THIS ESSAY was begun a long time ago, in 1962, when I spent a year in Rome on a Guggenheim Fellowship. That twenty one years were required to complete it is owing both to the character of the theory presented and to my peculiar habits of mind. The theory presented is a coherence theory of knowledge: the con ception of coherence is here dominant and pervasive. But considera tions of coherence dictate an attention to details. The fact of the matter is that I get hung up on details: everything must fit, and if it does not, I do not want to proceed. A second difficulty was that all the epistemological issues seemed too clear. That may sound weird, but that's the way it is. I write philosophy to make things clear to myself. If, rightly or wrongly, I think I know the answer to a question, I can't bring myself to write it down. What happened, in this case, is that I finally became persuaded, in the course of lecturing on epistemology to under graduates, that not everything was as clear as it should be, that there were gaps in my presentation that were seriously in need of filling.