Carl Liebermeister's hypergeometric tails (Q1344939)

In 1877, in a medical context, Liebermeister studied the possibility of distinguishing between equality and inequality of success probabilities in two (small) series of binomial trials. Starting from a Laplacian formula based on the existence of a uniform prior distribution and assuming that the two probabilities coincided, he considered the size of the tail probability (of the hypergeometric distribution). The author reconstructs Liebermeister's insufficient intermediate calculations and indicates that his test can still be applied and that his main formula has hardly ever reappeared.