... was one of the first people to compute orbits for double stars, and x Ursae Majoris was the system that he selected for the initial demonstration of his method, in 1831. Priority in that field, however, must go to Savary, who gave, in Connaissance des Temps for 1830 (which according to its title page was published in 1827), a description of a method of calculating orbits for double stars. In a supplementary paper in the same volume, he opens by saying (in French of course: the following is my translation), "It has occurred to me that it could be useful to add to [the formulae for computing orbits] their numerical application to a specific example. So I will suppose the following data, without attributing any reality to them" - and he lists four dates with associated separations and position angles, all highly reminiscent of, but not identical with, actual observations of x Ursae Majoris. Five pages later, at the conclusion of some stiff mathematics, he ingenuously remarks, "The data from which we started in the preceding calculations are very close to four observations of the double star x of the great bear"; he then goes on to tabulate all 11 of the observations available at the time, and to show that the computed orbit satisfies them very well. In fact, Sir John Herschel, whose orbit was, on his own admission, not nearly such a good match, complained with good reason that Savary had cheated - not Sir John's word: that is my précis of the gentlemanly circumlocution that occupies nearly a whole page (his paragraph 64) - by choosing to solve data that were only approximately the observed ones!