Waves by a moving perfect Reflector. 705 



Again, if the radiation is polarized perpendicular to the 

 plane of incidence, it is the magnetic forces that are given as 

 above ; say 



H = /> cosm(# cos + y sin 6 — ct), 



H' =b' cos m'(xcos6' +y sin 6' + ct). 



The sethereal forces are He and — H'c in the plane of inci- 

 dence, so that the total tangential component of the electric 

 (not sethereal) force in the reflector is 



Hccos0-H'ccos0'+i-(H + H'). 



This must vanish : so that 



b(c cos 6 + v) = b'(c cos 6' -V). • • • (C) 



Now 



(c cos 6 + v)' 2 = [c + v cos 6)-— (c 2 — v 2 ) sin 2 6 ; 



hence each term of (A) is equal to — "" , the sign 



v ' ^ ccostf — V| ° 



being fixed by continuity- It readily follows that V jb in (0) 

 has the same value as a'ja in (B). 



This agrees with Mr. Edser, and is the interesting result 

 of his investigation. Tt means that a moving total reflector 

 cannot operate so as to partially polarize an incident beam 

 of ordinary light. 



This result is in accordance with anticipation ; for inas- 

 much as the thrust of radiation in the parallel beam is simply 

 equal to the density of its energy, and does not even depend 

 on v/c up to the first order, the work done by the two thrusts 

 on the receding reflector must be the same, whatever be the 

 nature of the polarization of the beam, and therefore the 

 same statement must apply to the energy that remains and 

 goes off in the reflected beam. 



The question presents itself : What is the vector corre- 

 sponding to P in the investigation for direct reflexion quoted 

 by Mr. Edser on p. 509, which represents the actual shift in 

 the aether ? It appears now that it is not in the general 

 problem the time integral of the magnetic force ; for we 

 have taken it that it is the normal component of the mag- 

 netic force itself that is continuous across the surface. 



As regards the pressure of the radiation, I am unable to 

 follow Mr. Edser's investigation. From the last formula in 

 his paper it would appear that he makes it equal to the 



