54 FISHERIES OF ALASKA, 1908. 



water in these shallows would change its temperature much more 

 rapidly with variations of the air temperature than the main body 

 of the lake. Possibly the variable outflow of these shallows down 

 the lake accounted for the occasional sudden changes. 



Coloration of redjish. — During the early part of the season and the 

 height of the run at the lake, salmon were colored like those taken 

 by the fishermen for canning — silvery sides and dark-bluish backs ; 

 the color of the back was by no means uniform. An excellent op- 

 portunity was afforded to observe their appearance under water as 

 the fish filed through the gate; especially when the floating glass 

 window, made for the purpose, was used, could a clear and uncon- 

 fused picture be seen. The ground color of the back of the red 

 salmon varied from an almost olive green to darker shades of green 

 and blue almost to black. 



Straggling individuals with the distinct red coloration of the spawn- 

 ing season appeared even early in the season, the first recorded in the 

 notes being a male on June 30. At this time the 'trap contained also 

 several with considerable pinkish color. As the run began to dwindle 

 the number showing spawning colors increased. When the tally 

 ended more than half the few fish passing had turned red. Sixty 

 per cent of all tallied on August 8, and 85 per cent of all on August 

 9, had taken on the typical red color. 



Dead and injured jlsh. — The web caught all dead fish brought down 

 by the current, and these were from time to time thrown over below 

 the rack in order to relieve the pressure on it. Considerable interest 

 attaches to these salmon, although their death is probably an annual 

 occurrence, made more noticeable by the presence of the rack to 

 intercept them. They were not numerous until the latter part of 

 the season, only one having been found up to July 7. Between Juh^ 

 24 and August 8 more than 1,200 dead were thus picked off the up- 

 stream side of the web and thrown below it. They were chiefly red 

 salmon, with some trout and humpbacks, and probably a very few 

 of the other species of salmon. Many of them drifted against the 

 web while still alive, but too weak to stem the current, and in this 

 condition were drowned. Most of these fish had been injured in one 

 way or another before their death, and very many of them were 

 heavily fungused. None of them was spent. Most were unripe 

 females, and a few were ripe males. Many had gill net marks and 

 in some this was the cause of death. In one individual found dead 

 these twine marks had eroded the flesh so deeply that the fish was 

 readily broken in two on the lines of the mark by a slight pressure 

 of the hands. 



The origin of the commonest injury, however, is not readily ex- 

 plicable. It is a slough of the skin of the sides of the fish toward the 

 tail, the sore usually extending into the muscles and sometimes 

 deeply even before death, so that the muscle bundles may readily 



