Interaction of Genotype and Environment in Continuous Variation: I. Description

Genes can take part in four types of interaction in producing their effects. First, allelic genes may interact to give dominance, or even over-dominance where the heterozygote transcends in its effects the range of expression delimited by homozygotes. Second, non-allelic genes may interact in a variety of ways which have received various names in classical genetics, but which are commonly pooled under the comprehensive heading of epistasis in the study of continuous variation. Third, genes may interact with permanent or semi-permanent components of the cytoplasm. And fourth, they may interact with agencies external to the organism and hence referable to the environment. The distinctions among these classes of interactions may not always be final. Although persistent, components of the cytoplasm may themselves be products of earlier gene action, even the products of action of genes no longer present in the nucleus with which the cytoplasm is associated. Equally, an organism may characteristically alter its environment and in this way express the earlier action of genes. To the extent that this is the case, some cytoplasmic and environmental interactions may ultimately be referrable to wholly genic interactions; but even so, they may be still more conveniently handled in their separate classes. In such a matter experience must be the guide. Provision has been made for the part played by dominance, or allelic interaction, right from the first in biometrical genetics [Fisher 19181: indeed the method of describing, and hence of measuring the contribution that dominance makes to the variation, has varied but slightly from the pattern Fisher originally set. Fisher considered epistasis too, but the description and measurement of these non-allelic interactions have posed more difficult problems. A general method of describing and classifying them has, however, now been devised [Cocker-