agitate its atoms and throw them out of line. In steel, which is
iron with a small admixture of carbon, the atoms are not so free
as in soft iron, and hence, while iron easily loses its magnetism,
steel retains it, even under a shock, but not under a cherry red-
heat. Nevertheless, if we put the atoms of soft iron under a
strain by bending it, we shall find it retain its magnetism more
like a bit of steel.

It has been found, too, that the atoms show an indisposition to be
moved by the magnetising force which is known as HYSTERESIS. They
have a certain inertia, which can be overcome by a slight shock,
as though they had a difficulty of turning in the ranks to take up
their new positions. Even if this molecular theory is true,
however, it does not help us to explain why a molecule of matter
is a tiny magnet. We have only pushed the mystery back to the
atom. Something more is wanted, and electricians look for it in
the constitution of the atom, and in the luminiferous ether which
is believed to surround the atoms of matter, and to propagate not
merely the waves of light, but induction from one electrified body
to another.

We know in proof of this ethereal action that the space around a
magnet is magnetic. Thus, if we lay a horse-shoe magnet on a table
and sprinkle iron filings round it, they will arrange themselves
in curving lines between the poles, as shown in figure 28. Each
filing has become a little magnet, and these set themselves end to
end as the molecules in the metal are supposed to do. The "field"
about the magnet is replete with these lines, which follow certain
curves depending on the arrangement of the poles. In the horse-
shoe magnet, as seen, they chiefly issue from one pole and sweep