135 
Having considered how the magnetic and dia-magnetic 
condition of matter may be accounted for, the revulsive 
motions require considerable attention. These cannot, in 
any manner which I can conceive, be explained on any 
hypothesis, either of molecular action, independently of 
electric currents, or of electric currents directly induced 
by the magnetic force. If the revulsive motions were sup- 
posed to arise on the ceasing of the molecular disturbance, 
^ under such conditions that on the ceasing of the con- 
strained molecular condition, the atoms, by a species of 
libration, spring into the contrary arrangement, passing 
beyond the normal condition, and thus that magnetic 
action is succeeded by temporary dia-magnetic repulsion, 
and dia-magnetic action is succeeded by temporary magnetic 
attraction, we have, in addition to many other objections, 
the important one that the force exhibited by the revulsion 
is not in proportion to the force supposed to originate it; 
for, as appears from the before-described experiments, when 
the magnetic force is considerable, as in palladium, the 
revulsive force is very slight ; and when the magnetic force 
is slight, as in zinc, the revulsive force is comparatively 
very considerable. I have also shown by the experiments 
with spirals of copper wire, that the phenomena of revul- 
sion and of the transverse motions are closely connected 
with the conduction of electric currents. We do not per- 
ceive these motions when a complete circuit is not afforded 
for the electric current, nor have I been able to observe 
them in bodies of inferior conducting properties to the 
metals. 
The conducting power of the metal may, to some extent, 
affect the amount of revulsive force exhibited ; but the 
property of revulsion clearly does not depend on this, for 
platina is a much better conductor of electricity than lead, 
and in platina we have little, if any, revulsion, whilst in 
