Most equilibrium statistical mechanics textbooks do not discuss how we can empirically demonstrate `atomism, or how we can count the number of particles. Since the modern statistical thermodynamics is based on atomism, and since physics must be an empirical science, we must be able to know indeed there are atoms and we can count them.
Unfortunately, if we can observe only systems in equilibrium, even if in some cases we may infer the particular nature of the microscopic world, we cannot actually estimate the size of the molecules or count the number of them. Thus, we must be able to study fluctuations (or mesoscopic systems) under slightly nonequilibrium conditions. Thus, we study transport phenomena and Brownian motion first up to Section 8.
Problem source book by R Kubo and his colleagues
As explicitly introduced in the self-study guide, the author recommends Thermodynamics and Statistical mechanics as problem source book.
As a professional statistical physicist, it is not important that she or he can solve exercises and exam problems quickly. Such problems are easy to find with solutions; if you are a professor you could even ask your students for their solutions; as professional scientists, what you have to do is to find new interesting problems and to detect difficulties and flaws in the `established portion ‘ of your disciplines. Still, you must have firsthand look at what sort of problems we can solve easily. Thus, you should be able to solve problems in part [A] of each chapter.
There are much newer problem books such as a collection of qual problems, but the source books introduced here are systematically edited books with summaries of key points of major topics, so you could use them to secure your basic skill and knowledge of statistical thermodynamics. The author learned the subject essentially through solving all the problems in the books.
