WEISMANN'S THEORY OF HEREDITY 167 



(p. 130), but such cases appear to be very exceptional and the 

 segregation of the germ cells usually takes place much later and at 

 different stages of the development in different species of plants 

 and animals. The exact time at which the separation of the two 

 groups of cells takes place, however, does not seriously affect the 

 argument. In any case the ultimate distinction between germ 

 cells and somatic cells is supposed to lie in the fact that the 

 former retain each a complete sample of the ancestral germ 

 plasm, in which at any rate all the essential characters of the 

 organism are in some way or other represented, while the latter, 

 by a series of differential divisions, gradually undergo a further 

 segregation into the different tissue cells of the adult, each of 

 which contains (in an active condition) only a sample of that 

 part of the ancestral germ plasm which is appropriate to its own 

 particular requirements. 



Hence the germ cells, complete in themselves like so many 

 Protista, alone retain the power of giving rise to new generations 

 of complete individuals. The somatic cells have sacrificed this 

 power to their need for specialization. Certain phenomena, 

 however, such as the regeneration of lost parts in many animals 

 and the propagation of plants by buds and cuttings, necessitate 

 the supposition that some at any rate of the somatic cells must 

 retain a more complete sample of the ancestral germ plasm than 

 is necessary for their own development. For the hereditary sub- 

 stance of any particular cell Weismann adopts Nageli's term 

 " idioplasm," and in order to account for the phenomena just 

 referred to he is obliged to postulate the existence, in the 

 somatic cells in question, of " accessory idioplasm," which is only 

 called into activity under exceptional conditions, as, for example, 

 when it becomes necessary for a crab to regenerate a lost limb 

 or for an earthworm to renew a portion of its body which has 

 been bitten off by a bird or chopped off by a spade. 



Weismann's theory involves the assumption of great complexity 

 of structure in the germ plasm, which, as we have already 

 seen, he identifies with the chromatin substance of the nucleus 

 of the germ cells. He finds it necessary to assume the existence, 

 not only of " determinants," which correspond more or less closely 

 to Darwin's gemmules, and each of which is supposed to be 

 responsible for the development of some special inherited feature 

 of the organism, but also of structural units respectively of a 

 lower and a higher order. Thus each determinant is supposed 



