76 SALMON AND TROUT IN ALASKA. 



PARENT STREAMS. 



The election by salmon of certain streams as a spawning ground 

 to the exclusion of others has long been noted in the case of the sockeye 

 and king salmon. The choice on the part of the latter seems to be 

 based mainly on the volume, only stragglers of this species entering 

 the smaller streams like the Naha and Yes rivers. The larger streams, 

 such as the Stikine and Unuk, carry large quantities of silt and 

 thereby form extensive mud flats, which, with the drift carried on 

 the strong currents, make fishing difficult. For this reason no ade- 

 quate idea of the number of king salmon entering these streams has 

 been obtained. The only small stream in the region near Loring 

 which they are known to frequent, one tributar}" to Carroll Inlet, has 

 not been examined. They are commonly thought to have in Alaska 

 a predilection for streams of glacial origin, but all large rivers of this 

 section have some glacial water. 



The sockeye is recognized as exclusively' a frequenter of streams 

 with tributary lakes. The only suggestion adverse to this is a report 

 by Mr. Fred Patching, superintendent of the Fortmann Hatchery, 

 regarding a small stream in the vicinity of Wrangell, which Mr. Patch- 

 ing believes to have no lake water. The sockeyes are said to spawn 

 in a somewhat expanded portion of the stream, but this expanse has 

 not the character of a lake. As a converse proposition, not all lake 

 outlets, even of apparently suitable character, attract the sockeye. 

 The Anan stream, tributarv^ to Bradfield Canal, seems to be perfectly 

 adapted to all the sockeye's requirements, but no recognized run 

 enters it. 



The dog salmon apparently recognizes some distinctions, though it 

 is much more generally distributed than the two above-mentioned 

 species. For example, comparativelj' few of this cpecies enter the 

 Naha or Yes rivers, whereas Karta River, on the opposite side of the 

 strait, abounds with it. 



The coho is probably less particular in its requirements. The fry 

 were found, without exception, in every stream and brook examined; 

 even a tiny seepage rill entering Naha Bay which would become 

 dry with the first week of fair summer weather contained its little 

 school of coho fry. That some streams are profitably fished for cohos, 

 whereas others apparently as suitable are found not productive, is 

 probably due mainly to the stream character, this in the one case 

 being such as to promote the schooling of the fish at the mouth of 

 the stream, while in the other a greater length of estuary or greater 

 volume permits immediate ascent. 



The humpback, like the coho, is found everywhere, and being 

 exceedingly abundant is the more noticeable of the two. 



The evidence in the question as to the existence of a homing instinct 

 in the salmon is still lacking in many points. The exposition by 



