of Ether Waves in Air. 78& 



idea, but assuming there may be no more than four stray 

 electrons on the average to the cubic centimetre, there would 

 still be a calculable opacity due to this cause ; for a distance 

 of 10 21 kilometres, or a hundred million light-years, would 

 reduce the energy of direct radiation to 1/^th of its value. 



Professor Barkla has informed me that for X-ray pulses 

 it is not necessary for electrons to be isolated in order to 

 exert a scattering influence, but that electrons associated 

 with atoms of matter appear to be sufficiently free to 

 take part in the process. It may be that thus much ultra- 

 violet light is stopped, for to small enough waves the opacity 

 so caused would be considerable, because of the immense 

 value of N. That which is supposed by the above theory to 

 occur to waves of all kinds, in really ionized space, has thus 

 been already observed in the case of those minute waves to 

 which atom-bound electrons are partially and sufficiently free*. 



Returning to the consideration of the other phenomenon,, 

 the bodily propulsion or sweeping forward of electrons by 

 the wave : at first sight it is not easy to realize physically 

 why electrified particles should be thus propelled ; but no 

 flaw appears in the reasoning, and it may be classified as a 

 phenomenon akin to the magnetic deflexion of cathode rays.. 

 For (a) The electric force of a wave will obviously tend to 

 move a charged particle in its own direction in the wave- 

 front ; (h) directly it moves, it is a current, and is therefore 

 acted on by the magnetic component of the wave ; and (c) 

 this action will propel it in a direction at right angles to 

 both, i. e. normal to the wave-front. 



Taking the plane of the wave as xy, and the electric 

 intensity as X, the force acting on a charge e is X<3 in the 

 first instance; though in so far as the particle is propelled 

 along the ray it becomes subject to magnetic force also^ 

 and the whole x force is 



mx = X^ — jbuilez. 



Also my = 0, 



and m'z = fiB.e x. 



These are the equations ; and when they are worked out, as* 

 they are on pages 257-8 of the Philosophical Magazine for 

 August 1902, the result is a progression superposed upon an. 

 oscillation of twice the wave-frequency, a sort of fore- 

 shortened cycloidal motion, expressed by the equation: — 



the original wave being 



H = H cos (pt — qz) . 



