they existed as
molecular closed curves or loops, which, by the operation called
magnetisation, could, some of them, be opened out into loops of finite
area and spread out into space, where they are called "lines of force."
They then constitute the region called a magnetic field, which remains
a seat of so-called "permanent" magnetic activity, until by lapse of
time, excessive heat, or other circumstance, they close up again; and
so the magnet, as a magnet, dies. The magnetism itself, however, has
not really died, it has a perpetual existence; and a fresh act of
magnetisation can recall it, or something indistinguishable from it,
into manifest activity again; so that it, or its equivalent, can once
more interact with the rest of material energies, and be dealt with by
physicists, or subserve the uses of humanity. Until that time of
re-appearance its existence can only be inferred by the thought of the
mathematician: it is indeed a matter of theory, not necessarily
recognised as true by the practical man.

Our present view is that the act of magnetisation consists in a
re-arrangement and co-ordination of previously existing magnetic
elements, lying dormant, so to speak, in iron and other magnetic
materials; only a very small fraction of the whole number being usually
brought into activity at any one time, and not necessarily always the
same actual set. Only a small and indiscriminate selection is made from
all the molecular loops; and it can be a different group each time, or
some elements may be different and some the same, whenever a fresh
individual or magnet is brought into being.

All this can be said concerning the old process of magnetisation--the
process as it was doubtless familiar to the unknown discoverer of the
lodestone, to the ancient users of the mariner's compass, and to Dr
Gilbert of Colchester, the discoverer of the magnetised condition of
the Earth.

But within the nineteenth century a fresh process of magnetisation has
been discovered, and this new or electrical process is no longer
obviously dependent on the existence of antecedent magnetism, but seems
at first sight to be a property freshly or spontaneously generated, as
it were. The process was discovered as the result of setting
electricity into motion. So long as electricity was studied in its
condition at rest on charged conductors, as in the old science of
electrostatics or frictional electricity, it possessed no magnetic
properties whatev