IV.] FOUNDATION OF A THEORY OF HEREDITY. 237 



as an explanation, even though the nature of this difference is 

 entirely unknown, because polar repulsion is not developed 

 between the male and female halves of the nucleus, but within 

 each male and each female half. We are thus forced to con- 

 clude that increase in the quantity of the nucleus affords an 

 impulse for division, the disposition towards it being already 

 present. It seems to me that this view does not encounter an}*- 

 theoretical difficulties, and that it is an entirely feasible hypo- 

 thesis to suppose that, besides the internal conditions of the 

 nucleus, its quantitative relation to the cell-body must be taken 

 into especial account. It is imaginable, or perhaps even pro- 

 bable, that the nucleus enters upon division as soon as its 

 idioplasm has attained a certain strength, quite apart from the 

 supposition that certain internal conditions are necessary for 

 this end. As above stated, such conditions may be present, but 

 division may not occur because the right quantitative relation 

 between nucleus and cell-body, or between the different kinds 

 of nuclear idioplasm, has not been established. I imagine that 

 such a quantitative deficiency exists in an ^g%, which, after the 

 expulsion of the ovogenetic nucleoplasm in the polar bodies, 

 requires fertilization in order to begin segmentation. The fact 

 that the polar bodies were expelled proves that the quantity of 

 the nucleus was sufficient to cause division, while afterwards it 

 was no longer sufficient to produce such a result. 



This suggestion will be made still clearer by an example. In 

 Ascaris megalocephala the nuclear substance of the female pro- 

 nucleus forms two loops, and the male pronucleus does the 

 same ; hence the segmentation nucleus contains four loops, and 

 this is also the case with the first segmentation spheres. If 

 we suppose that in embryonic development, the first nuclear 

 division requires such an amount of nuclear substance as is 

 necessary for the formation of four loops, — it follows that an 

 Qgg, which can only form two or three loops from its nuclear 

 reticulum, would not be able to develope parthenogenetically, 

 and that not even the first division would take place. If we 

 further suppose that, while four loops are sufficient to start 

 nuclear division, these loops must be of a certain size and 

 quantity in order to complete the whole ontogeny (in a certain 

 species), it follows that eggs possessing a reticulum which 

 contains barely enough nuclear substance to divide into four 



