The initiation of the cambium in the ray parenchyma of a shoot has been explained by two contradictory hypotheses: (i) the initiation process is dependent on inductive influences arising from the vascular bundles, or (ii) the process is an autonomic one in the sense that the initiation is determined at a very young stage of the shoot and proceeds without influence from other tissues. Experiments in which small pieces of interfascicular tissue have been isolated and replaced in reverse between the vascular bundles in the hypocotyl have shown that the initiation site and the radial polarity must have been determined many days before the first demonstrable mitotic activity had actually occurred. The experiments provided no evidence pointing to the existence of an inductive effect arising from the vascular bundle or an influence from a surface-dependent gradient on the initiation of the cambium. It is suggested here that the entire initiation process, including fixation of radial polarity, is determined at the primary meristem ring stage.