FOUNDATION OF A THEORY OF HEREDITY. 227 



power to complete it, which is absent. We know that force is 

 always bound up with matter, and it seems to me that such 

 instances are best explained by the supposition that too small an 

 amount of that form of matter is present, which, by its controlling- 

 agency, effects the building-up of the embryo by the transforma- 

 tion of mere nutritive material. This substance is the germ-plasm 

 of the segmentation nucleus, and I have assumed above that it is 

 altered in the course of ontogeny by changes which arise from 

 within, so that, when sufficient nourishment is afforded by the cell- 

 body, each succeeding stage necessarily results from the preceding 

 one. I believe that changes arise in the constitution of the 

 nucleoplasm at each cell-division which takes place during the 

 building-up of the embryo, changes which either correspond or 

 differ in the two halves of each nucleus. If, for the present, we 

 neglect the minute amount of unchanged germ-plasm which is 

 reserved for the formation of the germ-cells, it is clear that a great 

 many different stages in the development of somatic nucleoplasm 

 are thus formed, which may be denominated as stages i, 2, 3, 4, &c., 

 up to n. In each of these stages the cells differ more as develop- 

 ment proceeds, and as the number by which the stage is denomi- 

 nated becomes higher. Thus, for instance, the two first segmen- 

 tation spheres would represent the first stage of somatic nucleo- 

 plasm, a stage which may be considered as but slightly different 

 in its molecular structure from the nucleoplasm of the segmentation 

 nucleus ; the four first segmentation spheres would represent the 

 second stage ; the succeeding eight spheres the third, and so on. It 

 is clear that at each successive stage the molecular structure of the 

 nucleoplasm must be further removed from that of the germ-plasm, 

 and that, at the same time, the cells of each successive stage must 

 also diverge more widely among themselves in the molecular 

 structure of their nucleoplasm. Early in development each cell 

 must possess its own peculiar nucleoplasm, for the further course of 

 development is peculiar to each cell. It is only in the later stages 

 that equivalent or nearly equivalent cells are formed in -large 

 numbers, cells in which we must also suppose the existence of 

 equivalent nucleoplasm. 



If we may assume that a certain amount of germ-plasm must be 

 contained in the segmentation nucleus in order to complete the 

 whole process of the ontogenetic differentiation of this substance ; 



