322 THEORIES OF INHERITANCE AND DEVELOPMENT 



Ascaris demonstrate the truth of this view in a particular case ; for here 

 all of the somatic nuclei lose a portion of their chromatin, and only the 

 progenitors of tJie germ-nuclei retain tlic entire ancestral heritage. Boveri 

 himself has in a measure pointed out the significance of his discovery, 

 insisting that the specific development of the tissue-cells is condi- 

 tioned by specific changes in the chromatin that they receive, 1 though 

 he is careful not to commit himself to any definite theory. It hardly 

 seems possible to doubt that in Ascaris the limitation of the somatic 

 cells in respect to the power of development arises through a loss of 

 particular portions of the chromatin. One cannot avoid the thought 

 that further and more specific limitations in the various forms of 

 somatic cells may arise through an analogous process, and that we 

 have here a key to the origin of nuclear specification without recourse 

 to the theory of qualitative division. We do not need to assume that 

 the unused chromatin is cast out bodily ; for it may degenerate and 

 dissolve, or may be transformed into linin-substance or into nucleoli. 

 This suggestion is made only as a tentative hypothesis, but the 

 phenomena of mitosis seem well worthy of consideration from this 

 point of view. Its application to the facts of development becomes 

 clearer when we consider the nature of the nuclear "control " of the 

 cell, i.e. the action of the nucleus upon the cytoplasm. Strasburger, 

 following in a measure the lines laid down by Nageli, regards the 

 action as essentially dynamic, i.e. as a propagation of molecular 

 movements from nucleus to cytoplasm in a manner which might be 

 compared to the transmission of a nervous impulse. When, however, 

 we consider the rdle of the nucleus in synthetic metabolism, and the 

 relation between this process and the morphological formative power, 

 we must regard the question in another light ; and opinion has of 

 late strongly tended to the conclusion that nuclear "control" can 

 only be explained as the result of active exchanges of material 

 between nucleus and cytoplasm. De Vries, followed by Hertwig, 

 assumes a migration of pangens from nucleus to cytoplasm, the 

 character of the cell being determined by the nature of the migrat- 

 ing pangens, and these being, as it were, selected by circumstances 

 (position of the cell, etc.). But, as already pointed out, the pangen 

 hypothesis should be held quite distinct from the purely physiologi- 

 cal aspect of the question, and may be temporarily set aside ; for 

 specific nuclear substances may pass from the nucleus into the 

 cytoplasm in an unorganized form. Sachs, followed by Loeb, has 

 advanced the hypothesis that the development of particular organs 

 is determined by specific "formative substances" which incite cor- 

 responding forms of metabolic activity, growth, and differentiation. 



1 '9i, P- 433- 



