ed when taken in small mass; M. Meslin
proves this in a very ingenious way as regards optical properties, and
Mr Vincent in respect of electric conductivity. M. Houllevigue, who,
in a chapter of his excellent work, _Du Laboratoire a l'Usine_, has
very clearly set forth the most interesting considerations on atomic
hypotheses, has recently demonstrated that copper and silver cease to
combine with iodine as soon as they are present in a thickness of less
than thirty millionths of a millimetre. It is this same dimension
likewise that is possessed, according to M. Wiener, by the smallest
thicknesses it is possible to deposit on glass. These layers are so
thin that they cannot be perceived, but their presence is revealed by
a change in the properties of the light reflected by them.

Thus, below fifty to thirty millionths of a millimetre the properties
of matter depend on its thickness. There are then, no doubt, only a
few molecules to be met with, and it may be concluded, in consequence,
that the discontinuous elements of bodies--that is, the molecules--
have linear dimensions of the order of magnitude of the millionth of a
millimetre. Considerations regarding more complex phenomena, for
instance the phenomena of electricity by contact, and also the kinetic
theory of gases, bring us to the same conclusion.

The idea of the discontinuity of matter forces itself upon us for many
other reasons. All modern chemistry is founded on this principle; and
laws like the law of multiple proportions, introduce an evident
discontinuity to which we find analogies in the law of electrolysis.
The elements of bodies we are thus brought to regard might, as regards
solids at all events, be considered as immobile; but this immobility
could not explain the phenomena of heat, and, as it is entirely
inadmissible for gases, it seems very improbable it can absolutely
occur in any state. We are thus led to suppose that these elements are
animated by very complicated movements, each one proceeding in closed
trajectories in which the least variations of temperature or pressure
cause modifications.

The atomistic hypothesis shows itself remarkably fecund in the study
of phenomena produced in gases, and here the mutual independence of
the particles renders the question relatively more simple and,
perhaps, allows the principles of mechanics to be more certainly
extended to the movements of molecules.

The kinetic theory of gases can point to unquestion