SOME CONTEMPORARY ADVANCES IN PHYSICS— X 113 



generally far beyond; that the calcium atom consists of the same sort 

 of kernel and two loosely-bound electrons, the scandium atom of the 

 same sort of kernel and three outer electrons. 



The Stationary States of the potassium atom have been interpreted 

 as corresponding to various quantized orbits which a single outer 

 electron describes around an unchanging residue; the lines of its 

 spectrum have been attributed to leaps of this electron from one orbit 

 to another, the residue remaining unaltered. There is nothing in- 

 compatible between this and the previous conception of the potassium 

 atom. 



The Stationary States of the calcium atom resemble, in their 

 arrangement, those of the potassium atom sufficiently to make the 

 same general sort of an explanation desirable, — to make it desirable 

 to suppose that one electron is loosely-bound and remote from the 

 nucleus, the other nineteen tightly-bound and near the nucleus; one 

 loosely-held electron versus nineteen tightly-held ones. But the 

 evidence from chemistry and electrolysis demands two loosely-held 

 electrons versus eighteen tightly-held ones. 



One might try to evade the dilemma by supposing that the calcium 

 atom is a sort of three-stage construction, with eighteen electrons 

 congregated in a kernel around the nucleus, a nineteenth far out by 

 comparison with the nucleus, a twentieth far out by comparison with 

 the nineteenth. For interpreting spectra, the residue of the atom 

 would be the kernel or "inert-gas shell" and the nineteenth electron, 

 the valence-electron would be the twentieth. For interpreting chemi- 

 cal data, the residue of the atom would be the inert-gas shell. This 

 conception would rescue the interpretation of the calcium spectrum 

 made after the fashion of the one just expounded for alkali-metal atoms. 

 It would probably demand a larger atom, or a more shrunken kernel, 

 than other data will allow. 



Or one might suppose that the nineteenth and the twentieth elec- 

 tron are on the whole about equally remote from the nucleus, and yet 

 it is possible for one of them to change over between any two of a 

 vast system of quantized orbits without greatly affecting the other. 

 There is certain evidence for this conception which I shall presently 

 narrate. 



Or one might suppose that the nineteenth and the twentieth electron 

 are a system by themselves, and that each Stationary State corresponds 

 to a particular configuration of this system, so that each line of the 

 spectrum is attributed to a leap not of either electron separately but 

 of both together. This idea seems to be gaining ground rapidly in 

 dealing with atoms composed of a kernel and several outer electrons, 



