Prof. E, Edlund on Unipolar Induction. 291 



polar induced current ; while the rotation of the magnet about 

 its axis has nothing to do with the phenomenon, as several phy- 

 sicists have assumed. If the magnet and cylinder be connected 

 together so as to form a compact body, the magnet is carried 

 in the direction of the rotation of the cylinder without any 

 modification in the induction of the magnet resulting, as we 

 have seen. The magnet acts in this case on the cylinder as if 

 the former were immovable and the latter alone rotating. The 

 opinion expressed by several physicists, that the magnet can- 

 not produce unipolar induction in a conductor with which it is 

 intimately united, is therefore not correct. If the radius of 

 the cylinder be sufficiently reduced for the cylinder to be in 

 perfect contact with the magnet, this circumstance will not 

 prevent induction taking place in the cylinder ; and since in- 

 duction will be produced perfectly irrespective of the thickness 

 of the cylinder, the latter can be entirely removed and the 

 electrodes of the galvanometer put into immediate contact with 

 the magnet without causing the induction to cease. The 

 magnet itself then performs the functions of a conductor, and 

 the induction does not result from its being put in rotation as 

 a magnet, but from its rotation as a conductor. We are here 

 in presence of the phenomenon of unipolar induction first pro- 

 duced by W. Weber*. 



If now we are asked how it is that a current is possible in 

 this case, we can reply that this must necessarily take place if 

 our idea of the nature of the galvanic current is correct — 

 namely, that it consists in the translatory motion of a fluid 

 going in the positive directionf , or of two fluids following op- 

 posite directions. To understand the necessity of the produc- 

 tion of a current in the case in question, one can adopt either 

 of the above-mentioned opinions. In the memoir cited (p. 3), 

 I have recalled the known fact that a metal ring surrounding 

 a magnet and traversed by a galvanic current transports itself, 

 if movable, along the magnet. It stops in the middle, where 

 it takes up a position of stable equilibrium if its galvanic cur- 

 rent follows the same direction as the molecular currents of 

 which we imagine the magnet to be formed. If, on the con- 

 trary, the current of the ring follows the opposite direction, 

 the ring will be in unstable equilibrium at the middle of the 

 magnet ; and if removed from that position, it will continue 

 moving further from the centre until it passes beyond the 

 poles of the magnet. This motion of the ring is determined 

 by the law which regulates the action of a magnetic pole upon 



* Pogg. Ann. vol. lii. p. 353. 



t " Theorie des Phenomenes Electriques," Kongl. Svenska Akad. Hand- 

 lingar, vol. xii. 



U2 



