repetition of M. Arago's experiments on magnetism, &c, 489 
direction with a velocity continually accelerating, till they 
move on together with equal velocities. 
31. It is manifest that, caeteris paribus, the greater the 
relative velocity, the more will the pole developed in the 
plate lag behind the magnet, or the magnet (in the reverse 
case ) behind the pole. The more oblique therefore will be 
the action, and the greater the resolved part of the force, 
and the velocity produced by it dato tempore. The same 
effect must also be produced by an increase in the absolute 
force, or lifting power of the magnet ; so that in such expe- 
riments there is an advantage in using large magnets which 
have great lifting powers, over small ones with intense 
directive forces, and this is perfectly consonant to experience* 
32. Hitherto we have only considered the case of rectili- 
near motion. If we regard the magnetism of the plate as 
very transient, and the velocity moderate, the whole space 
occupied by the magnetised portion of the plate will still be 
small, and confined to the immediate neighbourhood of the 
point vertically under the magnet. If the motion of the 
latter change its direction, the momentary pull communi- 
cated to the plate will always be in the direction of a tangent 
to the curve described. If therefore it describe a circle, it 
will tend at every instant to impress a gyratory motion on 
the plate about a centre vertically under the centre of its own 
motion, and vice versd, if the plate be made to revolve about 
a centre, it will tend to drag the magnet round with a conti- 
nually accelerated motion, provided its rectilinear recess from 
the centre of motion (or its centrifugal force) be prevented by 
a proper mechanism. The former is the case of a disc of 
copper suspended by its centre, and set in rotation by a 
