My purpose here is to elaborate the reasons I maintain that Salmon has not been completely successful in reporting the history of work on explanation. The most important limitation of his account is that it does not emphasize the critical necessity to embrace a suitable conception of probability in the development of the theory of probabilistic explanation.