70 
POPULAR SCIENCE REVIEW. 
siderations must apply, or there is no necessity to draw a dis- 
tinction in the mode of action in the two cases when the magnet 
rotates (on its axis), and when the wire rotates about the 
magnet, but the effect is simply one of relative motion. When 
a current is transmitted through the circuit ; the wire, w, reacts 
upon the system of force of the magnet, and both the magnet 
and the wire thereby tend to be propelled in opposite directions 
(or the action is mutual). Whether therefore the magnet 
rotates (on its axis) or the wire rotates about the magnet, 
merely depends on which is free to move, and therefore the 
rotation is due to the same cause in both cases, i.e. to the 
mutual reaction between the wire, w (traversed by a current), 
and the external lines of force of the magnet. On the other 
hand, the hypothesis that a rotating magnet and its system of 
force were independent of each other, necessarily would involve 
the assumption that there can be no reaction between the 
external wire and the magnet, which leaves us the only re- 
maining conclusion that the magnet is caused to rotate by the 
portion of the current within itself. Here we encounter the 
difficulty of a current within a body causing the rotation of 
that body without any external object for the current to react 
upon. This difficulty disappears at once on the view (already 
enunciated) that the rotation of the magnet on its axis is due 
to the reaction between it and the external wire, or to the 
same cause as the rotation of the wire (conversely) about the 
magnet. This view also brings the properties of the magnet 
into agreement with those of the helix (as in accordance with 
Ampere’s theory) ; for it is well known that the rotation of a 
helix on its axis is solely due to the reaction of the external 
portion of the circuit upon the helix. 
7. In connection with this subject, the following experiment 
was made (“Phil. Trans.” 1852, p. 33). A magnet had a por- 
tion of wire rigidly attached to it in the form of a loop (as in 
fig. 5), and magnet and wire were rotated together, when no 
current was found to exist in the circuit made up of the magnet 
and wire. This result was accounted for as follows, on the 
same hypothesis. When the magnet and attached wire rotated 
together as a whole, the magnet was supposed to rotate through 
its own system of force, thereby becoming charged up statically ; 
the wire was supposed to rotate through the external system 
of force of the magnet (the wire also becoming charged up 
statically), the magnet and the wire thus tending to generate a 
current in opposite directions, so that this was assumed to be the 
cause why no actual current was produced. The observed result 
may however be simply explained on the consideration that, 
when the magnet and wire are rotated together , and rigidly 
attached, the two are relatively at rest, and therefore no indue- 
