34 



rivers, tlierefore, judgment must be exercised. Supposing a two-inch 

 fly is the full size used, we should mount one of about an inch and 

 a half in length. 



When the angler's ingenuity, however, is so over-taxed in his 

 efforts to cover distant fisli with large flies, that by splashing the 

 water he alarms the aquatic camp, the onerous duty of changing 

 front becomes imperative. To begin with, an interval of ten minutes 

 before the renewal of hostilities is a sine qua iioii ; he has then to 

 make up his mind to get nearer the fish ; of course he will fare worse 

 if he gets farther away from them. For this reason his fly must differ 

 both in size and character. The size is reduced, because the fly will 

 inevitably travel across the water at less speed. And, to carry the 

 mind back a few pages, the fish, having seen a dusky pattern partly 

 composed of feathers impervious to the light, should now be offered 

 a gaudy specimen, even as radiant as " Beaconsfield " itself. 



Thus we have a common illustration of what we are pleased to 

 call " contrast." 



Our subject now takes a more definite turn. 



We have done with the sun, we have done with the shade, and 

 partly with geological formation, and will now consider the principles 

 to pursue at the riverside generally. 



As the variety poet says, " it takes a bit of doing," in the absence 

 of local advice, to decide off-hand the size of fly to put on ; but still, 

 the behaviour of the fish will soon give one the cue. 



If, for instance, a salmon only follows a fly, one smaller and 

 brighter must be used. Should, however, a swirl be seen, and no tug 

 or rise follows, the fish should be rested for about eight minutes. 



If a salmon comes sharply in rough water and misses the fly, tliree 



