46 



So much for the technical skill exhibited in this new departure 

 in fly-work. 



In the case of salmon, we know that external objects impress 

 themselves upon the ftsh in two ways. At one time, directly, through 

 some natural agency — light, contact, force of current, etc. At another 

 time, indirectly, as when they get the impression of a fly which 

 nourished them in their infancy, not from the insect itself, but from 

 a specific imitation containing something associated or connected 

 with it, which, as experiment teaches, is an appropriate stimulation of 

 their memory, one that revives the recollection of their favourite food 

 and stirs them to action. 



The natural impulse of salmon is guided by instinct, just as the 

 infant in arms sucks tiie bottle. But as they sometimes take the 

 wrong river, we cannot speak of their instincts as unerring. Very 

 slight stimulation which comes into their " thinking apparatus " — e.g. 

 excitement produced by an extra bit of blue, or an extra bit of red in 

 a fly — sets them in a state of passionate activity. 



How can this be " rubbish "? Of course it cannot be when we 

 compare notes and reflect. 



As a rule, the greater part of the ideas relating to fly-invention 

 are not reasoned out by the student in the first place, but are imparted 

 to him by the master-hand as part of his education. Not a few, on 

 the other hand, have become so settled by actual experiment as never 

 to be questionable, at any rate at certain periods ; and they only decay 

 by the substitution for them of new habits of thought occasioned by 

 observation of some such evil efi'ect produced on die fish, say, by 

 over-thrashing, when, of course, the conditions whicli led to the 

 making of the former " awfully killing fly " have become changed. 



