36 Transactions of the Canadian Institute. [Vol. IX. 



hatcheries, which take depends upon the abundance of fish on the spawn- 

 ing grounds. In 1901, the year of the "big" run, the hatcheries were 

 running to their full capacity and could have obtained far more eggs than 

 they could accommodate. In 1902 their take amounted to only seventy- 

 three per cent, of their capacity, in 1903 to twenty-five per cent, and in 

 1904 to only ten per cent., but in 1905 it again rose to ninety-three per 

 cent., notwithstanding an increase in the number of hatcheries in operation 

 from two to five. 



These data point clearly to a relation betw^een the abundant spawning 

 in 1 90 1 and the large run of 1905, and similarly a relation between the 

 diminished amounts of spawn in 1902, 1903 and 1904 and the diminished 

 runs of 1906, 1907 and 1908. It seems probable, therefore, that the life 

 of the Sockeye extends over four years. In other words, the eggs spawned 

 this fall will produce fish that will reach the spawning grounds, spawn and 

 die in 1913. 



The chain of argument outlined seems fairly reliable, but there are 

 reasons for supposing that there may be exceptions to the rule. It has 

 already been noted that each spring there are two sizes of young fish 

 starting on their migration to the sea, fry and yearlings. The fry, it may 

 be presumed, will return to fresh water three years later, but do the year- 

 lings, which have already spent a year in fresh water accompany the fry 

 with which they descended on their return migration or do they return a 

 year earlier with the fish that were spawned in the same season as them- 

 selves? In other words, do they remain two or three years in the ocean. 

 On this point we have no definite information and we are in the same 

 position as to the causes that determine their prolonged stay on the 

 spawning grounds. And yet these may be matters of no little importance 

 in connection with the possibility of favourably modifying the runs in the 

 poor years, especially in that of the year succeeding a good run. 



Some fish also seem to return to fresh water before they have reached 

 the fourth year of their existence. At least this seems to be the most 

 plausible explanation of the occurrence of greatly undersized fish on the 

 spawning grounds. These fish, which are relatively few in number com- 

 pared with those that are full grown, are less than half the size of their 

 companions, weighing only two to two and a half pounds and measuring 

 sixteen to eighteen inches, and they are all males. Notwithstanding their 

 small size they are sexually mature and deposit fertile milt. No similar 

 precociously mature females have been observed. They have been 

 taken in lakes tributary to the Columbia River and in Nicola, Frangois, 

 Stuart and Shuswap Lakes, tributaries of the Fraser River. In the 



