122 BELL SYSTEM TECHNICAL JOURNAL 



l^ach atom possesses one or more such systems of Stationary States; 

 and the i)articular tyix's which an element disphiys dei)end in a 

 detinitt- and fairh' clear manner upon the position of the element in 

 the Periodic Table, being in fact one of the most distinctive of the 

 periodically-varying qualities, luxch atom with an even number of 

 electrons exhibits systems which are all of odd multiplicity, and each 

 atom with an odd number of electrons exhibits systems which are all 

 of even multiplicity; thus magnesium, with twelve electrons, has a 

 singlet system and a triplet system, while sodium and once-ionized 

 magnesium, each with eleven electrons, have each a doublet system, 

 and neon with ten h^is a singlet, a quintet and two triplet systems/ 

 Rememl)ering what was said about the consecutive binding of elec- 

 trons, it will be noticed that these facts show a regular difference be- 

 tween the binding of the Mh electron when N is odd and the binding 

 of the iVth electron when N is even. Otherwise expressed, they show 

 that a kernel of A^ electrons treats an oncoming member in one or 

 another of two distinctive ways, according as A^ is even or odd. The 

 influence of magnetic fields on spectra likewise shows that this com- 

 plexity of the Stationary States is a quality not negligible, but primary. 



The features of the atom-model hitherto described must be supple- 

 mented with some new one if it is to cope with such facts as these. 

 We have represented (for example) the sodium atom in its 2p state 

 by a "valence-electron" cruising with angular momentum 2{h/2ir) 

 in an orbit around a "kernel" composed of ten electrons and a nucleus. 

 But there are two such states instead of one; if the angular momentum 

 of the valence-electron is to be equal to 2{h/2ir) for each of these, 

 some other not yet mentioned feature of the atom must discriminate 

 the two. One might, of course, again proceed as we did in discussing 

 the "primed terms," by assuming that the kernel of the atom is in 

 one condition when the atom is in the 2pi state, and in another slightly 

 different condition when the atom is in the 2p-2 state. This would 

 probably entail as many different conditions of the kernel as there 

 are pairs of terms in the sodium spectrum — a great number, and yet 

 small in c()mi)aris()n with the multitude which would be required by 

 other atoms; yet such may be the eventual theory. However, it is 

 possible to construct for these facts an atom-model out of two re- 

 volving parts, whereby different Stationary States of a group are 

 represented not l)y varying the condition of either part separately, 



' Hydrogen and ionized helium are not included under this rule. HeHum shows 

 a singlet and a doublet system together, a combination which violates the rule as 

 stated, unless the doublet system is really a triplet system in which two states of 

 each triad are too close together to be distinguished. 



