66 
POPULAR SCIENCE REVIEW. 
cisely the same way as it would be if (conversely ) it rotated and 
the magnet were at rest, or the effect depends simply on relative , 
not on absolute motion. 
III. The consideration of action and reaction also proves the 
same point. If a copper disk be oscillated about its own 
centre ( i.e . be made to rotate in alternate directions about its 
centre) above and parallel to the flat surface of the pole of a 
magnet, and concentric with the magnet’s axis, radial currents 
will (as is known) be generated in alternate directions in the 
disk at each alternate oscillation of the disk, i.e. the disk will be 
charged at its centre and circumference alternately positively 
and negatively. This alternate flow of electricity between the 
centre and circumference of the disk will tend to damp the 
oscillations of the disk, or to bring it to rest. Now on the prin- 
ciple that action and reaction are equal and contrary, the action 
between the magnet and disk must be mutual , and therefore 
the magnet, if freely suspended, would tend to be thrown into 
oscillation ( about its axis) by the oscillations of the disk ; or if 
the magnet were (conversely) itself artificially put in oscillation, 
the magnet would tend to put the disk in oscillation. But the 
magnet could not do this unless, by its rotation on its axis, it 
produced an inductive effect on the disk. The principle of action 
and reaction therefore affords additional confirmation of the fact 
that the rotating magnet has an inductive effect on the disk. 
IV. Let ns (fig. 3) represent a spherical magnet suspended so 
as to rotate freely inside a metallic ring (a slight interval exist- 
ing between the ring and the magnet), the axis of rotation of 
the magnet bein g perpendicular to the magnetic axis ns. Then 
it is an accepted fact that a current (in- 
Fig. 3. dicated by the arrows) will flow round 
the ring by the rotation of the magnet, 
the direction of the current reversing 
itself at each semi-revolution. The lines 
Q of force of the rotating magnet therefore 
intersect the ring, precisely as they 
would do if (conversely) the magnet 
were at rest and the ring rotated through 
the lines of force of the magnet. If now 
the magnet be rotated about its magnetic axis ns (fig. 4), cur- 
rents will also be generated, charging up the ring statically in 
the manner indicated in the figure, the sign of the electricity 
being opposite at the polar and equatorial parts of the ring (the 
charge being permanent so long as the magnet rotates). If 
the existence of this inductive effect be not admitted, then it 
would be necessary to conclude that the lines of force emanat- 
ing from the magnet, intersect the ring when the magnet is 
rotated about one axis (fig. 3), but not so when the magnet 
