( 249 ) 



So il appears tu mc tlial llie circitiiislances are much .slmj)ler lliaii 

 we shoukl conclude from the ratlier slarllir'fi,- coneliision of P]instkin, 

 that forces which do not imparl aiiv change of velocilv or shape to 

 a body, yel would change its energy. 



I lay some stress upon this j)oinl because it aj)pears to me that 

 everything in the theory of relativity may be interpreted in a nuich 

 more rational and intelligible way than many people imagine. So the 

 fad that according lo the dieory of relatixily two velocities cannot 

 be added in die ordinary way by means of die parallellogram is 

 often thought lo necessitate a new doctrine of kinematics. We must, 

 however, take into account that velocities, measured by the same 

 observer, may l)c added in the usual way. ()iily for velocities eva- 

 luated from coordinate systems moving with dilferent velocities this 

 is not the case. Those velocities, however, are measuied with dilferent 

 units of length and time. iVnd velocities measured with dilferent uints 

 cannot directly lie added. This was already the case according to 

 the old doctrine of kinematics. Kor that reason we do not want a 

 new one. 



Neither is the Lorkntz conti-action a sufticient reason to Sjieak oi 

 a new doctrine of kinematics. It appears to me that the best way to 

 formulate the discovery of Lorkntz is to say, that when a body is 

 set in motion, it experiences forces which try to make it contract in the 

 wellknown manner. It is however possible that those forces are 

 cancelled by othei- forces, and then the contraction does not take place. 

 So the contraction cannot take place when a body rotates : a begin- 

 inng contraction is in this case opposed by elastic forces. 



In the same way we formulate the law of Neavton by saying, 



that two masses at a distance y attract each other with a force /' — ^ — '-. 



Whether they will obtain the corresponding accelerations depends 

 upon the possible existence of other forces which perhaps cancel the 

 Newtonian force. So it appears ro me that the law of Lorkntz con- 

 cerning the contraction no more belongs to the region of kinematics 

 than the law of Newton concerning gravitation. 



§ 4. Mutual Mass. Let us imagine two electrons with e(pial 



charges <' but of opposite sign, and both with a total mass }ii. Their 



distance lie /'. Then we may distinguish three masses : /// in the 



e" 

 centre of each molecule, and a mass iii^., = , which reallv resides 



in the Held, but which I'or many purposes may be thought to be 

 concentrated in its centre of inertia i. e. in the point halfway bet vveeji 



