the same problems raised by workers in the other fields of
physics, and shall secure some very interesting bits of testimony as to
the solution of questions already outlined.

The line of observation which has led to the most striking results has
to do, as already suggested, with the conduction of electricity through
liquids and gases. It has long been known that many liquids conduct
electricity with relative facility. More recently it has been observed
that a charge of electricity carried by any liquid bears a curious
relation to the atomic composition of that liquid. If the atom in
question is one of the sort that can combine with only a single other
atom (that is to say, a monovalent atom), each atom conveys a unit
charge, which is spoken of as an ion of electricity. But if a divalent
atom is in question the charge carried is double, and, similarly, a
trivalent atom carries a triple charge. As there are no intermediate
charges it is obvious that here a very close relation is suggested
between electrical units and the atomic units of matter.

This, however, is only a beginning. Far more interesting are the results
obtained by the study of gases in their relation to the conduction
of electricity. As is well known, gases under ordinary conditions are
nonconductors. But there are various ways in which a gas may be changed
so as to become a conductor; for example, by contact with incandescent
metals or with flame, or by treating with ultra-violet light, with
Rontgen rays, or with the rays of a radio-active substance. Now the
all-important question is as to just what change has taken place in the
gas so treated to make it a conductor of electricity. I cannot go into
details here as to the studies that have been addressed to the answer
of this question, but I will briefly epitomize what, for our present
purpose, are the important results. First and foremost of these is the
fact that a gas thus rendered conductive contains particles that can
be filtered out of it by passing the gas through wool or through water.
These particles are the actual agents of conduction of electricity,
since the gas when filtered ceases to be conductive. But there is
another way in which the particles may be removed--namely, by action
of electricity itself. If the gas be caused to pass between two metal
plates, one of them insulated and attached to an electrometer, a charge
of positive electricity at high potential sent through the other plate
wi