372 THE MOOR AND THE LOCH. 



sea-trout during floods, but are seldom large enough for salmon ; 

 the consequence is, that all the heavy fish remain in the loch 

 till harvest, and are very apt to haunt the mouths of brooks 

 during a spate. This is the case even in Loch Lomond, where, 

 with the exception of the outlet (the Leven), there is no stream 

 where salmon dare enter, except at the spawning-time ; and it 

 is amazing to see great fish at that season in mountain brooks 

 that will scarce cover, far less conceal them. 



In Loch Awe, where the fine feeder Orchy is ready to re- 

 ceive them, salmon seldom go up the loch at all. When clear 

 of the river Awe (the outlet of the loch), they immediately 

 press on to the Orchy, and enter it. Consequently, salmon 

 are scarcely ever taken in the loch, except between the 

 mouths of these rivers. 



When salmon know they have the safety of a loch to look 

 to as the reward of their efforts for fresh water, the struggle 

 they hazard to gain their object is almost incredible. I 

 have watched a shoal boring up the Knock for Loch Baa 

 when they had to turn on their sides to force themselves 

 through the shallows, and the noise caused by these exertions 

 could be heard at a considerable distance. At first I fancied 

 some creature was fording the stream, and, on walking up saw 

 monsters of from 10 to 25 Ib. fighting against water often 

 only a few inches deep. 



I have always been convinced that trolling is ruination 

 to fly-fishing in small salmon lochs. All the shores, banks, 

 and rocks have their feeding or resting fish driven into deep 

 water, scared not only by the boat, but even by the long 

 trolling-lines sweeping over. 



Fly-practice in lochs from a boat is far less startling to 

 fish ; and the more complete the angler's knowledge of every 

 sunk bank, rock, or shallow point, the less likely will he be 

 to frighten them away. In order to obtain this angling 

 geography, the loch should be surveyed in dry weather ; and 

 when the different feeding-grounds are discovered, they should 



