SOME CONTEMPOR.IRY ADl'ANCES IN rilVSfCS—X 107 



ones of which are arranged in a peculiarly stable and symmetrical 

 fashion (as for instance, in a group of eight at the corners of a cube, 

 though this is by no means sure) ; while the atom of the next following 

 alkali metal consists of just this sort of arrangement or "inert-gas 

 shell," now to be known as the "residue" or "kernel," and of one 

 additional electron now to be known as the "valence-electron," usually 

 much farther away from the nucleus. 



If to such an atom-model we apply the doctrine of Stationary 

 States, we may infer that for each and every arrangement of the 

 electrons in tlie residue, or (to use more general terms) for each condi- 

 tion of the residue, there is a whole system of Stationary States dither- 

 ing from one another only in that the valence-electron treivels in 

 different ones among a system of quantized orbits. These orbits we 

 may suppose to conform to the quantum-conditions (2) and (3), at 

 least until convincing e\'idence is brought to the contrary. Such in 

 fact is the interpretation of the system of Stationary States, transi- 

 tions between pairs of which are responsible for the "optical spectrum" 

 of each alkali metal. 



An electron at a very great distance from the kernel of such an 

 atom will experience an attraction towards it, practically indis- 

 tinguishable from the attraction which would be exerted by a single 

 (hydrogen) nucleus of charge -\-e. One might say that the (Z— 1) 

 electrons surrounding the nucleus of charge -\-Ze effectively cancel 

 a portion -\-{Z—l)e of the nuclear charge; or to use a more common 

 word, that they "screen" it. As the imagined distant electron moves 

 inward towards the kernel, the screening will cease to be perfect. 

 An effect should occur analogous to the "stray field" which penetrates 

 the meshes of a grid; since the electrons of the kernel do not form a 

 continuous shell of electricity enclosing the nucleus, the latter should 

 make itself felt through the interstices, although this effect may be 

 diminished by the swift motion of the electrons. All this is specula- 

 tion of the wildest kind. The only deduction reasonably safe is this, 

 that very far outside the kernel the field will be very nearly the inverse- 

 square field due to a hydrogen nucleus of charge -\-e; very near to the 

 kernel the field will be quite incalculable '; while in between the very- 

 far-out and the very-near-in region, there will be an intermediate 

 region, in which there may be some chance of finding an adequately 

 approximate expression for the field. On the existence of such a 

 region, in which such an approximation is good enough to be valuable, 



' Unless it is violently sini])lifiecl by some agency or restriction of which at present 

 we know nothing. 



