ii6 SALMON AND TROUT. 



lake rivers are beginning to fail. Of these operating causes 

 two of the Sutherland streams afford good examples. One, 

 the Oikel, springs from a small exposed alpine pool some half 

 mile in breadth ; the other, the Shin (a branch of the Oikel), 

 takes its rise in the deep sweeping waters of Loch Shin and its 

 tributary lakes. The Shin joins the Oikel about five miles 

 from the sea. Early in the spring, all the salmon entering this 

 common mouth diverge at the junction, pass up the Shin, and 

 thus return, it would appear, to their own warmer stream ; 

 whilst very few keep the main course of the Oikel until a much 

 later period. 



Nor does it appear that these operative causes and their 

 resultant effects are confined to Scotland. An analogous 

 instance, indirectly traceable to the same cause, has been 

 pointed out by Dr. Heysham, in his ' Catalogue of Cumber- 

 land Animals,' as observable in several of the rivers of that 

 county : The salmon, during winter and spring, evidently 

 prefer the Eden to either the Esk, Caldew, or Peteril, although 

 the Eden and the Esk pour their waters into the same estuary, 

 and, in fact, are only separated at their mouths by a small 

 promontory. There is hardly an instance, Dr. Heysham 

 asserts, of a salmon entering the Esk until the middle of April 

 or beginning of May a circumstance always referred by local 

 fishermen to the difference in temperature between the two 

 streams. The waters of the Eden, they allege, are consider- 

 ably warmer than those of the Esk, which, from the shallow 

 and rocky character of the bed of the Esk, appears not im- 

 probable. 



Be this as it may, it is an indubitable fact that snow water 

 prevents salmon from running up even the milder stream of 

 the Eden. 



The Caldew and the Peteril, again, pour their waters into 

 the Eden, the one at, and the other a little above Carlisle ; 

 yet up neither of these rivers do salmon ever run, unless at the 

 spawning season, and then but in small numbers. 



The rule, however, which would appear to be inferred from 



