The fact that those who lack the blood group antigen A have anti-A antibody, and those who lack B have anti-B (Landsteiner, 1900, 1901), leads in a number of cases to an incompatibility between maternal serum and foetal antigens inherited from the father. Such serologically incompatible preg nancies have been thought to be a possible cause of various morbid conditions in foetus or mother (e.g., Gruhzit, 1923, 1924), although Tovey (1945) considers that placental resistance to the passage of agglutinin, and the presence of ABO substance in foetal plasma, together with the low sensitivity of foetal red cells, may greatly lessen the effect on the foetus. Hirszfeld and Zborowski (1925, 1926) found that, in the mating class Father A x Mother O, there were relatively fewer A offspring than in the reciprocal class Father O x Mother A, and also that there were relatively fewer O children from Mother A x Father O than from Mother O x Father A. Later results of Hirszfeld (1928) agreed with this, but Koller (1931) showed that the findings were not significant. An exhaustive review of the literature by Hirszfeld (1934) reversed his previous conclusions, and his further work (1938) did not confirm them. Levine (1943) assembled data apparently in support not of Hirszfeld's theory of constitutional incompatibility between a mother and a foetus of different groups but of Gruhzit's theory of serological incompatibility. Waterhouse and Hogben (1947) analysed twelve carefully chosen family studies made between 1927 and 1944 primarily to decide between the modes of inheritance of the ABO blood groups proposed by von Dungern and Hirszfeld (1910) and Bernstein (1924, 1925). Selecting the largest reciprocal mating class?namely Father A x Mother O and Father O x Mother A?in the pooled material from these studies, they found: (/) a significant shortage of families of the class Father A x Mother O as compared with its reciprocal; (//) a highly significant shortage of group A children from the class Father A x Mother O: