120 
BULLETIN OF THE BUREAU OF FISHERIES 
Rounsefell, 1931, p. 35-36). The herring of these “inside” waters belong to popula- 
tions quite distinct from those of Chatham Strait. 
It is well known that different populations may exhibit different structural 
peculiarities owing to differences in environment or to differences in heredity arising 
during long periods of isolation. The study of the individuality of the populations 
has been based largely on these structural differences. Whether the differences in 
the characters chosen are due to heredity or to environment has not been considered 
as being of great importance, as long as the characters are fairly stable within each 
population so that significant differences indicate very slight intermingling, if any, 
between adjacent stocks of herring. 
Success has finally been achieved for the direct method of tracing migrations 
by the release and recovery of tagged herring (Rounsefell and Dahlgren, 1933). 
This method may be called the direct method of racial investigation in contradis- 
tinction with the indirect method in which the movements or lack of movement of 
a population are inferred from the statistical analysis of morphological characters. 
Owing to the newness of this method which was first attempted in 1932, only a few 
results have been attained. Yet these few results offer such corroboration of our 
racial work as to inspire confidence in our results. 
In the determination of the individuality of populations by indirect methods it 
was deemed advisable, profiting by the experience gained in the preliminary racial 
work (Rounsefell, 1930), to concentrate on vertebral counts. This was the more 
necessary, owing to the difficulty of securing enough samples of fresh herring from 
various localities in the nearly perfect condition necessary for body measurements. 
The rates of growth and relative abundance of year classes have also been employed 
as indicators of populational differences. 
SPAWNING AND FEEDING LOCALITIES 
At the present time there are 3, or perhaps 4, spawning areas in southeastern 
Alaska, where the herring may be counted upon to spawn in abundance each spring. 
(See fig. 1.) Of these, the spawning grounds in Sitka Sound, on the outside of 
Baranof Island are probably the largest. Those at the entrance to Klawak Inlet 
in San Alberto Bay are undoubtedly a close second, and those centering near Juneau 
in Stephens Passage are easily third. The spawning grounds in Kootznahoo Inlet 
were once of great importance but have declined. 
As indicated in figure 1, there are a number of minor spawning grounds, a few of 
which were considered of importance in the past. There are certainly additional 
localities, not noted, where a few herring occasionally spawn. 
It may be of interest to note here the distances between the four major spawning 
grounds. These distances, measured approximately from the centers of spawning, 
are as follows: Sitka to Craig (Klawak Inlet), 120 miles; Sitka to Kootznahoo Inlet, 
70 miles by Peril Strait; Sitka to Kootznahoo Inlet, 140 miles by Cape Ommaney; 
Sitka to Juneau, 150 miles by Icy Strait; Sitka to Juneau, 120 miles by Peril Strait; 
Juneau to Kootznahoo Inlet, 65 miles; Craig to Kootznahoo Inlet, 140 miles; and 
Craig to Juneau, 210 miles. 
It is difficult to theorize as to the significance of the considerable distance between 
any major spawning grounds. It may mean that some of the minor spawning grounds 
are used merely by occasional schools straying from the main body of herring of any 
particular race. On the other hand it may indicate that there are two kinds or types 
