78 SALMON AND TROUT IN ALASKA. 



data not only on the life history of these species but especially on 

 the physical requisites for practical work. The returning adults in 

 these instances reached only the streams in which the fry had been 

 placed, the two species not commingling. It is yet too early to draw 

 positive conclusions regarding the relative influence of the two factors, 

 naturj^l suitability of the different streams, and a homing instinct. 



That any given stream within the natural range of a species is not 

 inhabited would seem to indicate its natural unsuitability. Whether 

 the example of the Chinook River, where the king salmon once 

 placed in the current of the river seems to find the locality at least 

 endurable, can stand as a refutation remains yet to be proved. That 

 it is not naturally populated by this species has been explained by 

 the peculiar nature of its location and the general character of the 

 stream and its bed. At least one of the principal guides directing 

 the salmon to their spawning grounds seems to be the current. That 

 the fish do not enter the Chinook River by choice is thought to be 

 due to the lack of any sensible current from that stream reaching 

 into the path followed by the schools in their movement from the 

 ocean into the Columbia River. Temperatures have not been 

 recorded. This same apparent influence of the current has been 

 noted in many instances. For example, in the Columbia, salmon 

 pass the Willamette during the summer season of flood, when the 

 tributary river has become back water, but enter it before and after 

 that season when the current is perceptible. At the Quesnel Dam in 

 the Fraser system sockeyes died fighting the swift current of the race, 

 while ignoring the feeble flow from the original fishway, 



"return of MARKED SALMON." 



The experiment in which the marked Naha fry subsequently 

 appeared as adults at Yes Bay, as noted above, nmst be again men- 

 tioned. This result is a direct refutation of the homing theory. The 

 relative numbers at Yes Bay and the Naha are the more remarkable 

 for the reason that no Naha fish are taken except for breeding, hence 

 almost all the individuals returning to that stream would be discov- 

 ered, while at Yes Bay many may have been taken in the cannery 

 seines and the record lost. In all experiments based on marking 

 young fish by various mutilations the possibility of regeneration 

 of the lost parts must be fully taken into consideration. In this 

 experiment it is believed the fins were cut much closer than in 

 the laboratory experiments. 



"distinctive AND CHARACTERISTIC RUNS." 



Upon this point some more or less conclusive facts have been deter- 

 mined. The group of streams fished by the Loring Cannery were 

 studied in 1903 and 1904, careful counts and measurements of a num.- 



