The title of this Symposium ‘From Mendel’s factors to the genetic code’ emphasizes a single aspect of genetic research. That is unfortunate in a subject of such wide scope. Even in the most general terms, it naturally involves far more than analysing the fine structure of the genetic material and the physiological paths by which the genes produce the characters for which they are responsible. For it is of course essential to study the functions of those characters: the ways in which they influence the adjustment of organisms to their environment and provide both the heritable variability and stability needed in evolution. A few aspects of that work are indeed reviewed in the papers contributed on this occasion. Consequently a proportion of them treat of subjects far beyond the range of the Symposium title, and the present article is of this kind. Genetic polymorphism comprises a distinct type of variation, recognizable and endowed with predictable properties. It is defined as, ‘The occurrence together in the same habitat of two of more discontinuous forms, or phases, of a species in such proportions that the rarest of them cannot be maintained merely by recurrent mutation’ (Ford 1940). The significance of this definition becomes apparent on analysis. Evidently it excludes geographical and seasonal forms, as well as comtinuous variation falling within a curve of normal distribution. It excludes also rare mutants, whether recessive or not, that are eliminated by selection and maintained in the population merely by mutation pressure. Such discontinuous variation must be controlled by some switch mechanism. This is nearly always provided by a gene or a super-gene, which latter may of course include an inversion or, in special circumstances, even a whole chromosome or group of chromosomes. In theory it might be, but hardly ever is, environmental. It is true that very distinct seasonal forms may be produced in that way, such as the spring and summer emergence of the butterfly Araschnia levana (Nymphalidae). But these involve whole generations. A much more delicate adjustment would be required to evoke sharply contrasting phases within the same brood, while it is difficult indeed to see how their proportions could be accurately adjusted to meet changes in the ecological situation, a drawback apparently fatal to the environmental control of polymorphism (Ford 1965).