SPECIAL CASES OF SEX-INHERITANCE 233 



The influence of food in sex-determination in man has 

 often been exploited. It is an ever recurrring episode 

 in the ephemeral literature of every period. The most 

 noted case is that of Schenk. In his first book he said 

 starvation produced more females ; in his second book 

 he changed his view and supposed that starvation 

 produces more males. 



Perhaps the most fertile source from which this view 

 springs is found in some of the earlier statistical works, 

 especially that of Diising. Diising tried to show that 

 more girls are born in the better-fed classes of the com- 

 munity, in the poorer classes more boys. The effective 

 difference between these two classes is supposedly one of 

 food ! For instance, he states that the birth-rate for 

 the Swedish nobility is 98 boys to 100 girls, while in the 

 Swedish clergy the birth-rate is 108.6 boys to 100 girls. 



Other statistics give exactly opposite results. Pun- 

 nett found for London (1901) more girls born amongst 

 the poor than the rich. So many elements enter into 

 these data that it is doubtful if they have much value 

 even in pointing out causes that affect the sex ratio, and 

 it is quite certain that they throw no light on the 

 causes that determine sex. 



In other mammals where a sex ratio not dissimilar 

 to that in man exists, extensive experiments on feeding 

 have absolutely failed to produce any influence on 

 the ratio. We have, for instance, Cu&iot's experi- 

 ments with rats, and Schultze's experiments with 

 mice. The conditions of feeding and starvation were 

 much more extreme in some cases than is likely to 

 occur ordinarily, yet the sex ratio remained the same. 



Why in the face of this clear evidence do we find 



