Eighty years ago: the beginnings of population genetics.

I suppose that every teacher of elementary genetics has at one time or another encountered the b lief that in the absence of counteracting factors there should be three times as many dominant as recessive phenotypes in the population. It was this statement that induced G. H. HARDY in 1908 to write his famous paper. He started out somewhat apologetically, saying: “I am reluctant to intrude in a discussion concerning matters of which I have no expert knowledge, and I should have expected the very simple point which I wish to make to have been familiar to biologists.” The principle was independently published a few months earlier by WEINBERG (1908), but this paper remained unknown to most English-speaking geneticists, so for many years the principle was called HARDY’S law. Since the 1940s, thanks to CURT STERN’S (1943) setting the record straight, it is referred to as the HARDY-WEINBERG law. It is so self-evident that it hardly needed to be “discovered”; SEWALL WRIGHT, among others, used it before he had heard of either HARDY or WEINBERG. Yet, trivial as it appears, the H-W principle is the foundation for diploid population genetics. The law can be stated in two ways. First, it says that if mating is at random in a large population and mutation, migration, and selection are absent, the genotype proportions do not change from generation to generation. This is almost a tautology: if nothing changes the frequencies, they won’t change. But the principle does make clear that, with inbreeding or assortative mating, the genotype frequencies can change while the allele frequencies do not. Second, and much more usefully, the law permits the prediction of genotype frequencies from knowledge of gene frequencies. If alleles A and a are in the proportions p and q, the three zygotic types AA, Aa, and aa are in the proportions p2, 2pq, and q2. Thus, equations can be written in terms of the more basic units of allele frequencies, and hypotheses about how phenotypes are inherited can be tested from population data.