1919] 



on Ether and Matter 



473 



the J come from very large masses of highly rarefied gas. For under 

 such conditions the atoms would have more room and could possess 

 far outlying or ultra-Neptunian electrons, and yet have total substance 

 enough to display their spectra. 



To contemplate the emission of radiation, both waves and particles, 

 we may picture one of the satellite electrons in a many-orbited atom 

 struck or so thoroughly perturbed by the sudden arrival of a foreign 

 charge as to precipitate it into the next inner ring, ejecting the 

 constituent of that ring into the one below, and so on, after the 

 manner of the " jack for mustard " game with a series of wooden 

 bricks set up on end. 



Wave emission should accompany each transition. The effect of 

 precipitating the innermost electron on the body of the nucleus is 

 not clear ; but a compound nucleus must be a strangely interlocked 

 conglomerate, and an explosion seems not unlikely : especially if one 

 of the supposed binding negative electrons were ejected. The 

 potential gradient close to a nucleus is prodigious. 



The effect of the arrival or departure of a charged particle at the 

 nucleus would be suddenly to change its intrinsic attracting force ; 

 and this of itself would render all the orbits elliptical for a time, 



with excentricity — ,'^ thus exciting radiation of several fre- 

 quencies. If the radiation ceased when the excentricity was got rid 

 of, a new circular orbit would be taken up, and thus perhaps discon- 

 tinuities might be accounted for in a dynamical manner. 



The effect of properly attuned X-rays or ultra-violet light, if it is to 

 be accomplished through resonance — and it is difficult to account for its 



* This can be proved as follows : — 

 For a circular orbit 



and r^ v^ = h- = fx r. 



When fj. suddenly changes to k fx I where k may be -^^— ) the velocity 



does not instantly change, but the orbit acquires an a and an e, such thajj 



a(l- e^)= {^ or e^ = i -^ 

 k /x k a 



also 



This last gives 



\r a/ 



= 2 





And the place where the sudden impulse occurred becomes an apse of the new 

 orbit, because r = a (1 ± e). {See also Appendix I.) 



2 K 2 



