The title of this book can lead to misunderstandings: apart from the Maxwell and Arrhenius demons, which in the last chapter are taken to absurdity, everything is very down-toearth and reasonable. It is doubtful whether the consistent omission of differential derivatives and integrals will prove popular, no matter how skilfully done, for a follower of the creed that school mathematics are the work of the devil will hardly have the ambition to achieve statistical understanding of thermodynamics. The book offers some valuable advantages. In the first chapter the author presents a very interesting history of the develop ment of thermodynamics without succumbing to the obvious temptation of using unclear expressions. Unfortunately, the history of thermodynamics as a didactic method has been rather overlooked, an unjustifiable trend as Zernike shows. The avoidance of differential and integral calculus necessitates, in the second chapter, derivation of the Clausius-Clapeyron equation by a cycle. Nevertheless, the term "function of state" is explained. Chapters 3 and 4 are the best part of the book because they contain at last statistical thermodynamics which any chemist can understand. This fills a sorely-felt gap, for in