It has been reported that the regression of human sex ratio (number of
males per 100 females at birth) on cycle day of insemination is U-sha
ped. The point is not yet established, so a search was made for possib
le effects of this phenomenon on the regression of sex ratio on durati
on of gestation (from last menstrual period to delivery). A previously
unremarked feature of this latter regression is that it, too, is U-sh
aped. It is accordingly suggested that the one regression is responsib
le for the other, and a simple mathematical model indicates that the m
agnitudes of the two regressions are compatible with this conclusion.