HERRING IN PRINCE WILLIAM SOUND, ALASKA 
269 
Table 3. — Catch in various localities in Prince William Sound 
[In thousands of pounds] 
In making a detailed study of the fluctuations in abundance, it is of great advan- 
tage to know whether the catch is being drawn from one or from several populations 
since the commercial catch may not be drawing proportionately on each population 
and each population may not be securing proportional annual increments. Owing 
to the difficulty of securing sufficient numbers of accurate counts of the fin rays and 
gill rakers, and to the great variability in the body proportions (which is especially 
noticeable in purse-seined material), the analysis has been based wholly on vertebral 
counts. 
Before comparing the vertebral counts from the various localities it is interesting 
to know what the causes are for variability in this character, whether they are genetic 
or environmental. For this purpose the counts from all localities in Prince William 
Sound for each year class (fish spawned the same year) from 1919 to 1927, inclusive, 
were treated as a single distribution and the mean computed. These means (Table 5) 
were correlated with the average air temperatures for March, April, May, and June 
from Seward, Cordova, and Latouche. (Table 4.) 
Table 4 .—Mean annual air temperatures of the combined months of March, April, May, and June 
from Seward, Cordova, and Latouche 
i 5.40° have been subtracted from the U. S. Coast and Qeodetic Survey temperatures to make them comparable to Weather 
Bureau data. This allowance has been made for the difference in the time of day at which the temperatures were taken. The cor- 
rection was empirically determined by taking the average difference between the two series for 1926, 1927, and 1928. 
92966—32 2 
