PRINCE WILLIAM SOUND, COPPER AND BERING RIVER SALMON STATISTICS 191 
Subsequent developments have demonstrated the fallacy of Moser’s conclusions, 
for, instead of being exhausted, the fisheries had hardly been touched, as is fully 
shown in the history of salmon canning on Prince William Sound in later years. 
From 1897 to 1904, the number of canneries taking salmon from the sound was never 
more than 2, and in the next 10 years only 1 was in operation. There was no unusual 
variation in the catch from year to year, and no evidence that the runs were being 
destroyed by intensive fishing. After 1914, however, important changes in the inten- 
sity of fishing began, due to the establishment of other canneries in the district, all 
of which entered this field to exploit the pink salmon fisheries. The character of 
fishing changed from gill nets and beach seines to a preponderant use of purse seines 
and traps. The catch increased proportionately with the increase of canneries and 
fishing appliances until 1920, when 15 canneries, operating 54 beach seines, 63 purse 
seines, 217 gill nets, and 47 traps were taking salmon from Prince William Sound. 
There is no such definite distinction between the salmon catches in different 
sections of Prince William Sound as exists between many of the fishing areas to the 
westward. In various districts that have been treated in Parts I and II the fishery 
draws upon the salmon produced by only one, or, at most, a few streams, and the 
catches made can be referred with considerable accuracy to the streams in which the 
fish originated. This can not satisfactorily be done in such a district as Prince 
William Sound where many of the important fishing operations are conducted in 
regions where fish are merely passing through and from which they disperse widely 
to spawning grounds in all parts of the sound. As will be shown later, similar condi- 
tions exist in southeastern Alaska and the same, even greater, difficulties are encoun- 
tered there in attempting to analyze the statistics. In certain well-defined and 
limited areas in Prince William Sound catches have been reported that unquestionably 
are properly allocated to the area in question, but this does not measure the total draft 
upon the salmon runs native to the area since the fisheries located in the channels 
through which the fish have passed have taken toll of the runs to an unknown extent. 
However, it has seemed best to preserve the data in as great detail as possible in 
spite of their deficiencies, and the table, therefore, gives the data for each definite 
geographic unit from which catches have been consistently reported. 
In addition the sound has been divided into 10 subdivisions, and data are given 
separately by localities for each one, with the final section of the table for each division 
showing the total catch in that particular area. The sound is also divided into two 
parts — eastern and western- — the line of separation extending from Point Freemantle 
on the north to Montague Point on the south. The subdivisions are considered from 
west to east, and a section of the table immediately following the tabulation of 
catches in the six districts of the western part shows the total catch in the western 
part of the sound. Data for the eastern part are presented in the same way, while the 
last division of the table gives the total catch of salmon by species and the number of 
fishing appliances used in Prince William Sound. These statistics are given in Table 1 . 
A considerable part of the catch in many years was simply reported as coming 
from Prince William Sound without reference to any of the bays or inlets. It was 
impossible to allocate these catches to specific waters but a more or less arbitrary 
allocation has been made between the eastern and western parts of the sound. Fur- 
thermore, small catches were made occasionally in known localities which were not 
of sufficient importance to be shown separately; these were put in with the unallo- 
cated catches. Other catches were made at places merely designated as Kni ght 
Island, Montague Island, and the like, without mention of the waters from which 
