REPORT OF THE COMMISSIONER OF FISHERIES. 31 



gravel in an old hatching trough held in the creek were found to be 

 a total loss, the eggs dying apparently on reaching the eyed stage. 



The collection of sockeye-salmon eggs at the Yes Bay (Alaska) 

 station extended from August 29 to September 27, 51,000,000 being 

 taken. Protracted rains after the middle of September caused very 

 high-water stages, which interfered to a considerable extent with 

 seining operations. The first eggs taken showed the eye spots by 

 September 27, and by November 1 all of them were eyed. The fry 

 were held on trays in the hatching troughs until the sac was absorbed 

 and they had commenced to come to the surface in search of food. A 

 branch of the creek in front of the hatchery approximately 2,000 feet 

 long was closed with a rack and stocked with fry, and in this pro- 

 tected area where the fry were fed, the edges of the grass and numer- 

 ous little indentations along the bank were literally swarming with 

 young fish. It is believed that this method of holding the fish more 

 'nearly approximates natural conditions and that the resulting 

 fingerlings may be safely liberated in the body of the lake in the 

 month of June, at which time but few of their natural enemies are 

 present. The fry held in the hatchery Avere fed from June 1 to 29, 

 when the food supply of salted salmon was exhausted and it became 

 necessary to plant the entire stock. Two million fry were held in 

 McDonald Slough, an arm of the lake located about 4 miles from salt 

 water. They were put into the lake on May 11, and by July 25 those 

 observed around the shores were 2 inches long. On September 10 

 the screen was removed and the fish were permitted to migrate into 

 the main body of the lake. An experiment was tried of planting eggs 

 within four or five days of hatching in the sand and gravel around 

 the shores of several neighboring lakes. It is believed that such 

 bodies of water as appear to have sufficient natural food may be 

 stocked by this method and that the young will escape to the sea 

 during high-water periods without injury. The usual run of hump- 

 back salmon entered the lake early in July, and approximately 

 246,000 eggs of that species were taken. 



Operations were conducted at Baker Lake and at six of its 

 auxiliaries in Washington. Five of these stations were open during 

 the entire year, and all species of Pacific coast salmon and the steel- 

 head were handled, the total egg collections of the group amounting 

 to 47,693,000. Although there was a slight decrease in egg collec- 

 tions at some of the stations, the work as a whole exceeded that of 

 the past year. At Baker Lake the work of constructing buildings 

 to replace those destroyed by fire several years ago was completed, 

 and a new trap was installed at the outlet of the lake. This trap is 

 located some distance below its predecessor and apparently functions 

 better. The new hatchery has a capacity for 30,000,000 eggs and 

 25,000,000 fry, when carried in the stacked tray system. The usual 

 fish-cultural operations and repairs were conducted at all of the 

 substations in Washington. Humpback salmon began ascending 

 Duckabush River in the vicinity of the bureau's station on Septem- 

 ber 2, and the run lasted an entire month. It is believed that fully 

 75 per cent of this run escaped the traps and spaAvned naturally. 

 Eggs to the number of 874,000 were collected, and the resulting fish 

 were returned to the river in the advanced fry stage. 



At Quinault Lake (Wash.) station, two concrete rearing ponds of 

 the long narrow type with sloping sides were constructed. They 



