Abstract
The most important thesis of "Of Miracles" has no special connection with miracles: I mean the perfectly general thesis that testimonial evidence should be evaluated by the method of balancing likelihoods, which is a relatively informal version of the calculus of changes (or of probabilities). C. S. Peirce argues that the method is radically unsuited to the assessment of historical testimony. In this paper, I do essentially two things: (1) set out both an informal and a formal account of Hume’s method; and (2) collect, systematize, and discuss Peirce’s somewhat scattered animadversions upon Hume’s use of this method. As part of (2), I explore some lines of thought that Peirce suggests but does not develop