AGE AND GROWTH OF THE CISCO 
267 
which the estimations are based. For example, the 1927 year class which as the IV 
group composed 5.6 percent of the 1931 collection and as the V group composed 7.3 
percent of the 1932 collection must be considered to represent a much less successful 
year than the 1924 year class which as the VII group of 1931 composed 11.6 percent of 
that year’s collection and a year later as the VIII group made up 7.9 percent of the 
1932 collection. Although the 1924 year class individuals in both years’ collections 
combined were only about one and one-half times as numerous as those of the 1927 
year class, the former year class, by reason of its 3 years greater age, had suffered 
the reducing effect of natural mortality over a longer period of time. Consequently 
the 1924 year class as compared with the 1927 year class may be considered much 
more successful than the comparison of the representation of the two groups in the 
collections would indicate. A similar line of reasoning leads to the conclusion that 
the 1925 year class (VI group of 1931) also represents a very successful year. The 
same conclusion may possibly apply to the 1923 year class which was well rep- 
resented as the VIII group of 1931. 
In the preceding discussion it was shown that in each population there existed 
a considerable variation in the numerical abundance or successfulness of the various 
year classes. The years that appear to have had a production of relatively successful 
year classes were: Trout Lake — 1926, 1927; Muskellunge Lake — 1928; Silver Lake — 
1926; Clear Lake— 1924, 1925, 1929. In Silver Lake the 1929 year class was poor, 
and in Clear Lake the 1926 and 1927 year classes must be considered relatively 
unsuccessful. 
Since the four populations show no agreement in the matter of goodness or 
poorness of the various year classes, it may be concluded that in each population the 
success of the individual year classes depends on the purely local conditions within 
the lake. This conclusion is of particular interest in view of the rather general 
opinion that fluctuations in the relative abundance of year classes have their origin 
in differences in weather conditions from one year to another. Inasmuch as these 
four lakes are grouped closely together (the maximum distance between any two of 
them is about 18 kilometers) they must all be subjected to approximately the same 
weather conditions. 
It must be remembered that relative abundance is not a wholly reliable index of 
absolute abundance. A poor year class may appear quite numerous in comparison 
to a yet poorer one. Jarvi (1920) gave an excellent illustration of this fact when he 
pointed out that the great relative abundance of third year fish in some years’ catches 
of the “kleine Marane” depended not upon the actual abundance of that particular 
age group but was rather the result of the scarcity of fish in their second year of life. 
The data of tables 45 to 48 make possible a comparison of longevity in the four 
populations. The Muskellunge Lake cisco has definitely the shortest average life 
span. The Trout Lake cisco falls second and the Silver Lake cisco third, while the 
average life span of the Clear Lake cisco is quite long. A possible explanation of 
these differences among the four populations as to the average length of life was 
mentioned in the section on condition in the different populations. 
AGE AT MATURITY AND SEX RATIO 
Although the collections of the cisco were made in July, August, and early 
September, several months before the time of spawning in November, the develop- 
ment of the gonads was such as to leave little doubt as to the sexual maturity of the 
