SALMON AND TROUT IN ALASKA. 



37 



the middle of Ma}*, when the temperature of the lagoon had reached 

 45° F., though the upper river was still about 40°, the young fish 

 began leaving in large numbers. The run, however, after lasting 

 about two weeks, fell off abruptly, before any marked rise in tem- 

 perature had occurred. In 1904 ice had left the lakes very much 

 earlier, but the temperatures were not notably higher the middle of 

 May than in the previous year; j^et on May 17 the movement was 

 at full height and continued so till the end of the first week in June, 

 when the water had reached a surface temperature of over 50°. It 

 then fell away little less abruptly than in the previous year. During 

 this period the fish showed no corresponding increase in size as the 

 season advanced, i. e., they maintained about the same average for 

 the four weeks covering the main run, only the remnant stragglers 

 late in June having increased in average length. 



In natural spawning the first deposited spawn must be in no small 

 part destroyed by the activities of late-coming fish. Of the eggs 

 deposited after low temperatures obtain, the product of the earlier 

 will perhaps to a certain extent be evened with the later hatching 

 in that it will not develop rapidly during the winter, and %vith the 

 rapidly warming water of spring the incubation value of each day 

 augments to lessen the total number of days required in the develop- 

 ment of the fry from the egg. Hence, while some fry may be ready 

 to move with the opening of the river from ice, the maximum num- 

 ber ^\^ll accrue gradually and fall away abruptly even though the 

 spawning be more or less evenly distributed over a greater period, 

 as appears to be the case in species migrating as fry. When the 

 young feed for a period or season before migrating this effect will 

 be obliterated by the superior ability of the larger of them to obtain 

 food and hasten their growth. 



That the migration of j^earlings takes place in such manner that 

 the average size of the migrants is about the same throughout the 

 season is shown in the following table: 



Lengths of Mioratinu Yearling Sockeyes at Station 2, Naha River, 1904. 



During the month occupied in the migratory movement there 

 should have occurred a growth of some 4 mm. in average length, 

 but this in nowise appears, and it may be that the late-hatched 



