efinite number
of holes in it, each of them four lines in depth and four in diameter.
Electrify this cylinder, and present to its superficies a small square of
gold-leaf, held to it by an insulating needle of gum lac, and bring this
square to an electrometer of great sensibility. The electrometer will
instantly show an electricity in the gold-leaf, similar to that of the
cylinder which had been brought into contact with it. The square of
gold-leaf having thus been discharged of its electricity, put it carefully
into one of the holes of the cylinder, _so_, namely, that it shall touch
only the bottom of the hole, and present it again to the electrometer. It
will be then found that the electrometer will exhibit no signs of
electricity whatsoever. From this it follows, that the electricity which
had been communicated to the cylinder had confined itself to the
_surface_.

If the time and the limit prescribed would admit, we could multiply
experiments, all tending to prove the same law; but we must be content
with the barely sufficient. But that the _chemical process_ acts in
_depth_, and first, therefore, _realizes_ and integrates the fluxional
power of magnetism and electricity, is involved in the _term_ composition;
and this will become still more convincing when we have learnt to regard
_decomposition_ as a mere co-relative, _i.e._ as decomposition relatively
to the body decomposed, but composition _actually_ and in respect of the
substances, _into_ which it was decomposed. The alteration in the specific
gravity of metals in their chemical amalgams, interesting as the fact is
in all points, is _decisive_ in the present; for gravity is the sole
_inward_ of inorganic bodies--it _constitutes_ their depth.

I can now, for the first time, give to my opinions that degree of
intelligibility, which is requisite for their introduction as hypotheses;
the experiments above related, understood as in the common mode of
thinking, prove that the magnetic influence flows in length, the electric
fluid by suffusion, and that chemical agency (whatever the main agent may
be) is qualitative and _in intimis_. Now my hypothesis demands the
converse of all this. I affirm that a power, acting exclusively in length,
is (wherever it be found) _magnetism_; that a power which acts _both_ in
length and in breadth, and _only_ in length and breadth, is (wherever it
be found) _electricity_; and finally, that a power which, together with
length and bread