This paper arose out of a study of the notes that Joseph Agassi and Czeslaw Lejewski took at Karl Popper's seminar on Logic and Scientific Method (1954–1955). It ponders on a basic logical distinction Popper had made : between sound inference (valid inference with sound premises) and proof (a collection of inferences that show that a given sentence follows from any premise). The difference between sound inference and proof seems crucial to Popper's epistemology, especially to his emphasis on the distinctness of epistemology and methodology. In this paper, (1) The distinction is explained; (2) The difference is presented as the basis for Popper's view of the history of logic; (3) Some modern hesitations about all this are discussed.