370 THE SCIENTIFIC ANGLER. 



of breakage, from the cause stated. The tail end of the 

 fish, particularly of a large one, is to be inserted first in 

 the net, except in special cases, as when for instance, the 

 quarry heads close by the rodster, up or down stream, 

 when the net may be used as a receptacle for him to head 

 into, when within reach. 



The best times to fish, the seasonable flies and other 

 information for the fly-fisher, we give in the Monthly 

 Notes. We now purpose dealing more generally with the 

 subject of fly-fishing in its various bearings. Fishing in 

 rapid streams requires tactics somewhat different from 

 those suitable for ordinary smooth running river, or still- 

 water fishing. In these no humoring actions are re- 

 quired to be given to the fly. To draw against or even 

 across stream in these circumstances is to extend to it an 

 unnatural motion. The flies must be cast a few feet 

 above the dimply indication of a rise, and then allowed 

 to float over. This may be repeated several times before 

 moving on, especially if casting over grayling, as these 

 fish are given to take the proffered lure more often than 

 not when passed over repeatedly. With regard to the 

 vexed question of up or down-stream fishing, no strict 

 rule need be observed ; a continuous resort to either is not 

 desirable. Adherents of the one deprecate the ever- 

 recurring nuisance of the line becoming slack when cast 

 up stream by the downward flow of water, thus lessening, 

 after each delivery, the chances of a rise by the fly being 

 brought again home to the feet. Upon the other side of 

 the question we have arguments in plenty against down- 

 stream fishing, the most important being the habitual 

 position of the fish heading up-stream, and therefore in 

 full ken of the operations instigated for his allurement.* 



Our method of fishing a strange stream is, after pros- 



*The proportion of " down stream " to " up stream " fishers in Amer- 

 ican waters is probably one-hundred to one. 



