The main aim of this essay is to show that Wittgenstein's general conception of logic, as expressed in the Tractatus Logico-Philosophicus, and his later conception of philosophical grammar can be reconciled with each other, provided that his earlier and his later views are modified in the light of certain specific criticisms. These criticisms are based on two grounds. One is Wittgenstein's early neglect of important differences between various philosophical activities and the methods employed in them, in particular between the roles of description and of reconstruction in philosophical analysis. The other is an incompleteness in his later description of ordinary language and of its connections with the specialist languages of mathematics and the sciences. An important example is his account of the nature of continuity, which Leibniz regarded as one of the two labyrinths of the human mind.