Karl Pearson's mathematization of inheritance: from ancestral heredity to Mendelian genetics (1895-1909).

Long-standing claims have been made for nearly the entire twentieth century that the biometrician, Karl Pearson, and colleague, W. F. R. Weldon, rejected Mendelism as a theory of inheritance. It is shown that at the end of the nineteenth century Pearson considered various theories of inheritance (including Francis Galton's law of ancestral heredity for characters underpinned by continuous variation), and by 1904 he 'accepted the fundamental idea of Mendel' as a theory of inheritance for discontinuous variation. Moreover, in 1909, he suggested a synthesis of biometry and Mendelism. Despite the many attempts made by a number of geneticists (including R. A. Fisher in 1936) to use Pearson's chi-square (X2, P) goodness-of-fit test on Mendel's data, which produced results that were 'too good to be true', Weldon reached the same conclusion in 1902, but his results were never acknowledged. The geneticist and arch-rival of the biometricians, Williams Bateson, was instead exceptionally critical of this work and interpreted this as Weldon's rejection of Mendelism. Whilst scholarship on Mendel, by historians of science in the last 18 years, has led to a balanced perspective of Mendel, it is suggested that a better balanced and more rounded view of the hereditarian-statistical work of Pearson, Weldon, and the biometricians is long overdue.