62 THE NATURAL HISTORY AND HABITS 



spawning in November and December ; a high or a 

 low temperature of the water during winter has the 

 effect of a corresponding difference in their assuming 

 the silver coating and leaving the river. Notwith- 

 standing the great throng going down in these months, 

 they are to be seen much later and earlier, and few or 

 more of them going down throughout the course of six 

 months, and likely longer ; for according to the length 

 of the spawning season must be the length of the 

 season that the smolts are going down. It was fully 

 the belief at one time that the whole crop of fry in the 

 rivers became srnolts at the same time in spring, and 

 all left the river with the first May flood. But that 

 would never do now ; for we have many times no 

 flood at all in May, therefore we would have of course 

 no smolts going down. To be sure a moderate spate 

 at times is favourable, for it relieves them from pools 

 in small streams and rivulets where they are confined, 

 but otherwise those that reach the open river will 

 go at their season, whether they get a spate or not, 

 and they do so, for they are often seen on their 

 passage when rivers are in low summer ply. 



I marked a number of smolts on their way to the 

 sea, and continued this process for several summers, but 

 yearly changed the mark from one part of the fish to 

 another. This marking was done particularly for the 

 purpose of ascertaining the exact time that the smolt 

 was absent from the river until it returned a grilse ; 

 because it was the opinion of some, and some even 

 assert it yet, that smolts are a whole year in the sea 

 before their return ; but if they were a year they 

 must be more they must have been fourteen months. 

 However, without marking them at all, we are certain 

 that salmon in no stage of their migrations remain 

 that time absent from the rivers. What I marked 

 was in April and May, and in course of June and 



