TABLE OF CONTENTS Our belief in the ultimate unity of scientific knowledge leads us to conclude that there must be some connection between the macroscopic (thermodynamic) and microscopic (mechanical) definitions of state. The first task of statistical mechanics is to find that connection. For the many particle systems that concern us, it is impossible to obtain an actual specification of the microscopic state and, even if such a specification were available, the equations of motion would be so complicated that they could not be solved. Also, we are not particularly interested in such detailed information. We are, however, very much interested in the possibility of interpreting thermodynamic data in microscopic terms, and in particular, it would be very desirable to obtain a method of computing macroscopic quantities from atomic properties, thereby making up for the inherent deficiency of thermodynamics.
In view of these considerations, any bridge connecting macroscopic and microscopic descriptions of material systems must have the following characteristics:
A great deal of microscopic information must be erased in constructing macro- from microstates so that we are not encumbered with an enormous number of microscopic parameters.
Enough microscopic information must be retained to enable the calculation of bulk properties from atomic properties.
Temporal causality and reversibility, which operate at the micro level, do not appear at the macro level. (Except insofar as introduced by special postulates such as the law of thermal conduction. The discipline of irreversible thermodynamics was developed to provide a causal description...
