AGE AND GROWTH OF THE CISCO 
299 
2-inch mesh nets. In 1928 the average length of 182 ciscoes was 138 millimeters, 
and in 1930 the average length of 490 individuals was 150 millimeters. The smaller 
average size of the fish of the 1928 collection may be accounted for by the use of 
mesh sizes smaller than 1% inches and by the relatively greater abundance of young 
fish in that year’s samples. The relative abundance of the different age groups as 
well as the average lengths of these age groups in the samples of the 2 years may be 
found in table 3. 
Table 71. — Effect of size of mesh of gill nets on the determination of the average lengths of the age groups 
of the Trout Lake cisco, 1931 collection 
[The first 2 rows show the average lengths of 5 age groups as based on samples taken July 22 in lli- and 1^-inch mesh gill nets. 
The third row shows the average of the 2. The fourth row shows the average lengths of the various age groups based on all 
fish taken in l^-inch mesh gill nets in 1931. Sexes combined. Number of specimens in parentheses] 
Size of mesh 
Age group 
Average 
length 
I 
II 
III 
IV 
V 
1Y\ inches __ - 
124 (1) 
134 (22) 
137 (16) 
139 (14) 
143 (20) 
147 (18) 
149 (15) 
155 (2) 
158 (4) 
140 (61) 
145 (58) 
1^2 inches - ___ 
Averages combined 
124 (1) 
132 (1) 
135 (38) 
137 (39) 
141 (34) 
142 (159) 
148 (33) 
148 (246) 
157 (6) 
156 (77) 
Average of all 1931 lH-inch samples 
147 (522) 
A single set of the “new” gear lifted July 22, 1931, provides the only data for 
Trout Lake for the comparison of growth as based on samples taken in nets with 
different size mesh. These data appear in table 71. In the first row appear the aver- 
age lengths for the various age groups as based on fish taken in the 1%-inch mesh net. 
The second row gives the same information for the sample from the 1^-inch mesh net. 
It will be seen that while the average length of all ciscoes taken in the 1 J^-inch mesh 
net was 5 millimeters greater than that of the ciscoes from the 1 %-inch mesh net, the 
differences, between samples of the same age group varied from 2 to 4 millimeters and 
averaged only about 3 millimeters. The third row shows the average lengths of the 
different age groups as based on the combination of the July 22 samples taken in the 1 /- 
and lK-inch mesh nets while the bottom row shows for comparison the average lengths 
of the various age groups as based on all the ciscoes taken in Trout Lake in the 1931 
season in 1 / 2 -inch mesh nets. It may be seen that with the exception of the I group 
the averages based on the l}£-inch mesh net samples for the entire, season differ but 
little from those based on the combined catches from l}i- and lb-inch mesh nets. 
While the necessity for the elimination of the I group from the Trout Lake growth 
data is at once apparent, the problem presented by the II group requires further 
consideration. In spite of the rather close agreement between the average length 
of the II-group fish as determined from samples from nets of l%- and lK-inch mesh, 
the sparse representation of this age group in both the 1930 and 1931 collections 
(table 3) throws suspicion on the reliability of the samples. This suspicion is sup- 
ported by the high calculated lengths at the end of the first and second years of life 
as based on both the 1930 and 1931 II groups. Consequently both of the groups 
were eliminated from the Trout Lake growth data. The II group of 1928, for which 
year no individual net records are available, was, however, retained, first because of 
its abundant representation in that year’s collection, second because of the good 
agreement between its calculated growth and the corresponding calculated growth 
both of older age groups of the same year’s collection and of samples of the same 
(1926) year class taken in the later years, 1930 and 1931, and finally because of the 
