70 SALMON AiTD TROUT IN ALASKA. 



elusion that the reproductive purpose is mainly instrumental in bring- 

 ing them into the streams. Correlated with this in determining the 

 particular time of year for the entrance into fresh water are sea- 

 sonal influences, temperature, density, currents, character of the 

 river, etc. There may be sufficient explanation of the variation in 

 time of runs in the proposition advanced by Moser, viz, that in streams 

 where the run is large both the precocious arrivals and the stragglers 

 are in sufficient number to be noted and recorded. But tliis can 

 scarcely account for the distinctly double maxima noted in the runs 

 of the Columbia and Sacramento rivers. The sockeye is said to 

 "run" earliest in the Aleutian Islands, where it arrives in May. It 

 is said to be found in the vicinity of Umnak Island all the year, in 

 which case it must be feeding. In Bristol Bay the middle of June is 

 given as the time of arrival, but the run is usually small until in July. 

 At Karluk it is earlier, sometimes the first part of June, while in 

 Southeast Alaska the arrival is delayed in certain streams until the 

 second week in July. 



THE DIFFERENT RUNS. 



In large rivers different runs of salmon are noted — for example, in the 

 Columbia and Sacramento the king salmon has two runs per year, 

 or two maxima; the early fish are supposed to travel to the head- 

 waters of the various tributaries which they enter, the later run not 

 to ascend so far. The April-June (?) arrivals spawn in July and 

 August; the later fish spawn September to November. The reason 

 for the separation is unknown. A well-marked division appears only 

 in the Columbia and Sacramento. In the former river both king and 

 sockeye are so separated. The Karluk is said similarly to have two 

 runs, one maximum about the last of June and one the first of August, 

 but this was not true in 1903 when the river was under study. A 

 double run might be supposed to bear some relation to the two 

 migration periods of young, i. e., as fry or yearlings, but this is not 

 borne out by the conditions in the Fraser, where such migrations 

 are also known. In the Sacramento, the distance being short and 

 the winters mild, the hatching, periods and time of reaching salt 

 water are distinct for each run. In the Columbia the elevation of 

 the headwaters, with attendant low temperatures and distance which 

 the fry must travel to reach the sea, may equalize the time of arrival 

 at salt water. 



The question whether an early hatched fry, the offspring of an 

 early run parent, will in turn become an early run adult is of much 

 importance in fish-cultural problems. In the Sacramento both runs 

 receive attention at the propagating stations. On the Columbia it is 

 mainly the late fish that are propagated, with what seems to be the 

 result that early fish are becoming fewer. In the Fraser the early run 

 may escape by reason of the close seasons. In short Alaskan streams 



