Statistical Theories of Matter, Radiation and Electricity

HE major subjects of this article are two extensions of what formerly was called atomic theory--that is to say, the attempt to explain as many as possible of the properties of pieces of matter large enough to be visible and tangible and ponderable, by visualizing these as swarms of tiny particles each endowed with only a very few and simple qualities. Among the properties of gases, for example, are pressure and viscosity and entropy and temperature. Conceivably one might invest the ultimate atoms with all four. The atomic theory of gases as it stands today, however, is the outcome of a very different procedure. It is the achievement of an effort to interpret these four properties and several more as features of a hypothetical assemblage of very many corpuscles all alike, and not possessing them nor any others except position and velocity and mass (and moment of inertia, sometimes) and the liability to make elastic impacts with each other. On the whole the effort has been remarkably successful. Therefore viscosity and temperature and entropy are not attributed to single atoms, but pictures and expressions for them are derived as qualities of the assemblage. The theory which leads to these results is called statistical; it is based on certain assumptions which, in the form in which they were originally made, constitute the classical statistics. The successes of the classical statistics are a part of the evidence that matter is corpuscular. Once they were nearly the whole of the evidence, for they antedated the striking demonstrations of individual atoms which now spring to the mind whenever one is asked to state the reasons for accepting the atomic theory.