AGE AND GROWTH OF THE CISCO 
225 
than the smaller ones, then the more rapidly growing fish within an age group would 
be less subject to capture by nets set on the bottom than the slower growing individuals 
within the same group. Further, the effect of such segregation would be greater as 
age increases. The above-mentioned assumption would serve well to explain decreas- 
ing growth as it has been observed in the Silver Lake cisco, but data for its verification 
arc not available. The chief objection to such an explanation lies in the fact that it is 
the smaller fish not the larger that would normally be expected to lead the more pelagic 
existence. 
SELECTION DUE TO DIFFERENTIAL MORTALITY, CORRELATED WITH GROWTH RATE 
It has been observed repeatedly that poikilothermous organisms tend to grow 
more slowly and reach a greater age in northern latitudes. The phenomenon appears 
to depend on the relation of temperature to the rate of metabolism. It is not wholly 
unreasonable to assume that the individual rate of metabolism might affect the indi- 
vidual growth rate and the individual length of life. Were mortality greater among 
individuals of more rapid growth, those individuals that survived longest would be 
those that had actually grown most slowly. The early growth as calculated from 
these slow growing survivors of the older age groups would naturally be small. Fur- 
ther, the effect on calculated lengths of this selection through differential mortality 
would tend to be increased with greater age, and thus discrepancies of the type 
observed in the Silver Lake data would be explained. 
The relation between individual growth rate and individual length of life has 
been studied experimentally by several investigators. Osborne et al. (1917) found 
that a temporary preliminary stunting delayed maturity and extended the life span 
of rats. Titcomb et al. (1928) and McCay et al. (1931) found that trout that did 
not grow lived longer than those which showed growth on a similar diet. Zabinski 
(1929) by effecting a retardation of growth was able to prolong life in the black 
beetle and the roach. McCay (1933) presented a brief discussion of the general 
problem of the relationship of rate of growth to longevity. 
OTHER POSSIBLE CAUSES 
It is recognized that the three possible causes of Lee’s phenomenon in the Silver 
Lake cisco discussed in the preceding paragraphs are by no means the only possible 
explanations of the observed discrepancies in calculated growth as based on fish of 
different age. These three suggested explanations were emphasized because they 
appear to be the most plausible in the face of the available data. Other explanations 
should be mentioned briefly along with the reason for their rejection. 
1. Growth may be better in some calendar years than in others. Tliis explana- 
tion cannot be accepted as the growth discrepancies were found to occur between 
members of the same year class, captured at different ages. 
2. Portions of the scale may be resorbed after being laid down. The examination 
of the scales offered no evidence for any kind of resorption. 
3. The scale fields may contract after being laid down. The nature of the 
structure of scales makes this explanation totally unacceptable. (See Van Oosten, 
1929.) 
4. More than one annulus per year may be formed. The clarity of scale markings 
on the Silver Lake cisco was superior to that of the Trout Lake and Muskellunge 
Lake ciscoes; yet the measurements of the scales of the last two populations gave 
