lled an
electro-magnetic grip upon the whole of the ether, and any change in the
former brings some change in the latter.

Lastly, the phenomenon called induction may be mechanically conceived.

It is well known that a current in a conductor makes a magnet of the
wire, and gives it an electro-magnetic field, so that other magnets in
its neighbourhood are twisted in a way tending to set them at right
angles to the wire. Also, if another wire be adjacent to the first, an
electric current having an opposite direction is induced in it. Thus:

Consider a permanent magnet A (Fig. 15), free to turn on an axis in the
direction of the arrow. If there be other free magnets, B and C, in
line, they will assume such positions that their similar poles all point
one way. Let A be twisted to a position at right angles, then B will
turn, but in the opposite direction, and C in similar. That is, if A
turn in the direction of the hands of a clock, B and C will turn in
opposite directions. These are simply the observed movements of large
magnets. Imagine that these magnets be reduced to atomic dimensions, yet
retaining their magnetic qualities, poles and fields. Would they not
evidently move in the same way and for the same reason? If it be true,
that a magnet field always so acts upon another as to tend by rotation
to set the latter into a certain position, with reference to the stress
in that field, then, _wherever there is a changing magnetic field, there
the atoms are being adjusted by it_.

[Illustration: FIG. 16.]

Suppose we have a line of magnetic needles free to turn, hundreds or
thousands of them, but disarranged. Let a strong magnetic field be
produced at one end of the line. The field would be strongest and best
conducted along the magnet line, but every magnet in the line would be
compelled to rotate, and if the first were kept rotating, the rotation
would be kept up along the whole line. This would be a mechanical
illustration of how an electric current travels in a conductor. The
rotations are of the atomic sort, and are at right angles to the
direction of the conductor.

That which makes the magnets move is inductive magnetic ether stress,
but the advancing motion represents mechanical energy of rotation, and
it is this motion, with the resulting friction, which causes the heat in
a conductor.

What is important to note is, that the action in the ether is not
electric action, but more properly the result of electro-