SEX-DETERMINATION IN AMPHIBIANS 333 
fully ten minutes before there was any indication of a division in 
the eggs of the control lot for the series. When the embryos were 
removed from the moist chamber and placed in water, seventy- 
seven hours after the experiment was begun, all of the jelly that 
had surrounded the eggs had disappeared and the embryos were 
lying on the nearly dry filter paper from which it took some time 
to float them off. Out of about 450 embryos that were taken from 
the moist chamber, 400 were selected for rearing. The tadpoles 
in this lot were noticeably larger than were any other tadpoles 
in the series, and they began metamorphosing less than five weeks 
after the experiment was started. The mortality during the 
development of the tadpoles was remarkably low, only 4.75 
per cent, so that selective mortality could have had very little 
influence on the proportion of the sexes in the lot of individuals 
carried through to metamorphosis. In the 381 individuals in 
which sex was ascertained 275, or 72.33 per cent, were females. 
This percentage of females is nearly 25 points higher than that 
in the control for the series (table 1), and much too high to be 
considered as a chance variation in the normal proportion of the 
sexes. 
In this, as in the other series of investigations made last spring, 
the experiment was repeated with the eggs from a different female 
in order to avoid the possibility of drawing conclusions from an 
unusual sex ratio that might be merely a chance variation. As 
a check for the experiment made with the eggs from female a, 
a lot of about 400 eggs, taken from female }, was fertilized out of 
water and kept in a moist chamber for fifty hours. In this in- 
stance, also, practically all of the eggs were fertilized, and the 
development of the tadpoles was similar in every respect to that 
of the tadpoles belonging to the corresponding experiment in series 
A. The mortality among these tadpoles also was very slight 
(6.50 per cent), and 374 individuals were carried through to meta- 
morphosis and their sex ascertained. This lot of individuals, 
as shown in table 5, contained 77.27 per cent of females, which is 
30.87 points above that in the control for the series. The sex 
ratio in this instance, 29.41 males to 100 females, falls far below 
that in any lot of toads so far examined. 
