SEX-DETERMINATION IN AMPHIBIANS 399 * 
or less normal manner died during the gastrulation period, so 
the number of tadpoles that could be taken for rearing was very 
small. The sex data obtained in this series are shown in table 3. 
No conclusions can be drawn from these results since there were 
so few individuals in the various lots. The experiments have 
been recorded simply because the sex ratios found agree with those 
obtained in similar experiments made last year. Altogether 
ten different experiments have been made in which various lots of 
eggs have been fertilized in acid solutions, and in each case a very 
low percentage of females has been obtained. Such a consistent 
series of results, in so many different cases, strongly suggests 
that the acid solutions have increased the tendency of the eggs 
to produce males rather than females, presumably by causing them 
to absorb an increased amount of water during the fertilization 
period. In all of these experiments, however, the mortality was 
very great, so it is possible that selective mortality was respon- 
sible for the results; though why acid solutions should invariably 
be more injurious to young females than to young males is not at 
all clear. It will be necessary to repeat these experiments on 
eges that are in a physiological condition to withstand the injuri- 
ous action of acid solutions before any definite conclusions are 
possible regarding the effects of such solutions on the sex ratio 
of Bufo. 
In another experiment eggs were fertilized in water that had 
been distilled in glass, in the hope that the zygote would absorb 
an increased amount of water and thus tend to produce a male 
rather than a female. The eggs, which were taken from female 
a, remained in the distilled water for thirty hours, the water being 
changed three times during this period. Practically all of the 
eggs experimented upon segmented in a normal manner and con- 
tinued their development. Not many of the 400 tadpoles taken 
for rearing died during their early development, and the entire 
loss was only 18.75 per cent. The 345 individuals that were 
carried through to metamorphosis were found to consist of 189 
males and 156, or 45.21 per cent of females. In this instance 
the sex ratio of 121.15 males to 100 females differs so little from 
that in the control for the series (table 1) that evidently the nor- 
mal proportion of the sexes was not appreciably altered by the 
