LIGHT OF DIFFERENT WAVE-LENGTHS BY FISH 25 



red plate without the slit was for the fish in some way 

 equal to white light through a 0.1 mm. slit, or the behavior 

 was a chance response. 



I decided to see whether the fish could discriminate 

 between the two light patches that had apparently in- 

 duced the rising action and were presumably equated in 

 brightness, i. e., the red without the slit and the white 

 through the 0.1 mm. slit. I tested them with the red on 

 one side and this very dull white on the other. The fish 

 which had given the best response to the white continued 

 to go to it and to feed before it as I shifted the plates from 

 side to side, though the brightness of the white presented 

 was now presumably reduced to that of the red. The 

 other fish gave slight evidence of a " position habit," but 

 on the second day responded accurately to the reduced 

 white. The width of the slit used with the white was again 

 changed to 5 mm., then to 2 mm,, then to 4 mm. Eighty 

 per cent of the choices continued to the white. There 

 seemed little possibility that brightness was controlling 

 the behavior when a continually shifting brightness for 

 positive stimulus was presented. The avoidance of a red 

 plate at first and the ability to distinguish it from the 

 white plate, even when the latter presumably matches it 

 in brightness, is a more consistent explanation. The shift- 

 ing brightness of the white plate did not prevent discrim- 

 ination. That this red plate had a very definite value 

 and a very different one from any dull white, was shown 

 also by a tendency on the part of the dace to avoid the 

 red illuminated area as compared to their indifferent ap- 

 proach toward the dimly illuminated white. But this 

 initial tendency to avoid red was very brief on the part 

 of the dace. It is doubtful whether such a tendency would 

 be shown by all individuals. 



3. Wave-Length Discrimination: Blue-Red by Dace 



a. Red chiefly at constant maximum intensity. 



To test the power of dace to discriminate longer and 

 shorter wave-lengths, the apparatus already described 

 was used with both the red and blue filters in place. By 

 this means two patches of light were presented to the fish, 



