The relationship of somatic polyploidy to cellular differentiation is discussed. Evidences for the occurrence of endopolyploidy on the basis of auxin‐induced divisions are questioned and considered valid in the light of radioisotope studies of DNA synthesis. The non‐universality of endopolyploidy is shown by known exceptions. The occurrence of endopolyploidy in varying degrees in gametophytes of the same species of fern which are of different basic chromosome numbers shows its variability within a given morphogenetic pattern. From these evidences, and from an analysis of the separate component processes of cell multiplication, it can be concluded that polyploidy is not a necessary precursor or concomitant of cell differentiation, but rather is a type of differentiation in itself. As such, it may be one of a cumulative set of steps to yet other differentiation, but this may be incidental to the intrinsic nature and role of the nucleus.