SALMON AND TROUT IN ALASKA. 71 



both early and late fish occupy the same grounds. The late comers 

 thus destroy the early eggs to a considerable extent. Partly from 

 this cause, perhaps, double runs are not well established ; perhaps the 

 fact that the volume of the streams is insufficient to carry the stimulus 

 of the fresh water to the spring feeding ground permits the fish to 

 remain together until all are induced by approaching maturity to 

 direct their movement toward the coast. 



Rutter found that of over 25,000,000 artificially hatched fry placed 

 in the Sacramento in 1897-98, in addition to the product of natural 

 reproduction, less than 1,090,000 remained as summer residents. 

 The following year the stations hatched about 17,000,000 from the 

 two runs, with probably a greater natural production than the pre- 

 vious year. The fry passed the lower river between the middle of 

 January and last of March, with maxima about six weeks separated 

 (January 25-30 and March 8-13), or less than half the interval that 

 separates the adult runs, if we assume that the fry taken in January 

 at Walnut Grove were from the early run eggs, as we must by con- 

 sideration of temperatures and the observed rate of travel. This does 

 not bear out the opinion that runs remain distinct. 



TEMPERATURE. 



Adjacent streams may have different seasons, i. e., time of run. 

 The variations in the small streams of the Loring district are of inter- 

 est. The early streams are the Naha, with dates of June 14-26 for 

 the first fishing; Karta Bay, June 13-26, and with these may be 

 included Klawak, June 16-29. The late streams are Yes Bay, July 

 11-17; the Moira Sound group (Dolomi, Kegan, and Nowiskay), 

 July 8-25; Quadra, July 13. The streams here listed as early 

 streams all drain lake areas of considerable surface and of a character 

 that would occasion an early rise in temperature. The late streams 

 have lakes of smaller total area or greater elevation and high sur- 

 rounding ridges. The temperature data at hand are entirely too 

 scant to demonstrate the difference suggested, but the few figures 

 obtained seem to point to such a condition. They indicate that the 

 streams are somewhat warmer than the surface in the adjacent salt 

 water at the time of the run, i. e., that the fish leave the colder for 

 warmer water. 



June 30 a line of observations was carried from the Naha to Karta 

 Bay. Off Loring the surface was 57°, density 1.0133. With some 

 variations the temperature reached its maximum, 59.5°, about midway 

 across Behm Canal, with density lowered to 1.0112; from this point 

 both temperature and density decreased until off Cape Caamano, 

 where the temperature had reached 54.5° and the density had fallen 

 to 1.0108-1.0110. Off the cape a momentary rise in density— 

 1.0136 — occurred. Crossing Clarence Strait the temperature varied 



