34 THE AMERICAN JOURNAL OF SOCIOLOGY
is so far finished that the organism produces a surplus of nutri- tion to be used in reproduction. Organically reproduction is also a function of nutrition, and, as Spencer pointed out, is to be regarded as discontinuous growth. The fact that an anabolic surplus, preparatory to the katabolic process of reproduction, is stored at an earlier period in the female than in the male, and that this period is retarded in the ill-nourished female, is a confirmation of the view that femaleness is an expression of the tendency to store nutriment, and explains also the infantile somatic characters of woman. Finally, the fact that polyan- dry is found almost exclusively in poor countries, coupled with the fact that ethnologists almost uniformly report a scarcity of women in those countries, permits us to attribute polyandry to a scarcity of women and scarcity of women to poor food conditions.
This evidence should be considered in connection with the experiments of Yung on tadpoles, of Siebold on wasps, and of Klebs on the modification of male and female organs in plants :
According to Yung, tadpoles pass through an hermaphroditic stage, in
common, according to other authorities, with most animals When the
tadpoles were left to themselves the females were rather in the majority. In three lots the proportion of females to males was, 54-46; 61-39; 56-44. The average number of females was thus about fifty-seven in the hundred. In the first brood, by feeding one set with beef Yung raised the percentage of females from 54 to 78 ; in the second, with fish, the percentage rose from 6 1 to 8 1 ; while in the third set, when the especially nutritious flesh of frogs was supplied, the percentage rose from 56 to 92. That is to say, in the last case the result of high feeding was that there were 92 females, and 8 males. 1
Similarly, the experiments of Siebold on wasps show "that the percentage of females increases from spring to August, and then diminishes. We may conclude without scruple that the production of females from fertilized ova increases with the temperature and food supply, and decreases as these diminish.'
Nor are there many facts more significant than the simple and well-known one that within the first eight days of larval life the addition of food will determine the striking and functional differences between worker and queen. 3
1 GEDDES and THOMSON, loc. cit., bk. i, chap. 4.
a ROLPH, quoted by GEDDES and THOMSON, loc. cit., bk. i, chap. 4.
- GEDDES and THOMSON, ibid.
