CHAPTER XX. 



DIFFICULTIES CONNECTED WITH THE ELECTRIC 

 THEORY OF MATTER. 



1. Concerning the formation of Spectrum lines. 



LORD RAYLEIGH has pointed out, in the Phil. Mag. 

 for January, 1906, that if the radiation from atoms is 

 due to shock and recovery, the series of vibrations 

 that would be obtained would show a simple relation 

 between the squares of the frequencies, as is th case 

 with plates and bells and other disturbed elastic 

 vibrators, and not between the simple frequencies 

 themselves ; whereas the spectrum observations of 

 Rydberg, and of Kayser and Runge, show that 

 simple expressions for the first power of the 

 frequency, and constant differences of frequency 

 among a series of lines, are really applicable to 

 the facts. 



Waves of this observed kind would be emitted by 

 electrons which radiate by reason of their unper- 

 1 turbed orbital motion, or other regular concomitant 

 of the constitution of the atom, but would not be 

 the result of oscillatory recovery from disturbance 

 of equilibrium. 



The natural frequencies of an undisturbed rotating 

 ring of corpuscles, for instance, would be functions of 



