74 CORA D. REEVES 



made accessible, and had he then offered the fish, along 

 with the blue, masses of equal brightness but of other colors, 

 and found all of them refused, he might then have concluded 

 that the fish did not discriminate them by color. What 

 he did was to establish an association with yellow by feed- 

 ing the fish on it. He then offered this yellow along with 

 blue, but without showing that the blue had been made 

 immune. The fish showed no differential choice, and there 

 was no reason to expect it. 



This experiment, used to prove the lack of ability on the 

 part of fish to discriminate on the basis of wave-length, 

 illustrates the difficulty of trying to prove a negative. 

 Fish are keen in detecting movement. There was appar- 

 ently no severe bumping against the glass when blue bits 

 were falling through the water with only a plate of glass 

 between them and the fish. The accessible yellow meat 

 dropped near the plate satisfied the hunger of the fish. 

 The yellow food served only to form the association " fall- 

 ing bits-food. ' But the inhibition against reacting to blue 

 was not established. 



Hess (1913) argues against the view that the red color 

 of the belly of the breeding " Saibling " may be observed 

 by other " Saibling " and is therefore an indication of color 

 vision. As evidence that red rays do not reach the spawn- 

 ing grounds of the brilliantly colored Konigseesaibling he 

 tested the penetration of red rays into water by a simple 

 experiment. He used a brass tube 4 m. long. On the side 

 of this at one end was a window 4 cm. in diameter, and 

 within the tube facing the window was placed a mirror 

 set at an angle of 45° with the axis of the tube. Attached 

 to the lower end of the tube opposite the window and at a 

 distance of 20 cm. was a red plate 12 by 20 cm. which could 

 be set at various angles to the window. The end of the 

 tube to which the red plate was attached was immersed 

 in the water to a depth of 4 meters, and the reflection of 

 the plate was observed in the mirror through the upper 

 end of the tube. When thus observed, the plate appeared 

 yellow or brown but not red. I have repeated the experi- 

 ment in air in diffused daylight. When the light is suf- 

 ficiently reduced, the results are as Hess states. This 



