The classical theory of monopolistic pricing (the Cournot-Marshall theory) is not entirely valueless, especially when overhauled so as to deal not only with the instantaneous maximization of monopoly gain but also with maximization over time. But it works with assumptions that are so restrictive as to exclude its direct application to reality. In particular it cannot be used for what it is being used in current teaching, namely, for a comparison between the way in which a purely competitive economy functions and the way in which an economy functions that contains substantial elements of monopoly.