soMi: C(K\ 1 1 .\ii\>H.u<) .inr.ixcis i.\ rinsics-iii s'h 



ever, l)een (lisi'criUHl ii) the "fine structure" of the K absorptioii-eclRes 

 of the huliter elements (notably the elements from sodiiun to potas- 

 siimi). In the second place, the process of emission is difTerent in 

 the two cases described by equations (I) and (-i). In the former 

 case, the rays were emitted as the valence-electron (or another re- 

 placing it, which comes to the same thing) redescended its ladder of 

 orbits: but when a deep-lying electron is extracted, the resulting 

 rays are emitted because of rearrangements of the other internal 

 electrons of the atomic electron-system, which occur irrespective of 

 whether the departed electron quickly returns to the atom, or remains 

 a long time away. 



I will now risk the making of a distinction \vlii( li may e\entuall\- 

 turn out not to be the most natural or practical, \>y reser\iiig the 

 name deep-lying electrons for those electrons which lie entireh' witiiin 

 at least one completed electron-shell of an atom, and designating 

 the others (exclusive of the valence-electron, which has already been 

 set apart from the rest) as the shallmv-lying electrons. It follows from 

 this definition that the first nine atoms of the periodic table, up to 

 fluorine (inclusive) possess only shallow-lying electrons; the next 

 eight (.Ve to CI) have one set of deep-lying electrons, the K set; the 

 next eighteen (.4 to Br) have at least four sets of deep-lying electrons, 

 the A' set and three L-sets (the last three can be grouped as one). It 

 follows also that ever>- instance in which an excitation-potential has 

 been measured, and the wave-lengths of the excited rays have also 

 separately been measured, is an instance in which a deep-lying electron 

 is involved. For example, the excitation-potentials invoking extrac- 

 tion of the AT-electrons have been measured from the top of the 

 periodic table down to the twelfth element {Mg), over which range 

 they decline from 11.5,000 volts to 1100 volts; the excited waves 

 have been measured over the same range and down to the eleventh 

 element (A'a), over which range they rise (for the principal ray) from 

 .\0.\ to 11.88A. At this point, and just before the A'-clectrons pass over 

 into the categorv- of shallow-King electrons at the ninth element, the 

 wave-lengths enter into the inaccessible range. The wave-lengths of 

 the rays excited when one of the L electrons is displaced have been 

 measured from the top of the table down to the twenty-ninth element 

 iCu) where, arriving at 13. .'iA, they tf)o pass into the immeasurable 

 class. 



The general conse(|uence of all this is, that the excitation-potentials 

 involving shallow-lying electrons must be below 1000 \olts; that, 

 conversely, the excitation-potentials observed between 2h \olts and 

 1000 volts are chiefly those of e.xcitations which consist in displace- 



