LOOKING UPWARD 37 



the light, and it would probably appear indistinct in 

 colouring, for one would be looking perpendicularly upwards 

 into the light. A trout, however, seldom sees a floating 

 fly by looking perpendicularly upwards at it. Indeed, at 

 the moment of taking it it must be at least the distance 

 from his eye of the tip of his neb, and while the fly is 

 approaching the trout more or less rapidly it must be 

 seen at an angle to the perpendicular from the trout's eye 

 to the surface. In these conditions the trout cannot 

 always have the fly between him and the strongest light. 

 If the sun be low and be shining directly down-stream 

 towards the trout, then the conditions would approximate 

 to those of the experiment quoted, and the light would 

 be entirely from behind the fly, and its colour, unless it be 

 transparent, would not perhaps be readily distinguishable. 



But if the sun were exactly the opposite way and were 

 shining exactly behind the line in which the fish is swim- 

 ming, would not one suppose that the approaching fly 

 would receive enough illumination to enable the trout to 

 apply to it such appreciation of colour as he possesses? 

 Between this condition of things and the condition pre- 

 vailing when the light is coming right down into the trout's 

 eyes, there must be a large range of conditions in which 

 a greater or less degree of visibility of colour would appear 

 to be possible. 



This is, of course, all upon the assumption (which may 

 not be correct) that the trout sees as man does. His eye 

 is adapted to the medium in which it works, and it is at 

 least conceivable that it is so constructed as to enable him 

 to overcome the difficulty of appreciating colour with the 

 fight behind it. It is certain that it has a faculty of choice 

 of fly, indicating a degree of appreciation of colour, or tone, 

 or texture in the dark or deep dusk, so that it might not 

 want much light on the underside of the fly to enable it to 



