190 ELECTRIC THEORY OF MATTER [CH. xx. 



to bring it out. But then anything in the nature of 

 collision would give a series of vibrations characterised 

 by the square of the frequency. Hence there is a 

 difficulty. 



The difficulty seems to be capable of being over- 

 come by the suggestion already made on page 182 

 that during the chemical collision in question, 

 the perturbation is of the nature of a disturbance 

 of the centre of gravity of a revolving system. 

 Such an eccentricity, or any other destruction of 

 symmetry, would at once develop strong radiating 

 power; but it might nevertheless leave the harmonic 

 constituents, and peculiarities of the radiation, to be 

 governed by the simple frequency law appropriate 

 to the revolving constituents, rather than to the 

 squared frequency law appropriate to their elastic 

 recovery from vibration. In other words, the fact 

 of considerable radiation would be due to the 

 collision, but the land of radiation emitted would be 

 due to the previously existing constitution, upon 

 which the disturbance had now conferred temporary 

 radiating power of considerable magnitude, tem- 

 porary, because the radiation itself must speedily 

 exhaust the energy of the disturbance and soon 

 restore the pristine condition. 



It may seem an open question whether a disturbance 

 of centre of gravity would allow quiet revolution to 

 continue as before, or whether it would not precipitate 

 a catastrophe. Simple mechanical considerations show, 

 however (Phil. Mag., March, 1904, p. 264), that 

 when the law of force is the direct distance, as it is 

 inside the hypothetical sphere of positive electricity, 

 no such calamity is to be expected ; but that, on the 

 contrary, a triangle or other group of corpuscles, no 

 matter how much displaced provided that they are 



