EFFECTS OF ROTATION OF MAGNETS. 
65 
above experiment admits of just interpretation consistent with 
the fact that the inductive effect depends on relative motion, no 
one will doubt that this is really the case. However, the point 
may be independently proved from other considerations. 
I. In experiments bearing on this question (“Phil.Trans.” 1 852, 
p. 30), a compound magnet was used, consisting of two rectangu- 
lar bar magnets placed side by side with a slight interval between 
them (as shown in cross section, fig. 2). The two magnets were 
rotated about a common axis x. Now if we take a purely 
theoretic view of this question, and suppose one bar to be 2 - 
removed, then it will be evident that by the rotation of r-y-i 
the other about the axis x (now external), the centre of LxJ 
gravity of the bar will be translated about that axis, in 
which case it is admitted that the lines of force partake of the 
motion of the bar and would intersect any external conductor, pro- 
ducing an inductive effect on it. Let now the other magnet be 
added, and the same holds true of it, i.e. its lines of force partake 
of its motion and intersect any external conductor ; and we there- 
fore have now the entire compound bar rotating on its axis, and 
its lines of force partaking of its motion and intersecting external 
conductors. This must be sufficiently clear, and if we discuss the 
matter further, it is with the view rather to consider some facts 
possessing a certain novelty in their aspects than to add proof to 
the above conclusion. 
II. Let us suppose one of the above magnets (fig. 2) to be 
again removed, and the remaining one to be made to revolve 
about the external axis x , a copper disk being placed concen- 
trically above this axis. Then by the excentric rotation of the 
magnet about the external axis x , the centre of gravity of the 
magnet is translated past the disk, so that the lines of force of 
the magnet intersect the disk and generate currents in it, which 
discharge themselves in the portion of the disk remote from the 
pole of the magnet. But if the second magnet be now added, 
the disk will then (by the rotation of the compound magnet) be 
continuously and symmetrically covered by the pole of the com- 
pound magnet, so that the currents have nowhere to discharge 
themselves. There is therefore no effect of a current in the disk 
observed. This is, however, no proof of absence of inductive 
effect on the part of the rotating magnet ; but from the very 
fact that the currents tend to be generated equally along all the 
radii of the disk, no back flow is possible, but the centre and 
circumference of the disk are simply charged up with electricity 
of opposite sign. If a wire be made to connect the centre and 
circumference of the disk, the accumulated electricity cannot 
discharge itself through the wire, because the wire itself is acted 
on by the rotating magnet in the same way as the disk, and no 
discharge is possible. The disk in this case is charged in pre- 
NEW SERIES, VOL. II. NO. V. F 
