14 
BULLETIN OF THE BUREAU OF FISHERIES 
are soon dislodged and the eggs must be laid and fertilized before this happens, we 
may infer that spawning occurs in these localities. 
There remains to consider the distribution of the sexes and the stages of their 
maturity. As the simplest we shall first consider the distribution of the stages of 
sexual maturity. If the numbers of immature and maturing males or females from 
different localities are compared, the greater proportion of maturing individuals will 
be found in outside waters. (See fig. 8.) This is more marked with the later stages 
of maturity. Thus, in Georgia, in 1932, an intermediate stage of ovary development 
was over twice as common in outside waters as in the creeks, while the last stage of 
ovary development recognized was 
six and one half times as common. 
As will be recalled, the spermato- 
phore-bearing females, which rep- 
resent the last stage of maturity, 
were all found in the Gulf or ocean 
with a single exception taken in an 
exposed sound. A satisfactory 
analysis of these data will require 
the inclusion of figures from 
another breeding season and must 
be presented at a later date. The 
present data can be interpreted in 
but one way — that by far the 
greater proportion of spawning 
takes place in the Gulf or ocean 
and that only a slight amount, if 
any, occurs on the inside. The 
spawning seems to be correlated 
with high salinity. 
SEX RATIO 
Figure 8.— Percent of maturing females in the inside and outside waters 
of Georgia from March to July 1931. 
A comparison of the propor- 
tion of the two sexes in the catch 
at different times of year and in 
different localities shows fluctuations associated with spawning. Since additional 
data will be required for a satisfactory analysis of these fluctuations, only brief 
mention of the facts will here be made. As may be seen from figure 9, in Georgia 
from February 1931 to July 1932 there is a well-marked annual cycle. During the 
winter (from the last half of September 1931 to the last half of April 1932, inclu- 
sive) the proportion is constant, the females constituting slightly over 52 percent of 
the 25,601 shrimp measured. During the similar part of the preceding year the 
females furnished essentially the same proportion of the 11,283 shrimp. In contrast, 
the period from the last half of May to the first half of September 1931, inclusive, and 
again from the first half of May to the first half of August 1932 (where, at this time, 
our record ends) there are sudden and remarkable fluctuations of the proportion, the 
females rising above 80 percent and falling nearly to 30. 
This period covers the time during which the adults mature, spawn, and disappear. 
A few weeks after spawning begins and when about half the females are maturing the 
