666 Mr. G. A. iSchott on Radiation from Moving Systems 



(2) Element JS along a transversal through E and 

 perpendicular to it. 



Let it be at Q. We have 



The radiation is 



S ^ = ^( 1 -^)( T ' 2+P ' 2 >' 2 ( u > 



The energy received per second by the whole of a fixed 

 surface, which passes through P and at time t coincides with 

 the wave-surface through P, is 



S= 4 ^Jj(l + *g)V 2 + PV 2 ^. • (12) 



It is instructive to compare this with (8). 



When U = 0, i. e. when the system (2) is at rest and 

 identical with (2')> hoth expressions are identical and give 

 the radiation from the system (2 f ) : it is 



R'=^f((T ,2 + P'V 2 ^ (13) 



"-^0 



R and S differ from E/ by the presence of the factors 



U / U\2 



1 + /— , ( 1 + Z — J respectively. 



§ 11. It is necessary for further progress to make some 

 assumptions as to the structure and orientation of the system 

 (2), or of the system (2') . For our purpose it is little use consi- 

 dering systems other than those which will account for spectra; 

 such systems satisfy the following conditions : — (1) they emit 

 waves of determinate frequency and of intensity corresponding 

 to that of a spectrum-line ; (2) they are permanent or nearly 

 so, so that their structure remains practically unchanged during 

 Jong periods of time ; and (3) they are stable, radioactive 

 systems excepted. I have shown * that a circle of equi- 

 distant electrons in uniform rotation satisfies conditions (1) 

 and (2) ; it follows from the example of J. J. Thomson's well- 

 known model of the atom that (3) can be satisfied provided a 

 suitable controlling field be chosen. But unfortunately a 

 single ring does not emit a sufficient number of waves strong 

 enough to account for all the lines of a spectrum. Therefore 

 we must suppose our system built up of a number of rings. 

 * Schott, loc. cit. 



