SALMON AND TROUT IN ALASKA. 65 



instinct supports the inference, as does also the character of the 

 mouth parts, teeth, and gillrakers. Two examples were seen at 

 Karta Bay the first of August filled with sand launces; another con- 

 tained a herring. One example, a female, taken at Quadra early in 

 August, was filled with crab larvae, the same material upon which 

 sockeyes were feeding at the time. 



HUMPBACK AND DOG SALMON. 



The humpback is ordinarily not feeding at the time it is taken in 

 the cannery seines. Of a number examined at Karta Bay early in 

 August, some of the males contained crustaceans and in one instance 

 a sand launce ; none of the females had food. At this time the males 

 exceeded the females in number in the ratio of 50 to 15. At Kegan 

 early in August a few humpbacks were found containing the macru- 

 ran from that locality. At Karluk, July 30, they were found to be 

 feeding about equally on sand launces and Crustacea. 



The dog salmon has not been examined for feeding habits. Like 

 the coho, it approaches the streams late, and as a rule the stomachs 

 are then empty. 



RELATION OF FOOD SUPPLY TO NUMBER OF ADULT SALMON. 



The abundance of food has a direct relation to the number of adult 

 salmon that may be produced. The propagation of countless myriads 

 of fry can result only in their cannibalism if the supply of other 

 food is lacking. The dependence of the different species of salmon 

 upon other fishes is probably about in the same ratio as their average 

 size. The largest, the king, is largely dependent on the herring, sand 

 launce, cod, etc., for its subsistence, as are, doubtless, also the coho 

 and dog. The two smaller species, humpback and sockeye, are bet- 

 ter adapted to live on the more minute life. They are not known to 

 take a troll at any time. Their food is, at least largely, the food of 

 the species which in turn support the king salmon. The interrelation 

 has reached a nice adjustment in nature which, while not understood 

 in detail, is known to exist. The extinction of the herring might 

 lead to increased food supply for the growing sockeye. At the same 

 time it would deprive the dogfish and king salmon of their natural 

 aliment and cause them to become a menace to the otherwise favored 

 species. Only upon complete knowledge of the various elements of 

 animal life can recommendations be safely made regarding more 

 obvious matters. Empirical regulations may be successful, but rational 

 legislation must ultimately rest upon results from the bolting-silk 

 tow nets and the microscope. 



