hardly be considered that the sex proportion of the children between two and five or five and ten was materially modified by immigration. It must be due primarily to the balance of births and deaths.
Various statisticians have called attention to the fact that the proportion of male children among those born in the city is smaller than among those born in the country. The mortality, and especially the infant mortality, are almost always greater in cities than in the country, and this mortality presses more heavily on male children than on female. Hence if fewer male children in proportion are born, and if those born are subjected to greater dangers in their tender years, plausible reasons are suggested for the development of an excess of females in cities and an excess of males in rural districts by natural increase alone. Unfortunately, the correctness of these reasons cannot be demonstrated. I am convinced that these biological forces do in fact coöperate with migration in the production of the results, and that, even if internal migration is decreasing in the United States, as I have sought elsewhere to prove, there is no reason to expect that in consequence the tendency here shown to a slight dissociation of the sexes will come to an end.
Cornell University.
