SALMON AND TROUT IN ALASKA. 31 



SUMMARY OF OBSERVATIONS. 



From these various records it would seem that in the larger rivers, 

 aa the Karluk and Fraser, many sockeye fry leave their nursery 

 waters as soon as they are able to swim and feed, or at the beginning 

 of the fingerling stage ; that in the smaller streams, as the Naha, this 

 number is small and, in comparison to the number resident for one 

 season, negligible; that in all cases a portion, perhaps the greater, 

 remain in the lakes for one winter and migrate in the following spring 

 as yearlings. Whether any spend a second wdnter in the fresh water 

 has not been demonstrated. This residence of large numbers of 

 young in fresh waters with a definite and apparently anticipated 

 movement seaward the second spring of their existence is unique in 

 this species. AVliile some king and coho young spend the summer in 

 the vicinity of their birthplace there is no evidence that these summer 

 residents do not pass to the sea as a rule during the fall, or for that 

 matter at any convenient time. Their continued stay seems to 

 depend largely on the period of rainless weather that obtains on the 

 Pacific coast during the summer months. Only a small remnant of 

 the lake-inhabiting iingerlings of these two species spend a winter in 

 fresh water. Humpback and dog salmon all leave- for the sea almost 

 as soon as they are able to swim. This habit of the sockeye is doubt- 

 less connected closely with the habit of lake spawning, or is even the 

 immediate effect of that cause. Further, it probably has given rise 

 to the dwarf form, of which no analogue exists in the other speciec. 



YOUNG SALMON IN FRESH WATER. 



THE SOCKEYE. 

 MOVEMENT OF FRY ABOVE THE LAKES. 



The sockeye fry when hatched or planted in the tributaries of a lake 

 linger a little while on the nursery ground. Small plants made in 

 pools near Loring in 1903 remained a considerable time after the 

 absorption of the yolk sac, and the same fact was noted at the nursery 

 pond at Fortmann Hatchery the following year. Similar observa- 

 tions were made also at Karluk. The first creek entering at the foot 

 of Karluk Lake is a small creek consisting of three pools connected b\^ 

 short rapids and fed entirely by springs. A few sockeyes frequent it 

 for spawning, the greater percentage, apparently, being weak fish 

 that are unable to continue the journey. It may be, too, that this 

 creek remains open late in the fall and fiiids favor with late-coming 

 fish. On July 14 a net was set across the mouth, taldng in daylight 

 4 adult salmon, 1 stickleback, and 1 sculpin. At night 6 adults and 

 87 young sockeyes, 7 sticklebacks and 1 sculpin were taken. Of these 

 young 11 were feeding and had reached an average length of 41 mm. 

 10731—07 3 



