226 
BULLETIN OF THE BUREAU OF FISHERIES 
quite consistent results in the calculation of growth. Accessory annuli were rare in 
the Silver Lake cisco scales. With the elimination of the more questionable scales 
it appears unreasonable to assume that the number of errors in the determination of 
ages was sufficient to account for the observed large discrepancies in the calculated 
growths. Certainly accessory annuli cannot account for the decrease with age in the 
diameter of the first growth field. 
5. The individuals of the population may have segregated themselves on the basis 
of maturity with the result that the smaller, immature members of each age group 
may have been absent from the samples. This explanation lacks plausibility in the 
face of the strong evidence presented later (p. 268) that some if not most fish mature as 
members of the I group, while all II-group fish certainly can be considered mature. 
It should be mentioned further that segregation on basis of maturity would be 
expected to be most pronounced at the spawning time, not in midsummer when the 
cisco is confined in Silver Lake to a narrow stratum of water in the lower part of the 
thermocline and the upper part of the hypolimnion. 
Regardless of the explanation or explanations accepted as to the cause of Lee’s 
phenomenon in the Silver Lake cisco, it appears that the observed discrepancies do 
not affect the validity of the method used in this paper for the calculation of growth 
from scale measurements. Nevertheless, there remains the question as to why the 
phenomenon should be peculiar to the Silver Lake population. 
GENERAL GROWTH CURVES FOR THE TROUT LAKE, MUSKELLUNGE 
LAKE, SILVER LAKE, AND CLEAR LAKE CISCO POPULATIONS 
GROWTH IN LENGTH 
For each population considered here all collections were combined 8 to give a gen- 
eral growth curve (fig. 1). The indicated lengths at the end of the various years of 
life are in general the grand average of all calculated lengths for these years. In the 
Trout Lake and Clear Lake ciscoes, however, the irregularities in the later years, 
that result largely from the small samples, were smoothed in accordance with the 
observed data on the annual calculated growth increments. 
The Silver Lake growth data raise a question as to the value and significance of 
any “general” growth curve in this population. The presentation of a general 
growth curve involves the tacit assumption that this curve can be taken to represent 
the course of growth of an individual that is typical of the population as a whole. 
However, the Silver Lake growth curve was derived from the combination of several 
age groups whose growth histories were fundamentally different. It appears char- 
acteristic of the Silver Lake cisco that individual growth history and individual life 
span are definitely correlated. Consequently the combining of all age groups to 
obtain a general growth curve involves the lumping together of a mass of heterogeneous 
growth material. The typical individual that such a curve is purported to represent 
is probably nonexistent. Nor can the growth curve obtained from a single age group 
be considered homogeneous. A sample of Ill-group fish, for example, can be expected 
to contain fish that would normally have died within the year of collection, and along 
with them others that would have survived to their fifth, sixth, or seventh year, or 
even longer. All these different groups within a single age group would have dif- 
ferent types of growth. 
8 Certain age groups, however, were eliminated as unreliable because of selection by gear. These were: Trout Lake, all I 
groups and the II groups of 1930 and 1931; Muskellunge Lake, all I groups; Silver Lake, all I groups. For Clear Lake no groups 
needed to be eliminated. 
