784 
BULLETIN OF THE BUREAU OF FISHERIES 
During the spring of 1934, 26,000 coho fingerlings, averaging 47.4 m m in length, 
which were made available through the cooperation of the Washington State De- 
partment of Fisheries, were marked by the author at Friday Creek, a tributary of 
the Samish River. During the same month, 9,800 coho fingerlings, averaging 49.2 
mm in length, were transferred from the Skykomish River and were marked and 
liberated in Friday Creek. In November of that year an additional 26,000 finger- 
lings from the same brood as the fish used in the first experiment were also marked 
and liberated at Friday Creek. This lot averaged 101.6 mm in length at the time 
of marking. Complete data on returns to the Samish River of six grilse from the 
third marking experiment were obtained during the spawning run of 1935, and the 
capture of two additional marked grilse was reported from a reliable source. 
The run of normally maturing three-year-olds appeared during the winter of 
1936-37, and 480 marked fish were recovered from the Samish River, 7 from the 
first experiment, 11 from the second, and 462 from the third. No recoveries have 
been made from nearby streams or from the Skykomish River. From these results 
it would appear that mortality is much higher for the smaller fish, and that there is 
a definite tendency for mature cohos to return to spawn in the stream from which 
they migrated to the sea. 
LOCALITY OF CAPTURE BY DIFFERENT TYPES OF GEAR 
CATCHES IN VARIOUS DISTRICTS 
Cohos have been second in demand only to kings for consumption as fresh fish, 
and large quantities have always been used in local markets. Because of their suit- 
ability for freezing they have surpassed all other species as a supply for the consider- 
able demand of cold-storage units which have maintained an active market since the 
earliest years of the present century. For these reasons the canned-pack figures for 
this species are an unreliable measure of the commercial catch in past years. Al- 
though they have been the mainstay of the cape purse-seine fishery throughout its 
history, Gilbert (1913) reporting over 850,000 cohos taken there as early as 1911, 
and have formed the major part of the offshore catch of trailers, no records of the 
high-seas catches have been kept for other than very recent years. 
It is impossible without thorough tagging experiments to determine the propor- 
tion of the cape catch provided by the populations of the Puget Sound-Fraser River 
region. Because of their widespread range of operation, part of the trailer’s catch 
landed in Washington may well be drawn from other sources. The purse seiners, 
however, are usually concentrated in the area off the entrance to the Strait of Juan 
de Fuca, and their catch doubtless consists mainly of the populations from the region. 
We may infer, from the far greater size of the runs entering the Strait of Juan de 
Fuca than of those conceivably passing the Banks en route for any other nearby 
district, that the major portion of the catch there is drawn from the regional popula- 
tions. 
In Puget Sound the trap fishery usually suspended operations in early years 
before the coho run had begun, except in the inside waters where the catch consisted 
