36 U. S. BUREAU OF FISHERIES 
tinued under the supervision of J. A. Crag and O. W. Lindgren, 
and was accomplished with the aid of W. P. A. and P. W. A. funds. 
During 1939 the construction of the screen in the Wapato Canal, near 
¥ akima, Wash., the work on the Sunnyside screen, and the construc- 
tion of the screen in the Echo Feed Canal in Oregon, were completed. 
Both of the canals are property of the U.S. Bureau of Reclamation. 
Contract work on the screen at the Black Canyon Dam canal of the 
Bureau of Reclamation, on the Payette River in Idaho, was begun 
in December and all contract and other work on the screen in ‘the 
Prosser Power Canal of the same agency, near Prosser, Wash., was 
almost completed at the end of the year. 
COHO SALMON 
The coho salmon of Puget Sound are not only of importance to 
the commercial fishery of the region, but likewise to the sport fishery 
which forms one of the Pacific Northwest's most valuable recrea- 
tional assets. The study of the life history of these salmon, and the 
causal factors responsible for fluctuations in their abundance, was 
conducted by George B. Kelez until May 1939. Owing to Mr. Kelez’ 
assignment to the study of the Bristol ‘Bay ; salmon, Dr. J. L. Wild- 
ing, “for merly with the Columbia River investigation, took over the 
study of the coho salmon. 
Marking experiments—Final returns from the 1937 marking ex- 
periments at the Bureau’s Quilcene (Wash.) Hatchery were collected 
at the hatchery rack in the fall of 1939. During this period the rack 
was covered twice by flood waters and some of the fish escaped to 
the upper river. Of 148 adult fish taken at the rack, 67 were marked. 
Thirty-eight of the recoveries were from the lot of 24,600 fingerlings 
marked and released in August and September 1937, 25 were from 
the lot of 20,039 fingerlings marked and released in December 1937, 
and 4 were from the lot of 17,094 fish marked and released in May 
1938. Two additional recoveries from the December 1937 release 
were made in salt water near the hatchery. 
The results of these experiments show that the fall release of finger- 
lings produced slightly better returns than the midwinter release. 
eh less expense for handling and feeding, and the spring release 
of fingerlings in their second - year produced much smaller returns 
than did the preceding releases. 
Returns from the belly-tagged fingerlings released at the State of 
Washington’s Samish River Hatchery in 19372 also were recovered 
in the fall of 1939. One adult fish was returned from the October 
release, 15 were recovered from the December release, and 3 recov- 
eries were made of fish which had not retained the tags. Although 
these returns were sufficiently large to warrant further experiments, 
mortality appeared to be considerably higher than that of fish marked 
by fin incision only. The survival of winter-tagged fingerlings was 
much greater than that of fall-tagged fish. 
Young fish, and fish-food studies —A study of the seasonal fluctua- 
tions in numbers of young fish and abundance of the fish-food or- 
ganisms in the various streams of the region frequented by the coho 
salmon was begun. This study was also. planned to include a corre- 
2 See ‘Progress in Biological Inquiries, 1937,” p. 30. 
