No. 485] 



THE DETERMINATION OF SEX 



319 



cate twins. It is not probable, however, that these errors are in 

 such a direction as would change the sexual proportion if we 

 could get absolutely correct statistics. 



The sexual ratio is remarkably constant in widely different 

 localities and at widely different times; in a given locality the ratio 

 is not altered by the varying social and material conditions of the 

 parents, as indicated by statistics. 



The study of duplicate twins shows that if sex is determined 

 by a series of accidental causes, such causes cannot be operative 

 after the fertilization and first segmentation of the ovum. 



The logical conclusion from the statistical data is that sex is 

 hereditary. Mendel's law does not apply. The constancy of the 

 sexual ratio for more than two hundred years may best be ex- 

 plained by supposing that sex follows Galton's law of ancestral 

 inheritance. 



If sex is hereditary, we may explain the significance of the sexual 

 ratio on the basis of natural selection by supposing that the pro- 

 portion of the sexes in any species is such as will give that species 

 the maximum reproductive power at the time of sexual maturity 

 of its individual members. The sexual ratio may be expected to 

 vary for different species, depending upon the mating and breed- 

 ing habits of any particular species. The sexual proportion may 

 be considered as one of the physiological adaptations of a species. 



The conclusions drawn from statistical data should be tested 

 experimentally. There are two experimental points of attack: 

 (1) Breeding experiments to determine whether the sexual pro- 

 portion can be altered by selection. (2) Experiments on the 

 separate blastomeres from one ovum to determine whether two 

 embryos of different sexes can be reared from the same egg, and 

 whether there is a period of sexual indifference in the development 

 of an individual. 



