular!}' 

 those of the inert gases — are too complicated even for these rules, and 

 possess a structure even more elaborate. Considering how difficult 

 it is to grasp the structures already described, one ma>' be excused 

 for feeling some dismay at the prospect. 



E 11. Effect of Eleclric Field on the Stationary Slates 



W'JU'ii an ck-rtric ticld is applied to a radialiiig gas, the lines of its 

 spectrum arc rcplarcil hy j^roups of lines, dticn rich and ciimijliratcd. 



l-ijC. 11 -Kcsohition of spcclnim-liiies into groups, displacement of lines, and rinit- 

 Reiice of missing lines, prodiued liy a strong electric field (increasing from the lop 

 downwards to nearly the bottom of the picture). (J. S. Foster, Pliysical Rn'inc} 



(Fig. 11.) F'rom these we infer, as heretofore, that the stationary states 

 are replaced by grou[)s of stationary states. The atom-model jiro- 

 posed for lu'drogen and ionized heliinn has been i\t r.inrdinariJN 

 successful in di'scribing the elTect of electric field iipnn llieir spectra. 



