SALMON AND TROUT IN ALASKA. 63 



surface to wings at Yes Bay, btit they probably came to the surface 

 in the evening at that season. 



It is remarked, in general, that the earlier runs of fish contain a 

 higher percentage of feeding individuals than the later. Much 

 depends also on the locality. At Dolomi practicalh'.all the fishing 

 is done in a small bay almost at the river's mouth. It is improbable 

 that the fish enter this bay in great numbers until ready to ascend 

 the stream. Of 511 examples opened August 4 to 11, 1904, none 

 contained food; many of these stomachs were shrunken, but some 

 were still lax. Of 200 examined August 21 to 23, 1903, none of the 

 first hundred contained food. These fish were of a distinctive type, 

 more or less characteristic, long, slender, and dark in color. Of the 

 second hundred, taken two days later, 9 were feeding on sand launces 

 and crustaceans. These were brighter fish of a different type — 

 deeper and more resembling the fish of other localities. 



The fishing at Nowiskay was done by a single crew, hence ordi- 

 narily all or nearly all were taken near the mouth of the river. In 

 1904, owing to the scarcity of fish, the crew sometimes hauled in the 

 lower end of the arm, which may help to account for the larger num- 

 ber of feeding fish taken that season. This same circumstance has 

 a bearing on the fish examined at Yes Bay in 1905. A greater part 

 of the fishing was done in Behm Canal than in former seasons. The 

 paucity of food material in the bays may sufficiently explain why 

 the salmon do not approach the rivers until about ready to make 

 the ascent. 



Digestion in fishes is ordinarily quite rapid, and for this reason we 

 may be sure that food found has been recentl}'^ ingested; but the 

 fishing crews are out with the daylight, atnd our knowledge of the dis- 

 tribution of forms at that hour is too limited to permit any conclu- 

 sions as to the depth at which salmon feed. 



The presence of such large percentages of feeding fish might lead 

 to the supposition that the fish were still on the natural feeding 

 grounds and had not yet been made subject to the changed instincts 

 of approaching maturity. This view is somewhat supported by the 

 absence of food as reported in the large schools of the greater rivers 

 at relatively earlier dates. On the other hand may be noted the fact 

 that the Alaskan schools are much smaller, hence the opportunity 

 to feed, if the instinct to do so still remains, is much better. A 

 salmon stomach at any time, if empty for a period, becomes con- 

 tracted and gives the appearance of being shrunken. This perhaps 

 is due to the lack of blood in the tissues and the action of the con- 

 tractile fibers. It is noted in migrating fingerlings as well as in adult 

 fish. With the presence of food and excitation of secretion, blood 

 will return to the tissues, the bulk of the ingested mass will stretch 

 10731—07 5 



