738 
BULLETIN OF THE BUREAU OF FISHERIES 
of the runs. Catches of coho salmon used were those taken during a period between 
September 16 and October 27; this period was decreased by one week in both 1929 
and 1930. The period used for chum salmon was from October 13 to November 6, 
except in years when the season was extended beyond the latter date. The periods 
for sockeye salmon were necessarily more varied than those for other species because 
of greater fluctuations in the time of run. Catches used were generally from the 
period between July 15 and August 15, although these dates were shifted when 
necessary, for example, to the period from July 29-September 8, in 1930, when the 
run was very late. 
For each species the average delivery by vessels of each size class was deter- 
mined by dividing the total number of fish caught, during the period selected, by 
the total number of deliveries made. No class was used in which less than 5 vessels 
fished with a minimum of 10 catches. For years in which the 6 size classes between 
10 and 39 net tons were represented, the average catches of the individual classes 
were determined as percentages of the average catch of all classes. For early years, 
when data were available for only the smaller classes, the average catches were 
determined as percentages of the average catch of the total class range represented. 
In order to make the data for early years comparable with those for later ones, the 
percentages of the individual size classes were proportionately reduced so that their 
sum was equal to the average sum of the percentages of an equal class range in the 
years when all classes were represented. The sums of the percentages in all years 
were divided by the number of years to determine the average percentage for each 
class, and the ratio of these averages to that of the class from 10 to 14 net tons was 
calculated for each species. 
These relative-efficiency ratios for each species, and for the average of all species, 
are presented in table 15. The sockeye salmon show the smallest and least con- 
sistent differences between vessel classes. Large catches of this species have fre- 
quently been made by vessels of all sizes fishing in certain limited areas on the Salmon 
Banks, near Lummi Island and at Point Roberts (fig. 2). Here peculiarities of winds 
and tides, or advantages of geographical location in relation to migration channels, 
have caused dense schooling for brief periods of time, and disproportional catches 
have been made by many vessels. 
Table 15. — Relative efficiency of Puget Sound purse-seine vessels 1 
Vessel size in five-ton classes 1 
Species 
5-9 
10-14 
15-19 
20-24 
25-29 
30-34 
35-39 
40 and 
larger 
Sorkeve 3 _ 
0. 66 
1.00 
0.99 
1.46 
1. 56 
1. 43 
1.55 
1. 59 
Pink i 
.92 
1.00 
1.27 
1.64 
1.85 
2. 02 
2. 33 
2. 25 
Coho_ 
.83 
1. 00 
1. 15 
1.69 
2. 19 
2. 27 
2. 37 
2. 12 
Chum 
.79 
1. 00 
1. 21 
1.43 
1.70 
1.78 
1.91 
1.98 
All species 
.80 
1. 00 
1. 16 
1. 56 
1.82 
1.88 
2. 08 
1. 99 
1 Proportion of the average annual catch of each species taken by each size class, calculated on basis of 10-14 class as unity. 
2 Size in net tons, official register. 
2 For even years only. 
• For odd years only. 
