THE SENSES OF FISHES. 13 



that fish can see indifferently well in either, of course,, 

 but rather from the fact that it sees so excellently in 

 water, and from the comparative fixity of the irides, I 

 would argue that the image it perceives through water 

 and air is ill-defined, blurred, uncertain, and altogether, 

 in most cases, grotesque and awesome to the piscine in- > 

 telligence. Ergo, the fish is startled by any moving 

 object. 



A well-known optical law, which does not affect the 

 main argument, must now be referred to. Eonald, in 

 "The Fly-Fisher's Entomology," gives it and, as I do 

 not think that its importance is sufficiently recognized, I 

 reproduce it with grateful acknowledgements : When 

 Mr. A. B., situated upon a certain eminence at a given 

 distance from a fish, which is near the bottom of the 

 water, looks over the edge of a bank, in its direction, he 

 might, if unacquainted with the laws of refraction, 

 imagine that neither the fish, nor any other fish below 

 the line of his direct vision, could see him ; whereas, 

 the tish could see A. B. by means of the pencil of light, 

 bent or refracted at the surface of the water, and the 

 image of A. B. would appear in the eye of the fish, short- 

 ened and transferred to a much higher point. The fish, 

 in fact, could see the whole of the man round or over the 

 corner of the bank by the aid of the water above it ; 

 but if the surface of the water should be about as low as 

 the fish's eye, then he could not see any part of A. B.'s 

 figure, because a straight or unrefracted pencil of light 

 would be obstructed by the bank. 



Increasing obliquity in pencils of light falling from an 

 object upon a surface of water is accompanied by still 



