450 HEREDITY AND DEVELOPMENT 



germ-plasm of the fertilised ovum, and are its lineal descendants ; 

 but the somatic cells are dominated by particular segregated and 

 liberated sets of determinants, whereas the germ-cells are those, 

 or the descendants of those, that retain the complete equipment. 

 In studying the development of the threadworm of the horse 

 (A scaris megalocephala), Boveri found that the two first segmen- 

 tation-cells both receive the four chromosomes characteristic 

 of the species; one gives rise to all the body-cells, the other 

 to all the germ-cells. In the lineage of the former there is a 

 visible reduction of the chromatin ; in the lineage of the other 

 there is no such reduction. This is perhaps the clearest of all 

 cases, and the case of some of the Diptera is almost as clear. 

 But theoretically it makes no difference how long the " germ- 

 track " may be, or how long it may be before recognisable 

 germ-cells are seen in the developing organism. In some familiar 

 cases the alternation of generations in Hydroids the repro- 

 ductive cells, as such, are not demonstrable till after the asexual 

 generation forms a sexual bud ; and yet, even here, we know 

 some very interesting facts regarding the germ- cell lineage. 



3. Note on Rival Theories 



Darwin's Theory of Gemmules. Darwin's provisional theory 

 of pangenesis suggests, as we have already seen, that particular 

 cells of the body give off representative gemmules, and that 

 these are collected in the reproductive cells. When the fertilised 

 egg-cell divides and redivides, the army of gemmules is contained 

 in each cell ; but at every stage of development particular kinds of 

 gemmules are stimulated to activity, and proceed to influence 

 the area in which they find themselves an area corresponding 

 to that from which they were originally given off. As Weismann 

 points out, this hypothesis requires us to postulate an enormous 

 number of specific stimuli, distributed through the crowd of 



