THE FISHES OF ALASKA. 237 



sticking out. The falls are in ?overal parts, three l)eing of considerable height, and with quieter water 

 between. The upper or main fall is far more serious, and the salmon make many unsuccessful attempts 

 to ascend it. The lip of the fall is very irregular and broken in many places, and at different levels 

 are found pools or corners with less turbulent water where the fish can rest on their way up. 



Just below this fall is a large relatively quiet pool extending the full width of the stream and perhaps 

 100 feet down the stream. This pool was literally packe<l with salmon; they .seemed to be lying nearly 

 as close together as possiljle and there must have been many thousands i:f them. Xearly or quite all of 

 them, unless disturbed, lay with their heads upstream. They were all restless and seemed to be moving 

 about more or less, usually trying to get nearer the fall — some pushing upward to the edge of rough water, 

 then jumiiing wildly, .sometimes gaining a little, but more often hitting the liank or some projecting rock 

 or swift water and gaining nothing. Immediately under the fall they were jumping all the time. During 

 the hour of observation there was scarcely a moment when one or more salmon could not bo seen in the 

 air or making their way against the nearly vertical ciuTent. The vantage ground from which they 

 started was not good and the jumping appeared to be aimless and at random. Often they would jump 

 straight tip when some distance below the fall, frequently the wrong way, sometimes even downstream; 

 sometimes striking the bank or a projecting rock, to be thrown Isack into the water, only to tiy the leap 

 again. Occasionally a salmon which had reached the foot of a descending sheet of water would, with 

 better judgment, jump toward the sheet and perhaps strike part of the way up, where it would maintain 

 itself for a moment, perhaps advancing a short distance, but finally losing, to be caijied or dashed, some- 

 times sidewise, sometimes headforemost, back into the pool below. During all the time of observation 

 by the authors not a single salmon was seen to make the fall, but others of the party saw some succeed, 

 and in the stream above the falls were seen several salmon that of course had gotten over. Into one pool 

 easy of access, at the foot of the fall and containing as many salmon as it could possibly hold, one of 

 us put his hand to lift a salmon out. By closing the hand over the gills it was possiljle to hold on to the 

 fish without any difficulty, and so long as this succeeded the other fish took no alarm, but when one 

 escaped it and all the others sciuTied down the falls in the most reckless way. 



The entire stream from the falls to its mouth was well filled with salmon. Xearly all appeared to 

 be humpbacks, although there were evidently a few sockeyes among them. Xeai'ly all of l»th species 

 were fresh from the sea and apparently in excellent condition. 



In many places humpback salmon were seen jumping, in ba\s, passages and river mouths. At 

 Sitka (August 20) boys were seen gaffing them in Indian River near its mouth. The males were greatly 

 humped. All reports were to the effect that the nm in the streams near Sitka was unusually large in 

 1903. The same was true at Killisnoo. AVe often saw humpback salmon along the shore where the only 

 fresh water was seepage throtigh the sand and gravel. They would often throw themselves out on the 

 beach in their efforts to reach fresh water. Humpbacks are said to occiu' in the Yukon in July and 

 August, though in limited numbers. We seined large numbers on July 25 at Pablof Bay, southeast 

 Alaska; also at Sitkoh Bay, July 2(5. The species was noted also in Silver Creek near Sitka. July 29. 

 Fiy 1.3 inches long were taken May 22 with a lot of coho fry in sloughs or little pools along the Karluk 

 River near its source. 



The run of humpbacks extends, perhaps, through a longer period than that of any other species. In 

 Southeast Alaska it begins in June and continues until September or even later in some places. Noith- 

 ward the period is somewhat shorter. On Puget Sound and southward it is more prolonged and 

 continues late in the fall. As already stated, the humpback salmon as a rule do not ascend streams to 

 great distances. If they enter large rivers at all they are apt to run into the first small tributary stream 

 which they reach. They prefer, however, the smaller coastal streams, and are therefore not often seen 

 in numbers at any distance from the sea. 



They are quite persistent, and. in a measm'e, successful in theh efforts to ascend streams in which 

 considerable falls occur. Observations similar to those on their attempts to ascend Afognak Falls were 

 made at Pablof Falls July 25, and at Dorr Falls, Xaha Stream, August 30. and indicate that the hump- 

 back jumps quite as well as any of the other species of salmon. Where the water conditions — depth, 

 current, width, etc. — are favorable it can probably make a vertical jump of 10 feet, but to ascend a fall 

 with that vertical height frequent attempts would probably be made before success was attained, as the 

 jumping appears to be more or less at random. It is perfectly evident that the salmon does not select a 

 particular point on or near the lip of the falls where it proposes to strike; it simply jumps aimlessly, 

 and sooner or later strikes the fall at a place where it is able to maintain itself and from which it can 

 ascend into more quiet water aV)i>ve the falls. 



