SEX DETERMINATION 23I 



going indefinitely by vegetative division, but require occasional periods 

 of sexual conjugation. This fact, however, has failed to stand experi- 

 mental testing;^ for instance, Woodruff^ has kept the Blepharisma undu- 

 lans for fifteen years without conjugation, while even the cells of higher 

 organisms can grow and multiply indefinitely under certain conditions, 

 as is shown by Carrel's famous strain of chick fibroblasts, now over 

 twenty years old. Mortimer^ has shown that parthenogenetic Cladocera 

 such as Daphnia can be kept indefinitely without sexual imion if the 

 environmental conditions are correct. 



Weissmann suggested that the importance of sexuality is that it leads 

 to the crossing of individuals and the production of new varieties. The 

 development of genetics has demonstrated the correctness of this view 

 and allows us to put it in more precise form. Crossing does not lead to 

 the production of new hereditary units, but it does enable the already- 

 formed units to be combined in new ways. There must certainly be an 

 enormous evolutionary advantage in this;* for one thing a vegetatively 

 reproducing species can only acquire two new characters when the two 

 mutations happen one after the other in the same clone, while in a 

 sexually reproducing population favourable mutations are spread as it 

 were through one another without hindrance. 



It is clear, then, that sexuality, in the sense of conjugation of two 

 haploid nuclei, is of great evolutionary advantage and will be preserved 

 whenever it occurs. Moreover, as Darlington has pointed out, the 

 mechanism which has actually been evolved is just sufficiently modified 

 from mitosis to give recombination by crossing-over as well as by 

 random assortment of chromosomes ; if the prophase of meiosis were 

 slightly more precocious than it is, the splitting of the chromosomes 

 which now occurs in pachytene would be postponed till the division 

 was finished, there would be no formation of diplotene loops, no 

 chiasmata, and no crossing-over. Thus in its details as well as in its 

 broad outline, the sexual process is admirably adapted to give recom- 

 bination of characters. 



Recombination is only important in hybrids. If self-fertilization were 

 the rule, and most organisms were therefore homozygous, the advan- 

 tages of the sexual cycle would be lost. Clearly then the function of the 

 differentiation between the sexes is to discourage self-fertilization and 

 encourage crossing. The evolution of hermaphroditism, which some- 

 times goes so far as to allow self-fertiHzation, has been discussed by 



^ Rev. Robertson 1929. 2 Woodruff 1935. 



3 Mortimer 1935. * Cf. Wright 193 1, MuUer 1932a. 



