26 SALMON AND TROUT IN ALASKA. 



During the season of 1903 a 35-foot trap was used except where 

 otherwise indicated in the table; in 1904 a 45-foot trap. 



In 1903, in 17 sets of the trap at Hirsch Rapids, station 2, between 

 April 21 and August 12, on only three occasions were sockeye fry 

 taken, twice a single example and the other time fifteen. In six 

 trials from April 21 to May 11 no sockeyes were taken; on May 17, 

 23, 24, and June 2, a total of over 2,000 sockeye yearlings from 52 to 

 97 mm. total length, together with over 2,500 humpback fry, were 

 taken. Thus it would seem if any great number of sockeye fry were 

 migratmg more would have been found m the net. 



In 1904, between May 17 and June 30, at the same point, over 

 8,300 sockeye j^earlings were taken, as well as over 3,300 humpback 

 fry and nearly 350 coho fry, but no sockeye fry. 



In 1903 a trap was placed in the Naha at station 1 as soon as the 

 lagoon was sufficiently clear of ice to permit use of a boat, April 12 

 to 19, thus probably securing the earliest movement of humpback 

 fry. At this time all the lakes above were entirely covered with ice. 

 April 19 a set at station 3, just below Jordan Lake, took only hump- 

 back and dog-salmon fry. It can not be that the run of sockeye fry 

 was over before the work was begun, since great numbers were only 

 recently hatched, nor is it likely that it occurs later in the summer, 

 for no fry of any kind were taken in a trap set August 12, 1903, July 9, 

 10, and September 2, 1904, nor in one set in the outlet of Yes Lake, 

 July 18 to 21, and August 5 to 26, 1905. 



A remote opportunity for error lies in the fact that, Roosevelt 

 Lagoon being salt water at the bottom, a deep-water fish, which the 

 yoimg sockeye undoubtedly is to an extent, might pass the rapids at 

 high water and go imder the trap. But this is extremely improbable, 

 since in British Columbia waters, where the migration has been 

 studied, the fry passed in the same manner and at the same season as 

 the yearlings. It is further shown to be unlikely by the fact that a 

 trap set below the mouth of Jordan Lake at station 3, May 15, 16, and 

 17, 1903, took only 2 or 3 sockeye fry, while securing some 12,000 

 humpback. At the same point June 2 and 3 only 14 sockeye fry were 

 taken. At this date the humpback run was practically over. Nor 

 did a trap set at the foot of Heckman Lake, station 4, May 12 to 14, 

 show any considerable movement, 6 sockeye fry being taken, with 250 

 yearlings. 



In 1903. the first movement of sockeyes at salt water was noted 

 May 17. May 11a trap set at station 2, in which over 1,000 hump- 

 back fry were taken, failed to secure any sockeyes. The migration 

 probably began on the high water of May 12 to 15; 300 yearlings were 

 taken on the latter date below Jordan Lake. May 17, 220 were 

 taken in a trap spanning about one-third of the channel at station 2 ; 

 on the 23d over 800 were secured by a trap of half the size, spanning 

 about one-fifth the channel; in this small trap 375 were taken on the 



