62 SALMON AND TROUT IN ALASKA. 



Stomachs of adult sockeyes, as the fish are taken for commercial 

 uses, are commonly empty. In many cases the shrinkage which 

 follows the termination of feeding is pronounced, but usually the 

 upper limb is still lax, with only that degree of contraction due to 

 its emptiness, and the pyloric limb and coeca are still undiminished 

 in size. On the other hand, schools are sometimes taken in which 

 the shrinkage of all parts of the digestive tract is pronounced and 

 the characteristic yellow slimy or curdy secretion fills the canal. 

 The stomachs not infrequently contain scales, teeth, bits of vegeta- 

 ble debris, etc., all of which is probably incident to the pursing of 

 large numbers in the nets and consequent struggles of the fish. In 

 many cases the esophageal cavity is filled with a clear viscid liquid, 

 which appears to be largely dissolved jellyfish. 



In the examples examined at Yes Bay in 1905, undissolved por- 

 tions of the common globular ctenophore frequently occurred. This 

 same form was taken in great abundance in the surface and sub- 

 surface to wings in the bay and in Behm Canal just off the bay. In 

 addition, fragments were found of what seemed to be the larger 

 ellipsoidal, purple ctenophore taken in towings in Behm Canal and 

 elsewhere only in depths of 50 fathoms or more. It is not known 

 whether this form rises at night, henoe its presence indicates nothing 

 as to the depth at which the salmon feeds. Much the greater part 

 of the food observed is crustacean. In the examples from Boca de 

 Quadra, 1903, which have the highest percentage (36.8) of food- 

 containing individuals of any lot examined, the material is almost 

 wholh' larval crabs, a few of the macruran forms also occurring. 

 In Kegan examples the common crustacean form is a small macruran, 

 though a few crab larvse also were noted. In Yes Bay the most 

 ^\ddely distributed form is a macruran, but crab larvae are present 

 in great quantity. Yes Bay salmon stomachs showed much the 

 same aggregation of forms that the tow net ^developed, which indi- 

 cates that the food is taken by "to^\nng." 



Among the stomachs of sockeye salmon taken at Karta Bay, in 

 Moira Sound, and at Hetta, were found some containing sand launces 

 (Amrnodytes alascanus), in instances as high as 83 individuals in a 

 single stoniach. In these waters large schools of sand launces some- 

 what more than half grown are numerous during July and August. 

 Schools of 3'oung herring occur earlier and during the same period, 

 but in only one instance was a herring found in an adult sockeye 

 stomach. In Yes Bay a small clam worm, identified by Dr. J. Percy 

 Moore as Callizona angelini (Kinberg) Apstein, was found in 32 

 stomachs, or over 4 per cent of those examined. Several of these 

 worms were sometimes found in a single stomach, as if they had been 

 assembled by selective feeding. A very few were noted at Boca de 

 Quadra and at Kegan. These worms were taken only in the sub- 



