THEORY OF DEVELOPMENT 445 



But if the various kinds of determinants are to get into ap- 

 propriate cell-groups, this cannot be a matter of chance. There- 

 fore, we must further postulate that from the first each deter- 

 minant has a definite position in relation to its neighbours, 

 that the germ-plasm is not a mere loose aggregate of deter- 

 minants, but that it possesses a structure, an architecture, in 

 which the individual determinants have each their definite place. 

 It must be borne in mind that the germ-cell is a unity, a potential 

 organism, and not a heap of hereditary contributions. Weis- 

 mann supposes that the determinants are kept in relation to 

 one another by " vital affinities," by internal forces, some ex- 

 hibition of which is, indeed, demonstrable, as when a chromosome 

 or ribbon of ids splits into a double ribbon of ids. 



But if the mechanism of the distribution of determinants is 

 by cell-division — one of the features of which is that the chro- 

 mosomes are halved with minutiose accuracy, so that each of the 

 two daughter-cells obtains a longitudinal half of each chromo- 

 some — how does it come about that different determinants pass 

 into different cells of the embryo ? This difficulty led to the 

 further hypothesis that, while ids may divide into two identical 

 halves, they may also divide into two dissimilar halves. Weis- 

 raann supposed that besides integral {erhgleich) division of the 

 nucleus, there is also differential {erbungleich) division. The 

 reality of this differential division — which many histologists 

 vigorously dispute — cannot be directly demonstrated any more 

 than the splitting up of a complex molecule into different mole- 

 cules can be demonstrated. But in both cases we may infer 

 the occurrence from the results. It is not a hypothesis, but a 

 fact, that a cell may divide into two daughter-cells, one of which 

 goes to form ectoderm, while the other goes to form endoderm, 

 and this implies some sort of differential division. What in- 

 ternal forces or vital affinities are concerned we do not know. 



If an egg-cell can divide differentially into a primordial 

 ectoderm-cell and a primordial endoderm-ceU, or into a formative 



