132 OBSERVATIONS ON A SALMON RIVER 



where the stream, which here flowed at a 

 great pace, divided into several channels, 

 separated by narrow spits of shingle. 

 Across one of these channels a dead birch, 

 brought down by the flood, had become 

 fixed. My fish elected to rush down the 

 birch-tree channel just as the men had com- 

 mitted the canoe to the next. Things looked 

 bad, but this obstacle was overcome: canoe 

 and fish both navigated their channels with- 

 out mishap — the boughs of the birch-tree 

 sloped down-stream, and the line passed 

 over them without getting caught. We met 

 our fish again below the narrow island of 

 shingle which had parted us, after which he 

 renewed his headlong course. Finally, hav- 

 ing found a resting-place to his liking, he 

 went to ground in a deep hole from which I 

 tried in vain to dislodge him: in spite of 

 pressure applied from above and below his 

 place of refuge, he was immovable. I tried 

 to hand-line him up from the bottom, but it 

 became clear that he had literally taken root 

 in the mass of brushwood and debris which 



