52 FAILURE OF NONDESCRIPT FLIES. 



summer, and hence the different sorts of dun 

 artificials, intended to be imitations of those in- 

 sects, are good general flies in every part of the 

 empire. Artificial palmers, which are imitations 

 of the caterpillars of different sizes and colours 

 common to the rivers of the British Isles, are 

 good baits, perhaps the best general ones, except 

 in those months when the living caterpillar does 

 not exist. 



After what I have now written, it will be ap- 

 parent to every one that I am in favour of close 

 imitation. I have tried the nondescript fly, and 

 found it fail tried it for two seasons on the 

 Thames without a shadow of success. Having 

 found how difficult it was to kill large Thames 

 trout with the ordinary artificial flies, I had some 

 nondescript ones dressed as attractively as imagi- 

 nation, guessing at probabilities, could make them. 

 During the seasons 1846-7 I used them with the 

 utmost perseverance, for I wanted to test the dis- 

 covery of the philosophers ; but the Thames trout 

 seemed determined not to afford me a single ex- 

 cuse for becoming a convert to the new doctrine 

 they would have nothing to do with my new- 

 fangled flies. In previous years I had killed 

 Thames trout with artificial flies, and I had made 

 others kill them with flies similar to those I had 

 used, viz. large red, black, brown, and furnace 

 hackles, and a very large imitation of the sand- 



