THE SYSTEM OF ARTIFICIAL FLIES. 7 1 



— viz., that Trout take the artificial for the natural fly — 

 argues that because the natural dry fly usually floats 

 passively down the stream, the artificial fly — zvct — should 

 do the same. This is another analogical fallacy, the 

 error of which would seem hardly to require an almost 

 universally opposite practice for its demonstration. 

 Ninety-nine men out of a hundred find it best to give a 

 slight movement to the fly in the water. 



I fear we must include in the same "unproven" cate- 

 gory, and for the same practical reason, the theory that 

 flies should usually be cast up stream, rather than down 



To sum up the foregoing arguments therefore, the true 

 rationale of the matter seems to me to be as follows : — 



T. Whatever Salmon, et Jioc genus, may do, Trout cer- 

 tainly take the artificial for the natural fly. 



2. But as the artificial fly is necessarily presented in 

 an abnormal condition — namely, wet instead of dry, 

 sunk instead of floating ; and as the resemblance which 

 wet feathers and silk under water bear to dry insect- 

 down, flufl", and wings on the water, is imperfect, — (3) 

 it is necessary for the purpose of hiding the counter- 

 feit, and partly also to hide the hook, to give the fly an 

 unnatural, life-like movement in the water ; adding to it 

 also an unnatural quantity of legs (hackles) which open 

 and shut, and move with the movements of the fly. 



4. These " movements " and alterations, however, 

 make it quite impossible for Trout to discriminate 



