A CURIOUS OCCURRENCE. 239 



having started from his stream, was caught in it that 

 night. 



John Crerar mentioned to me another instance where 

 a salmon, having broken a fisherman's line, went down 

 the Tay for a mile, and then up the Tummel three 

 miles, and was there caught the day following by the 

 same fisherman, who thus regained his fly with two 

 or three fathoms of line attached to it. 



On the other hand, I know of three well-attested 

 instances of salmon having been caught almost imme- 

 diately after they had broken the fisherman's line ; 

 but I conclude these fish were touched at first in a 

 part that was scarcely sensitive. A very curious cir- 

 cumstance of this sort occurred in the Isle Isla, where 

 a gentleman was broken by a salmon, which he caught 

 immediately afterwards. Upon landing it, he found, 

 to his amazement, that he had not touched the fish 

 itself the second time, but that his hook was linked in 

 the one left in his mouth previously. This was a very 

 delicate affair ; for had not the pull upon the fish been 

 moderate and even, he must inevitably have escaped. 

 As for my own practice, I never recollect having risen a 

 fish a second time that had touched my hook previously. 



What I have said regarding the number of fish lost 

 or set down in trolling is so universally acknowledged, 

 that this style of angling is seldom practised, except, 

 indeed, in fishing for kelts in very full waters, when no 

 one can throw completely over the casts without the use 

 of a boat. In this state of the river the flies are drawn 

 down the stream ; but when the water subsides, they 

 are trailed up it. It is practised also a day or two 



