Bateson was born in Whitby on the Yorkshire coast, the son of William Henry Bateson, Master of St John's College, Cambridge. He was educated at Rugby School and at St John's College in Cambridge, where he graduated BA in 1883 with a first in natural sciences.
Taking up embryology, he went to the United States to investigate the development of Balanoglossus. This worm-like enteropneust hemichordate led to his interest in vertebrate origins. In 1883-4 he worked in the laboratory of William Keith Brooks, at the Chesapeake Zoölogical Laboratory in Hampton, Virginia. Turning from morphology to study evolution and its methods, he returned to England and became a Fellow of St John's. Studying variation and heredity, he travelled in western Central Asia.
Bateson's work published before 1900 systematically studied the structural variation displayed by living organisms and the light this might shed on the mechanism of biological evolution, and was strongly influenced by both Charles Darwin's approach to the collection of comprehensive examples, and Francis Galton's quantitative ("biometric") methods. In his first significant contribution, he shows that some biological characteristics (such as the length of forceps in earwigs) are not distributed continuously, with a normal distribution, but discontinuously (or "dimorphically"). He saw the persistence of two forms in one population as a challenge to the then current conceptions of the mechanism of heredity, and says "The question may be asked, does the dimorphism of which cases have now been given represent the beginning of a division into two species?”
In his 1894 book, "Materials for the study of variation", Bateson took this survey of biological variation significantly further. He was concerned to show that biological variation exists both continuously, for some characters, and discontinuously for others, and coined the terms "meristic" and "substantive" for the two types. In common with Darwin, he felt that quantitative characters could not easily be "perfected" by the selective force of evolution, because of the perceived problem of the "swamping effect of intercrossing", but proposed that discontinuously varying characters could. Amongst other interesting observations he noted variations in which an expected body-part has been replaced by another (which he called homeotic). The animal variations he studied included bees with legs instead of antennae; crayfish with extra oviducts; and in humans, polydactyly, extra ribs, and males with extra nipples. Importantly, Bateson wrote, "The only way in which we may hope to get at the truth is by the organization of systematic experiments in breeding, a class of research that calls perhaps for more patience and more resources than any other form of biological enquiry. Sooner or later such an investigation will be undertaken and then we shall begin to know."
In 1897 he reported some significant conceptual and methodological advances in his study of variation. "I have argued that variations of a discontinuous nature may play a prepondering part in the constitution of a new species." He attempts to silence his critics (the "biometricians") who misconstrue his definition of discontinuity of variation by clarification of his terms: "a variation is discontinuous if, when all the individuals of a population are breeding freely together, there is not simple regression to one mean form, but a sensible preponderance of the variety over the intermediates… The essential feature of a discontinuous variation is therefore that, be the cause what it may, there is not complete blending between variety and type. The variety persists and is not “swamped by intercrossing”. But critically, he begins to report a series of breeding experiments, conducted by Edith Saunders, using the alpine brassica Biscutella laevigata in the Cambridge botanic gardens. In the wild, hairy and smooth forms of otherwise identical plants are seen together. They intercrossed the forms experimentally, “When therefore the well-grown mongrel plants are examined, they present just the same appearance of discontinuity which the wild plants at the Tosa Falls do. This discontinuity is, therefore, the outward sign of the fact that in heredity the two characters of smoothness and hairiness do not completely blend, and the offspring do not regress to one mean form, but to two distinct forms.”