UNION OF THE TWO THEORIES We have now to consider the attempts that have been made to transfer the localization-theory from the cytoplasm to the nucleus, and thus to bring it into harmony with the theory of nuclear idioplasm. These attempts are especially associated with the names of Roux, De Vries, Weismann, and Hertwig ; but all of them may be traced back to Darwin's celebrated hypothesis of pangenesis as a prototype. This hypothesis is so well known as to require but a brief review. Its fundamental postulate assumes that the germ-cells contain innumerable ultra-microscopic organized bodies or gemmules, each of which is the germ of a cell and determines the development of a similar cell during the ontogeny. The germ-cell is, therefore, in Darwin's words, a microcosm formed of a host of inconceivably minute self-propagating organisms, every one of which predetermines the formation of one of the adult cells. De Vries ('89) brought this conception into relation with the theory of nuclear idioplasm by assuming that the gemmules of Darwin, which he called pangens, are contained in the nucleus, migrating thence into the cytoplasm step by step during ontogeny, and thus determining the successive stages of development. The same view was afterwards accepted by Hertwig and The theory of germinal localization is thus transferred from the cytoplasm to the nucleus. It is not denied that the egg-cytoplasm may be more or less distinctly differentiated into regions that have a constant relation to the parts of the embryo. This differentiation is, however, conceived, not as a primordial characteristic of the egg, but as one secondarily determined through the influence of the nucleus. Both De Vries and Weismann assume, in fact, that the entire cytoplasm is a product of the nucleus, being composed of pangens that migrate out from the latter, and by their active growth and multiplication build up the cytoplasmic