284 
BULLETIN OF THE BUREAU OF FISHERIES 
Examination of the data shows that during the second year of life of the Ill-group 
fish there occurred sufficient growth compensation to change the length distribution of 
the group from the bimodal to the unimodal condition. The existence of intense com- 
pensatory growth is indicated in the bimodal character of the curve for the second-year 
growth increments of the III group. 
A more adequate conception is given of the growth compensation in these two age 
groups by the study of the correlations between the lengths attained at different times 
in the individual life histories, and the amounts of growth made by the individuals 
during different years of life. The correlation was computed for each age group for the 
following combinations of characters: (1) Calculated length at the end of the first year 
of life and calculated length at the end of the second year of life; (2) calculated growth 
in length during the first year of life and calculated increment of growth in length dur- 
ing the second year of life; (3) calculated length at the end of the first year of life and 
actual length at the time of capture. The results of these computations are given in 
table 63. Here L; is the calculated length at the end of the first year ; L 2 is the calculated 
length at the end of the second year; ALi is the calculated growth during the second 
year; and L T is the actual measured length at the time of capture. 
Table 63.—Muskellunge Lake cisco, II and III groups of 1931 — Correlations between calculated 
length at the end of the first year and calculated length at the end of the second year, between the 
amount of growth during the first year and during the second year, and between the calculated length 
at the end of the first year and total length at the time of capture 
Data upon which the correlation is based 
Coefficient of correlation and 
its probable error 
II 
III 
Li and L2 
0. 513±. 046 
—0. 311±- 056 
. 404±. 052 
0. 708±. 027 
-0. 826±. 017 
. 399±. 045 
Tables 64 to 67 show the data from which the correlations between Lj and L 2 
and between Li and ALi were calculated. The examination of tables 64 and 65 shows 
clearly that in both the II group and the III group the smaller fish at the end of the 
first year of life tend to be the smaller fish at the end of the second year of life, and that 
this tendency is the greater in the III group. Tables 66 and 67 show that in both age 
groups the individuals that grew least in the first year of fife tend to grow most in the 
second, and that here again the tendency is more pronounced in the III group than in 
the II group. 
Table 64. — ■ Relationship between calculated length in millimeters at the end of the first and second years 
of life in the Muskellunge Lake cisco, II group of 1931 
