20 U. S. BUREAU OF FISHERIES. 



At the close of the season the bureau's launch was placed on the ways at Coose Bay 

 and left in care of Mr. Millett. who has a launch of his own at the same point. Sup- 

 plies were stored at Fass Ranch for use early next season, which does away with the 

 necessity of the ])urchase of any supplies except perishables before the arrival of the 

 cannery fleet. After this work was accomplished the party returned to Iliarana 

 Village, thence over the portai^e to the bay and ^ia launch to Anchorage, by rail to 

 Seward, and thence by regular steamer to Seattle, arriving September 29. 



At Kokhonak Falls a fishway was blasted out of the rock through which it was 

 hoped the fish could pass into the river above. A cut was made in the rock about 

 30 inches wdde by 3 feet deep and 10 feet long, with a 20 per cent drop, carrying about 

 1 foot of water, which flowed into a large pocket, then through a series of natural pockets 

 to the foot of the falls. The fish ascend with little difficulty to the large pocket but 

 find trouble in reaching the last pocket at the foot of the cut, where a leap of about 

 4 feet is necessary. As the pocket is comparatively shallow, the fish are unable to 

 get sufficient power for the leap, although some accomplish it. More blasting and 

 possibly some flume work will be necessary to make the passage thoroughly efficient. 

 The location of the fishway is excellent, as it is protected by a rock bluff projection 

 from the shore which precludes damage from ice or drift, and also has the advantage 

 of being the side where salmon are most in evidence endeavoring to ascend. It was 

 estimated that 1,500 red salmon were working at the falls and some were spawning 

 near its base in comparatively deep water on our last visit August 20, about three 

 times as many as the previous year. 



Summary. — ^The lateness of the spring season interfered greatly with the work, as 

 in most cases the migration was nearly over before the ice had left the lakes. At 

 both Iliamna and at Wood River the small migration was a contributing factor, as the 

 attraction was not sufficient for large schooling of trout. The total take for the season 

 was 34,758 fish, totaling 122,762 pounds. 



RUNS OF SALMON. 



While the nvmiber of fieh ascending was more than double that of the preceding 

 year, the run was of such short duration that no large individual packs were possible. 

 It was estimated that about 950,000 cases of salmon were packed during the season, 

 practically all reds. All canneries curtailed their preparations at least 20 per cent 

 of capacity, but their packs fell short of anticipation. 



Salmon in large numbers first made their appearance July 1 in the K\-ichak River 

 and the following day in the Naknek River and vicinity. In the Nushagak sector tlie 

 run began .luly 5. At each point the numbers immediately increased to such propor- 

 tions that the different canneries, with one or two exceptions, set a limit of 1,200 fish 

 per boat, which condition continued over a period of about a week, when the run 

 tapered off rapidly and the limit was removed, but the crest of the nin had passed and 

 few large catches were again reported. An increase was noted July 15, but im- 

 mediately the run fell to small and, in most cases, unprofitable catches. 



ESCAPEMENT. 



While the escapement in the different rivers and lakes was excellent, it is believed 

 to be due in part to the efficiency of the stream patrols. An estimate of the numbers 

 reaching the lakes is as follows: 



Wood River 1,000, 000 



Kvichak River 1, 000, 000 



Naknek River 800, 000 



Egegik River 500, 000 



Ugashik River 500, 000 



The run was inspected in the different rivers and checked as nearly as possible, 

 and while the estimate is known to be faulty, it is believed to be close enough for 

 purposes of future observations. 



Ooservations were almost wholly in regard to red salmon, but other species were 

 noted in the Naknek-Kvichak section in rather small nimibers. In passing up the 

 Naknek River A'isits were made to each tributary stream. There were but two of 

 any importance in connection with the salmon runs^King Salmon Creek and Ralph 

 River. The former, about 12 miles up, is small but has fair possibilities. King 

 salmon were noted ascending in small numbers. Later fair numbers were observed 

 spawning in the main river over a stretch of about 3 miles below the foot of the rapids. 

 Ralph River, a tributary about 7 miles below the rapids, is a fair body of water with 



