Faraday-Maxwell Mechanical Stress. 619 



Within this region A ... B let there be a ring (II) in all 

 respects similar to I, every electron in the one system having 

 its counterpart in a like electron of the other. We shall 

 suppose that, when t = 0, not only is the assemblage o£ 

 •electrons II similar to the assemblage I and similarly oriented, 

 but each electron of II has the same .v-component of velocity 

 as the corresponding electron of I and the same c-component, 

 while the ^/-component of the II-e!ectron's velocity exceeds 

 that of the I-electron by bj^/irp. Thus the compound 

 system consisting of II with its surrounding aether is only 

 distinguished from the system I with its surrounding aether 

 by a general translational velocity bJ^Trp, which cannot 

 modify the internal dynamics of the compound system II. 

 This system accordingly, from the time £ = onwards, imitates 

 the compound system I in every detail of motion of every 

 electron ; the correlation being maintained so long as no 

 electromagnetic disturbance has reached the region occupied 

 by I, while the state of the region occupied by II continues 

 to be defined by (9). In particular, as no electric current is 

 flowing around the rotating ring I, neither does any current 

 ■flow around the rotating ring II : and since the direction of 

 the axes of rotation of the rings may be any we please, it 

 must be concluded that the region between the planes A and 

 B is free from magnetic induction ; which is contrary to the 

 supposition with which we started. 



15. A further word may help to make the matter clearer. 

 Let a sphere S be described completely enclosing the system 

 II at time £ = (tig. 1), and let P be a concentric sphere 

 whose radius exceeds that of S by Vt, where Vis the velocity 

 of radiation. Then, in accordance with well-known principles 

 of wave-propagation, the state of things within S between 

 £ = and t = r is completely determined by the state of things 

 within P when t = 0. If. then, the sphere P lies wholly 

 between the planes A and B, the dynamical circumstances of 

 the system II, from £ = () to t = t at the least, will be precisely 

 the same as if the range of equation (9) — instead of being 

 limited by the planes A and B at time £=0 — extended to infinity 

 in all directions. Thus no argument against the conclusions 

 of §§ 11-14 above can be founded on the circumstance that 

 the hypothetical aether-drift corresponding to (9) is limited 

 by the planes A, B ; or that the whole extent of quiescent 

 aether beyond the limits C, D provides a standard relatively 

 to which the velocity of the drift can be measured. 



16. The foregoing is offered as a rigorous proof of 

 the proposition that the magnetic vector in free aether is 

 not to be identified with translational aetherial velocity — a 



