SUMMARY AA'D CONCLUSION 205 



lengthwise, and are distributed in the usual manner, and reappear 

 ill the same unuiber in all later stages. In other words, the second 

 polar body here plays the part of a sperm-nucleus, precisely as main- 

 tained by Boveri. 



In all individuals arising from eggs of the first type, therefore, the 

 somatic number of chromosomes is eighty-four ; in all those arising 

 from eggs of the second type, it is one hundred and sixty-eight. It 

 is impossible to doubt that the chromosomes of the first class are 

 bivalent; i.e. represent two chromosomes joined together — for that 

 the dyads have this value is not a theory, but a known fact. It 

 remains to be seen whether these facts apply to other parthenogenetic 

 eggs ; but the single case of Artemia is little short of a demonstration 

 not only of Hacker's and vom Rath's conception of bivalent chromo- 

 somes, but also of the more general hypothesis of the individuality 

 of chromosomes (Chapter VI.). Only on this hypothesis can we 

 explain the persistence of the original number of chromosomes, 

 whether eighty-four or one hundred and sixty-eight, in the later stages. 

 How important a bearing this case has on Strasburger's theory of 

 reduction (p. 196) is obvious. 



H. Summary and Conclusion 



The one fact of maturation that stands out with perfect clearness 

 and certainty amid all the controversies surrounding it is a reductioji 

 in the number of chromosoines in the ttltimate genn-cells to one-half the 

 number eJiaracteristic of the somatic cells. It is equally clear that this 

 reduction is a preparation of the germ-cells for their subsequent union, 

 and a means by which the number of chromosomes is held constant 

 in the species. As soon, however, as we attempt to advance beyond 

 this point we enter upon doubtful ground, which becomes more and 

 more uncertain as we proceed. With a few exceptions the reduction 

 in number first appears in the direct progenitors of the germ-cells by 

 a segmentation of the spireme-thread into one-ha^f the ns?tal number of 

 rods. This process is, however, not an actual reduction in the num- 

 ber of cJiromosomcs, but only a preliminary " pseudo-reduction " in 

 the number of chromatin-///c?'i'.s-rjr. In what we may regard as the 

 typical case {e.g. Ascaris) the pseudo-reduction first appears at the 

 penultimate division ; i.e. in the grandmother-cell of the germ-cell 

 (primary oocyte or spermatocyte). It may, how^ever, appear at a very 

 much earlier period, even in the embryonic germ-cells, the reduced 

 number appearing in every succeeding division until the germ-cells 

 are formed. This is the case in the salamander and in Cyclops. It 

 appears in its most striking form in the higher plants, where the re- 



