THE ATLANTIC SALMON 

 fortunate than others in meeting with plentiful food in the sea. For 

 this, or for some other unexplained reason, they are not all ready or 

 willing to return to the river at somewhat more than three years old 

 (which was the age of the marked grilse recaptured during the summer 

 of 1906, if it is assumed that all the smolts marked in 1905 were two 

 years old). 



Previous to these experiments it was universally assumed that all salmon 

 made their first reappearance in fresh water as grilse in summer and 

 autumn, and that all salmon spawned, under normal conditions, each year 

 of their lives. Fishermen were, and are, familiar with a class of small fish 

 running up very early in the season, similar in weight to grilse — ^that is, 

 from 4 to 10 lb.; but these have hitherto been regarded as fish that, having 

 spawned as grilse in a previous season, were returning to the river to 

 repeat the process. The result of the Tay netting in the first half of the 

 season of 1907 has shown this belief to be entirely erroneous. It will be 

 explained later how it was ascertained that these fish, and those which are 

 next to be mentioned, were returning to the river for the first time. 



The average weight of summer running salmon in the Tay and most 

 other rivers considerably exceeds that of the spring run of small fish. 

 It had never been suspected that these larger fish were maiden salmon 

 which had never been in the fresh water since they left it as smolts; yet 

 such was proved to be the truth. The heaviest marked fish retaken in 

 1907 weighed 27 lb., and was caught on August 9, its weight being 

 282 times greater than when it went to the sea two years and three months 

 previously. 



The results disclosed during the season of 1908 were equally surprising. 

 The four marked fish retaken were maidens, revisiting the river for the 

 first time, and were just completing their fifth year of life. The big fish 

 — 35 lb. — had been in the sea two years and eleven months, during which 

 period its weight had increased by 430 times, being on the average of about 

 half an ounce per day. 



In these experiments the actual age of the fish retaken admits of no 

 possible doubt. Six thousand smolts, all presumably two years old, were 

 marked in the Tay in the spring of 1905 ; no others were marked in a similar 

 way in any other river or in any other year. The data in calculating the 

 age of these marked fish cannot be impugned ; but the evidence upon which 

 Messrs Galderwood and Malloch base their diagnosis of these fish as 

 " maidens " must be sought in another direction. 



11 



