ATTRACTION BY INDUCTION. 399 



The diagonals being then conjugate, equations (23) of 949 show, 

 in fact, that the current which at any given instant traverses the 

 movable coil, in which there may be an electromotive force e of 

 induction, has the value 



t = 

 or 



2 = 



according as the liquid occupies the branch & or the branch &'. 

 Now, these two values are identical if we allow for the direction 

 in which the electromotive forces E 6 and E 6 / have been measured. 



The deflection itself depends not only on the amplitudes, but 

 on the phase of the two currents I and i (963). It would be 

 null if the electromotive force of polarization were itself null, for 

 the intensity of the current would be null in the movable coil. 



996. ATTRACTION BY INDUCTION. Messrs. Guthrie and Boys* 

 have endeavoured to deduce the conductivity of liquids from the 

 intensity of the currents induced in their mass by the motion of a 

 magnet. These currents, being closed and without electrodes, 

 cannot produce any effect of polarization. 



The liquid is contained in a glass vessel suspended by a metal 

 wire about which it can rotate. A system of external magnets, 

 producing a sensibly uniform horizontal field, can be rapidly ro- 

 tated about the common axis of the wire and vessel. Screens 

 interposed prevent any motion being transmitted through the air. 



Other things being equal, the currents developed in the liquid 

 mass are proportional to the intensity of the field, to the velocity 

 of the relative displacement, and the conductivity of the liquid. 

 They tend to oppose the motion. If there were no friction of 

 the liquid against itself, or against the sides, the vessel would 

 remain at rest, and the liquid would finish by acquiring the same 

 velocity of rotation as the magnets. On the other hand, if the 

 mass of liquid formed a rigid system with the vessel, this would 

 be dragged in the direction until the torsion couple counterbalanced 

 the moment of the electromagnetic actions. In fact, the various 

 concentric layers acquire a very slow rotation, the angular velocity 



* GUTHRIE and BOYS. Phil. Mag. [5], Vol. x., p. 398. 1880. 



