390 Fly-rods and Fly-tackle. 



parent that its small diameter made a great difference 

 in its favor. 



During all this the idea was gradually gathering force 

 that these experiments only tended to show how the ob- 

 ject appeared when viewed by a fish lying directly be- 

 neath it ; and upon trying some flies, and finding that 

 only with difficulty could the most gaudy be distinguished 

 from those sober in color, the box was dropped, and light 

 sought in another direction. 



A bath - tub of considerable size, its length facing a 

 window and the sky, was filled with water to the depth 

 of fourteen inches. Two mirrors were submerged in the 

 water, one at each end of the tub, and so inclined that 

 by looking down upon them the reflected image of any- 

 thing in or upon the water could readily be seen. A 

 joint from a rod was used to manipulate the leader to be 

 experimented with, and by moving it to and fro in the 

 water, it could be viewed at almost any degree of ob- 

 liquity. 



Here, again, the results were a surprise. Though I 

 have habitually used a colored leader, still I had supposed 

 color was of questionable utility. Such seemed not to be 

 the case. The coffee color was still the most conspicu- 

 ous, but it was but little more so than the natural-colored 

 gut, which latter, in all positions and angles, looked like 

 a streak of silver. The mist-colored leaders, in some posi- 

 tions, had the same appearance, but always it seemed in 

 a less degree; while at times, and at certain angles and 

 directions of motion with reference to the light, they 

 seemed more or less to disappear. The darkest tinted, a 

 decided azure, gave the best result. I was unable to de- 

 termine with satisfactory certainty in what positions in 

 reference to light, etc., this partial or total disappearance 



