IMITATION OF THE NATURAL INSECT 171 



was rather clumsy, and he took fright and disap- 

 peared. Getting as close as I could to the water, 

 I hid behind a bush and watched for the fish to 

 return, which he did in just four minutes, timed 

 by my watch. This in itself was interesting, as 

 it tended to show how long the incident lingered 

 in his memory. The eddy which he occupied 

 was formed at the bottom of a rather swift little 

 run by a large boulder that deflected a part of 

 the stream toward the bank and started it up- 

 stream again. The fish stationed himself ex- 

 actly in the centre of this up-stream current, 

 which was not very strong, and immediately 

 began feeding. He rose three or four times a 

 minute, sometimes oftener, according to the op- 

 portunities presented. There were very few 

 insects in the air, but apparently a great many 

 upon the surface of the water. I think perhaps 

 a half dozen or so of different sorts alighted di- 

 rectly in the eddy, all of which the fish accounted 

 for, but the majority of rises were to insects 

 that were carried down-stream upon the surface, 

 and collected in the eddy. They were of all 

 sizes and shapes, from the tiniest Diptera, which 

 interested him much, to a small, dead butterfly, 

 lying flat, which he examined closely, but de- 

 clined. It was this discrimination that puzzled 



