In 1699 John Craig published an underappreciated book on the probability of historical events. A new interpretation of Craig's work is offered, and it is argued that his formula for the probability of testimony was tantamount to a logistic model for the posterior odds. A modern model for the transmission of evidence is given that is qualitatively similar to Craig's, and Craig's attempt to date the Second Coming of Christ is discussed. Finally, Craig's model is fit to more modern data (from 1882) on the birth date of Laplace.