TRIPLET CALVES. 263 



tistics for man, that here the number of sons for a thousand 

 daughters drops from 1057 ^'^^' single births to 548 for quadruple 

 births. That this decHne is progressive is shown b}^ the table 

 given by this author (Joe. cit. p. 298) which is here reproduced. 



Number of Sons for 

 1000 Davighters. 



Single births - : 1057 



Twin births , 1043 



Triple births 1007 



Quadruple births 548 



Such a preponderance of females in multiple births is not 

 confined to the human species. The same thing is true of mul- 

 tiple births in sheep and other usually uniparous animals. The 

 obvious explanation which suggests itself to account for these 

 deviating ratios is a prenatal mortality which is- differential in 

 respect to sex. It is well known, from statistical studies of 

 the matter in man, that the prenatal rate of mortality is greater 

 for males than for females. If, as seems reasonable a priori, 

 the conditions of foetal existence become more severe as the 

 number of young borne in the uterus at the same time increases, 

 any inherent differences in respect to foetal mortality rate be- 

 tween males and females might be expected to be accentuated 

 by these conditions. The figures quoted from Nichols show 

 exactly the trend Avhich would be expected if there were an 

 increasing proportion of deaths of male embryos with multiple 

 gestations of advancing order. There are, however, certain 

 facts which make it very doubtful indeed whether any sig- 

 nificant portion of the observed disturbances in the sex ratio 

 in multiple gestations are due to differential prenatal mortality. 

 In the first place the same change in the sex-ratio is observed 

 in multiple gestations in domestic animals where the prenatal 

 mortality is insignificant in amount. In the second place the 

 known facts regarding the distribution of the sexes in multiple 

 births do not accord: with the sex-differential prenatal mortality 

 hypothesis. 



The preponderance of females is not the only disturbance of 

 the normal single birth sex ratios which occurs in multiple 

 gestations. Another, and probably even more significant devia- 

 tion is that which concerns the distribution of individuals of 

 the two sexes in the sets of offspring born together. If indi- 



