148 The American Salmon-Jisherman. 



fly to a smaller one of the same variety, allow three 

 minutes by the watch to elapse if the fish is small, and 

 five if it is large, and then have at him again. 



Do not cast at the fish, but off to one side so that the 

 current will swing the fly over it as before. Indeed, un- 

 less the angler is really a first-rate caster, and unless 

 every condition favors a really pretty cast, I doubt the 

 expediency of ever casting directly at a fish where the 

 current is such that a fly cast to one side will be swept 

 over it. A current is an able ally, and it and the angler 

 combined can present the fly in a far more attractive 

 manner than can either alone. 



Should the fly be again refused,* into the canoe with it 

 as before, change to one quite different in appearance, 

 rest the fish once more, and then tender the new fly in 

 the same manner. This can be continued until either the 

 angler or the fish has had enough of it, for the angler 

 may be morally certain that the salmon has returned to 

 and will remain at the place from which he first rose. 

 While it is quite true that a salmon will at times take a 

 fly it has risen to, but otherwise ignored, without any in- 

 terval of rest between the casts beyond such as is neces- 

 sarily incidental thereto, still if such a cast does fail, it is 

 pretty certain to change the suspicions of the fish into a 

 settled distrust which every subsequent effort to remove 

 will be futile. The other method, it is believed, will be 

 found far more profitable in the long-run. 



There is nothing in salmon-fishing more interesting, at 

 least to me, than a direct issue of this kind. The capture 

 of no other fish of anything like its size gives the pleas- 



* See p. 101, et seq. 



