72 SALMON AND TROUT IN ALASKA. 



around 55° — density rising from 1.012 to 1.0216— with which 

 increased density the temperature fell again to 51°, and this, falling 

 in places to 50°, was carried with a density of about 1.020 through 

 Kasaan Bay into Karta Bay, when the density abruptly dropped to 

 about 1.012, with almost no change in temperature. 



These figures show a general decrease in temperature correspond- 

 ing to a rise in density. They show further that the surface density 

 and temperature of Kasaan Bay is little affected by the fresh water 

 from the Karta stream, whereas Behm Canal is markedly affected by 

 its affluents. At this season sockeyes are running into both the 

 colder waters of Karta Bay and the warmer waters of Naha Bay. 

 The Karta Bay fish are reported to be seen first farther south along 

 the west side of Clarence Strait. Hence they may never come in 

 contact with the currents from Behm Canal. 



Another series of observations was made July 30 and 31, 1903, 

 from Kegan stream to Dolomi via Old Johnson Stream opposite 

 Kegan. In the Kegan Basin the temperature was 60° and density 

 1.003; at the Kegan fish scow, anchored just outside the mouth of 

 the stream, the temperature was about 53° and density somewhat 

 over 1.020; in the middle of Moira Sound, between the two streams, 

 the temperature fell to 51° and density rose to near 1.022. In Old 

 Johnson Bay a temperature of 54° and a density of over 1.020 were 

 found. Passing down the sound, the temperature fell gradually, but 

 varying, to 51°, with corresponding rise in density to near 1.022. The 

 inflow from North Arm (Nowiskay) could scarcely be noted. The 

 lowest temperature was found off Point Adams, 48°, with a density 

 of over 1.022. The stream at Dolomi exhibited no influence outside 

 the arm into which it flows. At the fish scow in the inner bay 

 the temperature was 57° and density 1.0032. Five minutes' run of 

 the tug reached water of 52° F. and density of over 1.020. 



If the fish travel on the flood tide, it* must be necessary for them to 

 approach these small streams very closely in order to feel the influ- 

 ence. The observations recorded above are all surface, where the 

 greatest change would be made by the fresh water. It would seem 

 scarcely possible to detect the river water outside the small basins or 

 estuaries at a few feet below the surface. 



It is not impossible that the incoming schools closely follow the 

 coast line. For example, it has been stated'* that sockeyes enter 

 Uyak Bay, on the east side of Harvester Island, pass to the south 

 and east side of the bay, thence to the mouth of Larsen Bay, after 

 which they follow the west side of the bay at some distance from 

 the shore and pass out again into Shelikof Strait at the point where 

 first seen. Since Uyak Bay contains no sockeye streams, such 



a A. B. Alexander, ms. notes. 



