Many protein conjugates, prepared in vitro with reactive simple chemicals, are highly effective reagents for inducing the formation of antibodies which are specific for the simple chemical components of the conjugates (1). Antibodies may also be formed directly in response to injections of certain simple chemicals, providing the simple substances used are capable of reacting in vivo to form protein conjugates (24). These generalities are only partly applicable to the induction of contact skin sensitivity. In order for simple chemicals to induce the latter hypersensitive state, the in ~vo formation of protein conjugates (6, 7) is also obligatory (5). However, with a few possible exceptions (8), protein conjugates prepared in ~itro have little, if any, capacity to induce contact skin sensitivity (9). Part of the present work is devoted to the presentation of evidence which supports and extends older observations (8, 9) concerning the unexpected inability of protein conjugates, made in tritro, to induce contact skin sensitivity. The different conditions required for induction of antibody formation and for induction of contact skin sensitivity constitute an interesting distinction between these two inducible responses. But this difference introduces a problem in respect to induction of contact skin sensitivity: If the formation in vivo of protein conjugates is obligatory, why are "synthetic" conjugates, prepared in the test tube,