208 AUTUMN ANGLING 



artificial breeding of salmon. " Let us suppose," 

 it argues, "that we have a river with a gravelly 

 bottom, but destitute of fish of any kind. Let 

 us further suppose that we stock it with twelve 

 thousand salmon ova, and then leave it to take 

 its chance. Something like the following results 

 would ensue: The ova would hatch in April 

 or May; very shortly fish of prey would make 

 their appearance," &c. Now, in place of " sup- 

 posing" that "fish of prey would make their 

 appearance" soon after the introduction of salmon 

 into this formerly lifeless river, I firmly believe 

 that even the salmon smolts themselves, like the 

 Caithness nightingales, would never return after 

 the first migration. 



In Australia, where the salmon are forced to 

 come back to some one of the rivers of that 

 country or perish in the sea, it may be differ- 

 ent, but any river of this country communicating 

 with the sea, and destitute of salmon, grilse, or 

 sea-trout, will remain so if there were twenty 

 breeding-ponds on its banks. 



I have not fished the Tay since the establish- 

 ment of the Stormontfield ponds, but I think 



