[ 179 ] 



XIII. On Unipolar Induction : Another Experiment and its 

 Significance as Evidence for the Existence of the ^Ether. 

 By E. H. Kennard, Ph.D.* 



§ 1. Introduction. 



*<TTNIPOLAR induction in the general sense may be 

 KJ defined as induction due to motion alone." This 

 condition requires that the magnetic induction shall remain 

 constant at all points fixed relative to any part of the 

 material system involved, so as to exclude any possible 

 effects due to a changing magnetic field ; and the only form 

 of motion satisfying this requirement is rotation about an 

 axis of magnetic symmetry. The fundamental problem of 

 unipolar induction is therefore this, whether the induced 

 E.M.F. is determined by the absolute rotation of the system 

 or by the rotation of its parts relative to each other. 



Theory has answered the question in three principal ways. 

 The oldest view, put forward by Faraday and adopted by 

 Lorentz, refers the effects to an induced electromotive 

 intensity given by 



-[•xBl, 



C 



where B = magnetic induction and v = velocity relative to 

 the sether (assumed stationary); the effect will therefore 

 depend in part upon the rotation of the system as a whole. 

 The " moving line " theory adopts the same expression, but 

 interprets v as velocity relative to axes fixed in the material 

 magnetic system ; this view is virtually included in Neumann's 

 theory of electromagnetism, and is based upon complete rela- 

 tivity, so that the effects depend only upon relative rotation 

 between the parts. 



Experimentally, the question can be answered only by 

 observations upon open circuits. The first investigation of 

 this kind seems to be one made by the author f, in which an 

 iron bar magnetized by a stationary solenoid was set in 

 rotation inside an insulated metal cylinder connected to 

 earth, and a charge was looked for on the cylinder due to a 

 possible E.M.F. in the earthing wire. The result w:js defi- 

 nitely negative. Barnett's objection J that the negative 

 result might conceivably be due to the non-rotation of the 

 solenoid is valid, but seems decidedly weak, for it assumes a 



* Communicated by the Author. 



t E. H. Kennard, Phil. Mag. June 1912, p. 937. 



X S. J. Barnett, Phys. Zeits. Sept. 1, 1912, p. 803. 



N2 



