“For Geographical Purposes”: Andrew Ellicott, Lewis and Clark, and the Longitude

Andrew Ellicott was a largely self-taught surveyor and practical astronomer who, between 1784 and 1813, helped locate and mark several state and international boundaries for the newly formed United States. Along with University of Pennsylvania professor Robert Patterson, Ellicott also helped train Meriwether Lewis, of Lewis and Clark fame, in astronomical methods for finding latitude and longitude. For longitude, Ellicott and Patterson both expressed a strong preference for the newer lunar distance in contrast to an older method using eclipses of Jupiter’s satellites, and both held the value of the newer method that had been demonstrated by comparing longitude determinations made by Ellicott. Interestingly, while their arguments are similar to what we now call “statistical” reasoning, they made no use of mathematical probability. This essay enlists Ellicott’s work to better understand non-probabilistic, quantitative argument as used at the turn of the eighteenth century.