T 11 K SALMON FAMILY. 213 



number taken by the angler at that time, compared with 

 Salmon, is often five to one. 



The early emigrants of course are prepared to spawn first ; 

 but as a general rule, all the Salmon leave the pools and 

 rapids, and collect for this purpose on the shallows and in 

 the tributaries of the rivers by the middle of October,* and 

 the spawning season, instead of extending over a period of six 

 months, as it does in Scotland, hardly embraces as many weeks, 

 for by the last of November the rivers are generally closed 

 by ice, and the spawning beds sealed against the fecundating 

 influence of the air, and many of them are frozen hard even 

 to the bottom by the middle of December. The probability 

 therefore is, that incubation is arrested for months by the 

 spawn being encased in ice ; so, it follows, that a much longer 

 period (perhaps even double the time) is required for the ova 

 to hatch, than is necessary in the temperate waters of Scot- 

 land and Ireland. I have never read or heard of any 

 person having tested it in America by experiment, as has 

 been done in Scotland, but it is probable, that spawn de- 

 posited in American Salmon rivers late in October or in 

 November, does not produce the young fish until the ensuing 

 month of June or July. But it is not my object now, to show 

 that the general rule for the time of incubation, as laid down 

 by Scotch naturalists and observers, is inapplicable to the 

 northern waters of America, from great disparity of winter 

 temperature ; but to give the result of their experiments 

 carefully conducted through a long series of years as to the 



* In Scotland and Ireland, where the rivers are open all the year, 

 Salmon begin to come into fresh water in January and February, and 

 continue to do so every succeeding month until October. The rivers of 

 British America being ice-bound four or five months in the year, the time 

 of coming and going is limited to about four months. 



