20 



BULLETIN OF THE BUREAU OF FISHERIES 



humpbacks, and trout, as well as young fish. The cause is unknown, unless it was due to over- 

 crowding of humpbacks, with a possible fall of the water level in the river. Mr. Wood states that 

 a few days later the river was still packed with live fish. It is estimated there were over 4,000,000 

 humpbacks passed through the weir this season. 



Three men made a trip to Karluk Lake from September 16 to 24, inclusive. No unusual num- 

 ber of dead salmon containing spawn were seen above Larsen Bay portage, except in two creeks 

 in the lake, as noted later. 



Red salmon in considerable numbers had spawned and were spawning in the river from one- 

 half to three-quarters of a mile below the lake, beginning at a point where the. river makes a 

 right-angled turn around the shoulder of a mountain. 



It is believed, after viewing the river, lake, and spawning creeks, that only one-fourth to one- 

 third of the humpbacks reached the lake. In the river the dead spawned-out humpbacks were 

 piled in windrow after windrow, parallel with the stream, for its entire course. 



Each creek is mentioned below in geographical order, starting at the outlet of the lake and going 

 up the west side and returning on the east side. 



Grassy Point Creek. — The first creek, going up the west side of the lake, just beyond Grassy Point. 

 This creek did show unmistakable signs of some catastrophe to adult fish, similar to the lower 

 river. Numbers of dead red and humpback salmon, that had not spawned, were seen in an ad- 

 vanced state of decay, yet not so far advanced as the ones in lower river. A few reds were spawning 

 in the upper part of the stream; and some humpbacks were still alive in lower part, but spawned out. 

 Judging from the distribution of dead fish and eggs adrift, most of the humpbacks stayed in lower 

 part of stream, where is the best spawning gravel, and here is where the dead, unspawned fish were 

 seen. 



Halfway Creek. — Indications of a considerable number of humpbacks previously in this stream. 

 No dead unspawned fish of either species. Several hundred reds spawning. Piles of dead eggs 

 were in the creek. About two-thirds were humpback eggs. Red eggs were observed, that had been 

 killed about the time they were to come out of the tender stage. Several dozen red eggs were 

 noticed containing live fish past the tender stage. Quite a number of trout were in evidence. 

 School of 200 reds was seen through the wall of breakers at the mouth of creek. 



Meadow Creek. — This creek was well populated with spawning reds for about one mile. Judging 

 from the dead fish, the creek was full of humpbacks a short time before. More than a dozen dead 

 reds, full of eggs, were noticed, but no dead humpbacks. The reds were in a better state of preserva- 

 tion than the humpbacks in lower river. One thing noticeable in the piles of dead fish washed 

 against rocks in the stream was that there would be two or three reds against the rock and then 

 seven or eight humpbacks on top of them. These reds were spawned out. 



Cascade Creek. — Very few fish of any kind have used this stream. Some dead humpbacks were 

 seen one-fourth to one-half mile from the lake, and a few reds were spawning. 



Beach at head of lake. — Large numbers of humpbacks have spawned here on top of a good many 

 reds. Some other reds were spawning. 



O' M alley River. — Red salmon were spawning in this stream in larger numbers than ever noticed 

 before by the writer. Believe it would average a pair to each square yard, except in the so-called 

 "pothole," where they had spawned earlier. Two visits in August in previous years disclosed 

 comparatively few fish in this stream. Apparently the red spawning here is at its height in Septem- 

 ber. Thousands of humpbacks had spawned here and some red eggs had been dug out. The reds 

 were now digging out the humpback eggs. Behind every rock and in every eddy piles of hump- 

 back eggs lay. Within 22 steps the writer counted 12 piles that would average 5 gallons to a 

 pile; and behind a small island about 6 feet in diameter there were more than a 50-gallon barrel 

 full of humpback eggs. These eggs were all dead; had been dug out and drifted around before pass- 

 ing the tender stage. A small percentage of red eggs was among them. In fact, more or less red 

 eggs were noticed adrift in every stream where humpbacks had spawned. Upon examining the live 

 red eggs, including those partly white, there were about the following proportions: Fifty per cent 

 were well eyed, 25 per cent were fertile but in tender stage (most of them beginning to show white 

 spots) ; and 25 per cent were not yet to the tender stage. The dead red eggs — that is, eggs that had 

 turned entirely white — were more numerous than the live ones. All of these live eggs will probably 

 be picked up by the birds and trout before they hatch and after the dead eggs decompose. 



