ON LAKES 99 



chance. One is obliged to respect the gillie. His 

 theory looks eminently reasonable ; and, to carry it 

 out, by rowing in cunning curves he puts himself to 

 considerable pains. Still, I cannot be so confident 

 as he is. Often you hook a salmon when, the day 

 being nearly over, you are going directly to the 

 landing-stage. If the straight course is the wrong 

 course, how does that happen ? 



This is an interesting problem. I perceive three 

 suggestions towards solution. One of these, which 

 I shall mention last, has been, in Trout Fishing, 

 offered by myself. Each of the others is wholly 

 original, and will probably be considered untenable 

 by such as are given to quick judgments. 



In the first place, I can conceive it possible that 

 in water about thirty feet deep, or deeper, a boat 

 might be rowed right over a salmon at the bottom 

 without the fish seeing it. Any one who has studied 

 the ways of salmon in a river must have noticed 

 signs of certain peculiarities of their vision. If by 

 any means, such as by being in the stream, you can 

 approach a salmon straight in the face, you may get 

 within three or four yards of him before he sees you 

 and flashes off; this indicates that, though the fish 

 does see in front of him, he does not see far in that 

 direction. If you approach sideways, as from the 

 bank of the stream, he darts off much sooner, though 

 not so soon as a trout would ; this indicates that his 

 lateral line of vision is longer than his line of vision 

 to the front. As we have already had occasion to 

 note, he sees upward also ; but how far upward does 



