414 Interference Experiments in a Highly Rarefied Gas. 



So we can go one step further. 



We assume an atom model with electrons rotating round 

 a central body, lines of force originating from the electrons 

 and disturbances travelling along them. A line of force 

 originating normally to the plane of rotation of the electron 

 will not have much motion, and a train of oscillations (even 

 of considerable length) travelling along this line will be in 

 something like a straight line. But on a line of force in the 

 plane of the rotation of the electron, a train of oscillations 

 will not be emitted as a whole in a certain direction ; if the 

 motion of the electron is only rapid enough the train of 

 oscillations (representing a succession of up to many hundred 

 thousand periods) will be emitted in a spiral with a very 

 considerable number of turnings. These spirals spread. 

 As a finite source of light will emit them in random directions, 

 these spirals will produce effects similar to those of spherical 

 waves when falling upon an interference apparatus. 



From the fact that very distinct interference-fringes can 

 be got experimentally, we would have to deduce that the 

 energy emitted from a source of light in straight trains is 

 very small compared with the energy emitted in such 

 spirals. 



These views form perhaps the foundation for a combined 

 explanation of the purely optical phenomena and or! the 

 ionization effects. 



Such a conception of the energy emission of the elementary 

 oscillator would avoid a dualism to which Marx * alludes 

 in the explanation of his photoelectric experiments. He 

 thinks of waves with points of concentration, and finds that 

 the amount of energy localised in these points need not be 

 more than a small fraction of the energy contained in the 

 wave. 



Summary, 



(i.) Interference experiments are described where the 

 source o£ light and the interference apparatus are in 

 the same highly evacuated vessel. 



(ii.) The source of light employed is the resonance radiation 

 of Hg- vapour excited to continuous luminosity by a 

 special Hg-vapour lamp. 



(iii.) The experiments show that a hi resonance " hypothesis 

 developed for an explanation of the purely optical 

 phenomena by units does not represent the facts. 



* E. Marx, Ann, d. Phys. xli. p. 167 (1913). 



