itions more readily. The molecules of a magnet, then, are
differently arranged from those in an unmagnetized piece of iron or
steel; and, for every new arrangement of the molecules of a mass of any
kind, we always have some new physical property developed. The same
identical substance may appear as charcoal, coke, plumbago, anthracite
coal, and diamond. Hence a magnet is a machine in which other forces
acting upon it are transformed in character, and re-appear as
attractions and repulsions of other kinds of matter: this transformation
cannot take place, and hence magnetism cannot become apparent, only upon
the condition of another force acting in concert with it; and, if at any
time it may seem to be acting without such external force, it is done at
the expense of the heat it has absorbed, and therefore the magnet must
at such time be losing temperature proportional to the work done. This I
have discovered to be true by making a magnet to exert its force in
front of a thermo-pile, which uniformly exhibits a cooled face under
such conditions. What the particular form of the motion may be that we
call magnetism, is not yet made out; but that it is some form of motion,
is very evident. The following experiments may throw some light upon it.
Last August Mr. Kerr read a paper before the British Association of
Science, in which was detailed the following experiment: The pole of an
electro-magnet was nicely polished so as to reflect light like a mirror.
A beam of sunlight was permitted to fall upon it, and be reflected to a
convenient place for examination. A current of electricity was sent
through the coil, which of course rendered the iron magnetic; and it was
noticed that the light that was reflected from the pole was circularly
polarized: that is, the motion of a ray, instead of being a simple
undulatory movement, was now made to assume such a motion as the water
from a garden-hose has when the nozzle is swung round in a circle while
the water is escaping from it. After reading the account of it, it
occurred to me that the converse experiment might be tried; that is to
say, the effect of a circularly polarized beam of light upon a piece of
steel. By concentrating a large beam of ordinary plane polarized light
with a quartz lens, and passing it through a quarter wave-plate at the
proper angle, a powerful beam of circularly polarized light was
obtained. At the focus of this beam a fine cambric needle without
magnetism was plac