AGE AND GROWTH OF THE CISCO 
269 
With the exception of the irregular data of Silver Lake, the females tend to 
become relatively more abundant as age increases. The tendency toward a progres- 
sively greater preponderance of females finds an extreme condition in Trout Lake 
where male individuals are rare beyond the IV group. The Clear Lake fish differ 
from those of the other three lakes in that the males outnumber the females in the 
first five age groups. In the VI group the numbers are equal while in the age groups 
above 6 years the females are much more numerous than the males. Differences 
among the populations in the rate of change of the sex ratio with increased age are 
reflected in the average sex ratio for each population as based on the combination 
of all age groups. 
It is believed that the data of table 49 are a reliable index of the sex composi- 
tion of the four populations. Cahn (1927) and Van Oosten (1929) pointed out that 
males arrive at the spawning ground earlier than the females. European investi- 
gators of coregonids have observed frequently that not only do males arrive at the 
spawning grounds earlier than females but that they remain there longer. As a 
consequence, if samples are taken from the spawning run the conclusions concerning 
the sex ratio would not only be affected by the time within the spawning period of 
the taking of the sample but also would tend to contain a too high percentage of 
males. The collections used in this investigation are not open to such a criticism. 
All samples were taken during the summer, several months before the spawning 
time, and there is no evidence pointing toward a segregation of the sexes at this 
period of the year. 
The data relative to the sex ratio in the four populations differ from those given 
by Cahn (1927) for the lakes, Mendota and Oconomowoc, in southern Wisconsin 
and by Van Oosten (1929) for the Lake Huron herring. Cahn found 101 males and 
62 females in his collections from Lake Mendota and 429 males as against 315 females 
in his samples from Oconomowoc Lake. He combined the collections from the two 
lakes to obtain a ratio of 100 males to 71.2 females. Cahn attributed the greater 
abundance of males to the relatively higher mortality of the females during the warm 
weather of late summer. His data on sex ratio were not correlated with age. 
Van Oosten (1929) found males and females approximately equally abundant 
(49.5 percent males) in his samples of the Lake Huron herring taken at Bay City, 
Mich. He did, however, find differences in the sex ratio in the different age groups. 
Females were the more numerous in the younger age groups (I and II in the method 
of age designation used in this paper) while the males were the more abundant in 
the older age groups. Van Oosten attributed the shifting sex ratio to an earlier 
attainment of sexual maturity on the part of the females and a consequent tendency 
for them to appear in the commercial catch at an earlier age. 
The change with age in the sex ratio of the four cisco populations of this study 
must be explained on the basis of a differential mortality of the two sexes. Since 
there is no fishery for the cisco in these lakes, there can be no disturbance of the 
natural relationship through the catching of more individuals of the one sex than of 
the other. The reason for the observed differences in differential mortality in the 
four lakes presents a problem difficult of explanation. Attention should be called 
to the fact that with respect to the relative abundance of males in the entire samples 
the lakes fall in the same order that they show with respect to the amount of growth 
in weight, that is the population with the slowest growth (Trout Lake) has relatively 
fewest males. The possible significance of this parallelism will be considered later. 
