224 
BULLETIN OF THE BUREAU OF FISHERIES 
are considered in relation to the total discrepancy. Thus it may be seen that the 
observed discrepancies in the calculated growth of the Silver Lake cisco did not origi- 
nate in a change of the body-scale relationships with increasing age. 
Table 9. — Calculated growth of the Silver Lake cisco, collection of 1931 
Age 
Length 
Uncorrected calculated lengths 
Corrected calculated lengths 
Li 
La 
La 
U 
U 
U 
Lj 
Li 
La 
L 3 
Li 
l 5 
Le 
Lj 
VII.. 
201 
194 
188 
183 
177 
171 
141 
77 
78 
80 
80 
83 
104 
105 
111 
119 
126 
132 
137 
151 
135 
142 
150 
159 
166 
158 
168 
168 
176 
172 
177 
182 
186 
189 
196 
84 
84 
86 
86 
89 
108 
108 
116 
123 
129 
135 
139 
152 
138 
145 
152 
160 
167 
160 
160 
169 
176 
173 
178 
182 
187 
189 
196 
VI 
V 
IV 
Ill 
II.... 
I 
Further light is thrown on the question by the examination of the actual average 
measurements of the growth areas of the scales upon which the Silver Lake growth 
calculations were based (table 8). The comparison of the diameter measurements 
of table 8 with the corresponding calculated lengths of table 9 shows clearly that the 
disagreements in the calculated growths of the different age groups appear also in 
the actual scale measurements upon which the growth calculations were made. 
Van Oosten (1929) also observed that Lee’s phenomenon was to be found in actual 
scale measurements as well as in calculated growths. 
The “apparent change in growth rate” in the Silver Lake cisco must, then, be 
considered a real change rather than an apparent change due to changing body-scale 
relationships. Since, as was pointed out previously, the phenomenon appears con- 
sistently both in comparisons between different age groups of the same year class and 
in comparisons between different age groups of different year classes taken in the 
same calendar year, it cannot be explained either on the basis of inherent differ- 
ences in the capacity for growth in different year classes, or on the basis of varying 
environmental conditions affecting growth in different years. It must rather be 
considered the result of some selectional factor correlated with the length of time 
intervening between the years for which the computations and scale measurements 
were made and the year when the fish was captured. Attention will be called 
briefly to possible sources of selection. 
POSSIBLE CAUSES OF LEE’S PHENOMENON IN THE SILVER LAKE CISCO 
SELECTION BY GEAR 
A more detailed consideration of the question of the selective action of the gear 
used will be presented later in this paper. (See section on selective action of gill 
nets.) It may be stated here, however, that this factor most probably does not 
operate on the Silver Lake ciscoes except in the I group and that therefore it cannot 
explain Lee’s phenomenon in these fish. 
SELECTION DUE TO DISSIMILAR DISTRIBUTION WITHIN THE LAKE OF THE VARIOUS 
ELEMENTS OF THE POPULATION 
Since the gill nets used in collecting the cisco samples for these studies were set 
directly on the bottom they fished only the lower few feet of water. If it is assumed 
that the larger individuals of the population tend to lead a more pelagic existence 
