of Relativity in Electromagnetism. 419 



It is worthy of note that whereas in general the forces 

 acting on the electron in a uniform magnetic or electric field 

 are different from those of the Maxwellian theory, yet the 

 masses acted upon are identical. Becquerel rays moving 

 obliquely to the field should therefore exhibit a different law 

 of deviation from that expected on the basis of the Maxwellian 

 theory- Here there is the possibility for carrying out an 

 experimentum cruris to decide for or against the principle of 

 relativity as here introduced. 



§ 5. It is of especial interest to investigate the bearing of 

 our principle on the phenomena of radiation. It is evident 

 that relativity excludes any influence of the translatory motion 

 of the earth on terrestrial optics ; and it is likewise apparent 

 that for relative motions of a source of light and an observer 

 the principle of Doppler holds and that the phenomena of 

 aberration are easily accounted for. 



The phenomena hitherto considered were governed by the 

 Maxwellian equations for vacuous space ; and the question 

 imposes itself whether the adopted principle of the relativity 

 of motion is capable of accounting for phenomena in which 

 the structure of ponderable matter plays an essential role. 



The electron theory, as is well known, interprets the pro- 

 perties of matter by assuming that it is made up of complex 

 assemblages of negative and positive charges. Thus, while 

 our equations still hold in the interstices of matter for the 

 single electrons, the observable properties are due to the com- 

 bined effects of numerous assemblages of electrons. We 

 obtain these effects by a process of averaging, and as Lorentz 

 has shown the result is a set of equations for ponderable 

 matter which is identical in form with that of Maxwell, — 

 leaving out of regard phenomena of dispersion. 



Now it is evident that with our principle of relativity we 

 arrive at the same results as Lorentz as long as we restrict our 

 investigation to bodies at rest. But there are two well-known 

 experiments which have always been adduced as proving the 

 existence of an aether at rest. These experiments are that of 

 Ronto-en on the magnetic effect of the rotation of a dielectric 

 between charged condenser-plates, and that of Fizeau with 

 the propagation of radiation in streaming water. It will 

 suffice to show that the former experiment is in perfect accord 

 with our hypothesis. For let a liquid dielectric flow between 

 charged condenser- plates. At the boundary of the dielectric 

 free charges will form and will take part in the motion of 

 the liquid. This motion constitutes a convection current, and 

 as such will be proportional to the free charge ; and the 

 magnetic field outside the condenser will therefore likewise 

 be proportionate to the free charge. The latter for a given 



2 G2 



