792 
BULLETIN OF THE BUREAU OF FISHERIES 
each of these was then determined as a proportion of the average annual catch of the 
base group. The average annual catch of any trap which fished for a lesser period 
of years than did the standard traps was determined as a proportion of the average 
of the total catches of the base group for the same years as those in which that par- 
ticular trap fished. 
The annual catches of the traps were then divided by the proportional weights 
of the same traps, and the average of the resultant figures for any 1 year is the index 
figure for that year. The index figures for the three areas, tabulated in the last three 
columns of table 39, are not directly comparable as they now appear, but measure 
only the degree of change from year to year in the individual areas. The relative 
changes in the different areas will be considered in conjunction with the index de- 
rived from purse-seine catches. 
CALCULATION OF PURSE-SEINE INDEX 
Inasmuch as a considerable portion of the coho salmon taken in Puget Sound 
waters have been caught by means of purse seines, a determination of changes in 
abundance of the species based on purse-seine data provides a valuable comparison 
with the indices from trap catches. 
The purse-seine index is similar to those derived from traps in that it is a measure 
of relative variation, from year to year, in the average catch of a unit of fishing effort. 
However, its construction is materially different in that the total seasonal catches of 
individual vessels are unknown, hence the size of the average delivery was used as 
the unit of measurement instead of the annual catch. 
In order to eliminate the influence of deliveries made by the vessels fishing for 
other species of salmon than coho, only such deliveries as were made between Sep- 
tember 2 and October 20 were included. Data were also limited to vessels of more 
than 9 net tons and less than 40 net tons. This restriction excluded both the very 
small vessels, which were not regular purse seiners, and the largest vessels, which 
fished on Puget Sound only occasionally in the fall. 
Since the average delivery of the small vessels operating in early years could 
not be compared directly to that of the large-sized, modern vessels, the catches neces- 
sarily were weighted to compensate for the changes in efficiency. In determining the 
weighted average delivery of the fleet, the vessels of 10-14 net tons were considered 
as unity, and the weighted number of deliveries of vessels in larger size-classes were 
the product of their actual number of deliveries and the vessel efficiency of that par- 
ticular size-class, taken from table 15. For each year from 1911-34 the sum of the 
number of fish in the catches of all vessels in the fleet was divided by the weighted 
number of deliveries. The weighted average delivery figures represent the average 
catch in terms of one size-class of vessels, hence they are directly comparable through- 
out the series of years. These figures are presented in the last column in table 40. 
The other columns in the table show the same data broken down according to group- 
ings of vessels of various sizes, 
