Summary.Major problems in the field are presented through a brief review of Burt's work and a critical account of the Hawaiian and British schools of biometrical genetics. The need for a covariance term is stressed and an estimate, based on several studies, is given. The merits and demerits of Jencks' study are discussed. There follows an account of the principle of genetic variation with age, a new concept to the heredity‐environment field. The new solution, based largely on Jencks' approach and incorporating the genetic variation principle, is then outlined. Its results may be regarded as a compromise between Jencks and Burt in that it agrees with Jencks on the matter of covariance (about 18–19 per cent) but is closer to Burt as far as the relative importance of hereditary and environmental differences is concerned (about 75:25). The consistency of the solution with other studies in the field is demonstra