The book attempts to develop an account of simplicity in terms of testability, and to use this account to provide an adequate characterization of induction, one immune to the class of problems suggested by Nelson Goodman. It is then shown that the past success of induction, thus characterized, constitutes evidence for its future success. A qualitative measure of confirmation is developed, and this measure - along with the considerations of simplicity - is used to provide an account of the consilience of inductions, and also an inductivist account of the structure and progress of scientific theory. An appendix extends the treatment of simplicity to statistical distributions and provides a reasonable interpretation of the maximum entropy principle. Thus, this book is an attempt to characterize induction in terms of a well-defined notion of simplicity and to use that characterization as a basis of an account of empirical, and in particular, scientific reasoning.