Fundamentals of Statistical Physics

Abstract

When we know completely the initial condition of a physical system, quantum mechanics may predict exactly its future (apart from the partial unpredictability of the results of a measurement process). If the system is only partially known, however, we are not forced to give up completely the idea of making predictions on its future behavior. Statistical physics is the branch of physics that deals with systems only partially known. The lack of maximum knowledge of the initial condition of the system is often due to the extremely large number of degrees of freedom, as in the case of gases, but this is not necessarily always true, and statistical physics works equally well for simple systems when, for any reason, our knowledge is less than complete. Following the classical text of Tolman [448] we may say that the general nature of the statistical mechanical procedure for the treatment of complicated systems consists in abandoning the attempt to follow the precise changes in state that would take place in a particular system, and in studying instead the behavior of a collection or ensemble of systems of similar structure to the system of actual interest, distributed over a range of different precise states.