THE SALMON FAMILY. 229 



distributed over a space of five or six months ; but irorri 

 information received from the best sources, and from reason- 

 able conclusions drawn from premises already laid down, it i« 

 doubtful whether the migration of Smolts — that is, in large 

 numbers — from the rivers of New Brunswick and Canada. 

 commences until late in the summer or early in autumn. I 

 did not take a Smolt during my last summer's fishing in the 

 Nipissiguit, although I caught numbers of the fry with bars 

 and red spots. Nor, in my inquiries, did I meet with any 

 person who had ever seen the young fish in the Smolt state 

 in that river. I therefore concluded, that the bars and spots 

 on the young fish disappeared later in the season, after the 

 angling, which lasts until the middle of September, was over, 

 and that it was not until after that time that they assumed 

 the silvery coat of the Smolt and migrated to sea. 



It is also a matter of doubt whether the young fish returns 

 as a Grilse until the following summer. The long time 

 necessarily occupied in incubation in those cold wateie, and 

 the length and low temperature of the winter which follows. 

 debarring it from feeding and retarding its growth, it appears, 

 would prevent its attaining the Smolt state until the autumn 

 of the second year, and too late to return as a Grilse before 

 the river was closed by ice. 



In opposition to this idea, it may be argued that Grilse, 

 when they do enter the rivers the next summer, after remain- 

 ing at sea all winter, should be very large. This is not the 

 case, nor does it necessarily follow, if we consider that the 

 extremely cold winter which closes harbors and bays, and 

 frequently fringes the shore for miles out with ice, must, of 

 course, lessen that class of animal life which constitutes the 

 ihief food of the Salmon at sea. 



From all accounts there is a disparity in the size of 

 A.merican and European Grilse. It appears from the " Book 



