Although the mechanism of antibody formation has naturally been a principal problem of immunology from its beginning, it remains still unsolved. Even a tentative hypothesis, giving a rational chemical picture of how antibody formation may occur should at least serve to focus though t and suggest experiment upon this problem. It would seem that the specific correspondence of antigen and antibody can best be explained by some sort of directing mechanism in antibody synthesis. A hypothesis involving such a mechanism ishere proposed and certain of its implications are pointed out. “By definition an antigen is any substance which, introduced into the tissues orcirculating fluids of an animal, causes the appearance in these fluids, sooner or later, of substances which react specifically with the antigenic substance; i.e., specific antibodies” (1). The essential fact to be explained is that every or nearly every known sort of soluble protein which contains its full quota of amino acids is antigenic, i.e., gives rise on injection to antibodies which possess specific elective combining affinity with itself.