HISTORY OF THE PACIFIC SALMON. 587 



.stomachs are invariably empty. This, however, may be from 

 lack of means rather than from want of will. When one hears 

 of a shoal of fish, seven miles broad and of unknown length, 

 heading for the land through the Straits of San Juan, one can 

 well imagine difficulties in the commissariat. Part only of this 

 l)ig shoal enters the Fraser River, the remainder moves on up the 

 coast of British Columbia. The photograph (text-fig. 103), which 

 I give by kind permission of Mr. Parry, late of Granite Creek 

 Hatcheiy, shows a detachment on its way up Scotch Creek, a 

 subtributary of the Fraser River, about 300 miles from the sea. 



The big run of which I have been speaking strikes the south- 

 west coast of Vancouver Island in July and August, coming from 

 the north-west, but a few Sockeyes run as early as April. In 

 the Fraser River itself the main run is in August, and some con- 

 tinue to come in until October. In the far north of Alaska the 

 main run is as early as June, which goes some way towards 

 showing that the feeding-grounds ai-e in the north, unless a 

 natui-al instinct to get spawning over before the winter sets in has 

 determined the habits of the fish frequenting the most northerly 

 rivers. In the Fraser district spawning begins in August, and 

 may go on until November ; spawning takes place only in streams 

 I'unning into or out of lakes. In this district, from which 

 most of mv information and all my specimens have come, several 

 hatcheries have been established, which in 1910 liberated 

 134,639,200 Sockeyes in British Columbian waters. Besides these 

 4,544,825 were liberated from U.S. hatcheries in Puget Sounds 

 and 257,021,790 in Alaska. 



I am indebted to Mr. W. J. Sim, who was emplo3^ed at Granite 

 Creek for some years, for the information that the fry are liberated 

 from the Canadian hatcheries as soon as the yolk-sac has been 

 absorbed, at a period of the year which varies from January to 

 April, according to the date of spawning. The main liberation is 

 about the first week in February. In that district the fry may 

 remain in the creek for two months or less ; then they move into 

 Shuswap Lake. The exodus from the lake takes j)lace from June 

 to September. The fry are supposed to be under one year old at 

 the time of their exodus from Lake Shuswap, but no one really 

 knows how long they may have remained in this, or in an}^ other 

 lake, and it is admitted by some observers that they maj^ be in 

 their second year. Dr. Greene, writing in the Bulletin of the U.S. 

 Bureau of Fisheries, vol. xxix. 1909, on the Migration of Salmon 

 in the Columbia River, and quoting Evermann, says the fry 

 begin their seaward journey not sooner than September of their 

 fi.i*st, and not later than July of their second year. That means, 

 I presume, that they leave the lakes when from nine to eighteen 

 months old. But others, e. g. Rutter, Bulletin U.S. Bureau of 

 Fisheries, vol. xxii. 1902, p. 102, say that they begin to descend the 

 rivers as soon as they can swim, and I'each the sea in about three 

 months. Mr. Sim says he is certain that there are no fry in the 

 rivers after September, and that they could not possibly avoid 



