( 250 ) 



the two electrons. If wc impart to this svslein a velociiy i\ then 

 the monieiituui will l)e (2/// -j- "'12) ^^ ^)- ^^ *^i'g of the electrons is set 

 in nioiion and the other remains in rest, the moment nm will be 

 (,/y, _|_ ^ v;?^^U\ for the mass n/,.,, remaining halfwav between the two 

 electrf)ns, moves with a velocity h v. This however is only trne for 

 cimisi stationary motio]i, and we innst keep in view that the requirements 

 for (piasi stationary motion are in this case by no means so easy to 

 be fnltilled as in the case of a single electron. If e. g. the electron 

 vibrates witli a wavelength < y, then the mass residing in the tield 

 and contributing to di..^., cannot be assumed to have everywhere the 

 velocity h iv This mass therefore may not be thought to be concen- 

 trated in the centre of inertia and the mass of the electron may not 

 be augmented with h in^.^. 



Let ns consider electrons on the sun. They have a greater poten- 

 tial energy than those on earth. Are we justified in ascribing a greater 

 mass to them and in expecting that the period with which they 

 vibrate will accordingly be greater ?') In order to answer this question 

 we must investigate whether this potential energy shares the motion 

 of the electrons or not. If we assume that gravity propagates with 

 intinite velocity, we shall have to assume that the gravitational energy 

 moves with the electron, and then the mass of electrons on the sun 

 would really be greater than that on earth. If on the other hand 

 gravitation propagates with the velocity of light this conclusion woidd 

 not be justitied. 



If the shifting of the spectral lines in the light of the sun as 

 expected by Einstein therefore does not occur, this fact does not 

 prove that we are wrong in ascribing a mass to the energy. But it 

 proves that gravitation propagates with finite velocity. Ifon the other 

 hand the effect did occur, it would show that gi-avitation propagates 

 with infinite velocity or at least with a velocity which is very great 

 compared with that of light. The effect would therefore be in direct 

 contradiction to the hypothesis of relativity. 



§ 5. We will still consider the following special case. A rod of 

 1 cm^ cross section experiences a pressure tjj- in the direction of its 

 length. We will call the ends of the rod ^ and j5 and choose the direc- 

 tion from .4 to B as positive A'-axis. The rod moves with a velocity 

 i^ in this direction. If" be the density of the energy of the rod. The 

 amount of energy which [)asses through a stationary plane of unit 



1) This agrees with the calculations of L. Silberstein, Phys. Zeitschr. XII, 

 p. 87, 1911. 



3) A. Einstein, Jahrbuch der Radioakt. u. Elektr. IV, p. 459. 



