LIGHT OF DIFFERENT WAVE-LENGTHS BY FISH 55 



During these tests the intensity of the blue Hght was 

 changed so that there was added evidence that the choice 

 of blue by the small sunfish (p. 42) was not merely a re- 

 sponse to light of a given constant intensity. But the 

 disturbances in the behavior of Large Sunfish and YP 

 did not occur when the blue intensity was changed, hence 

 cannot be attributed to that cause. Since the slit was 

 used in a part of these trials on the blue side only and in 

 a part of them on both sides, the series affords additional 

 evidence that the position of the slit did not guide the fish 

 in their choice (see p. 46). 



G. Results of Experimental Work with the use of Food- 

 Association. 



In the foregoing experiments with stimulus patches of 

 different wave-length it is believed that the responses of 

 the fish were made to the light on the stimulus plates and 

 were not dependent on any other part of the apparatus 

 nor on its manipulation. The radiant energy of the plates, 

 as measured by physical instruments does not appear to 

 have been a controlling factor. The fish were therefore 

 responding either to the wave-length difference between 

 the plates or to a brightness difference. At one point in 

 the series the colored plates were matched in brightness 

 for the human dark-adapted eye and the behavior of the 

 fish indicated that they were matched in brightness for 

 them. Since it is possible that fish have a capacity to dis- 

 criminate very small intensity differences between stimulus 

 patches of like quality, a capacity even greater than that 

 of man, it became necessary to test them with respect to 

 this. When thus tested the dace failed, under the condi- 

 tions of the experiments, to learn to discriminate intensi- 

 ties of white light which differed as 1 :4 while sunfish did not 

 discriminate intensity differences of 1:2. As it is certain 

 that the intensity difference between the colored patches 

 discriminated by the fish are far less than these, it follows 

 that this discrimination is based on a difference in the 

 wave-lengths of light of the two plates. 



At the critical value of the red variable (slit width of 

 about 0.9 mm. in most of the graphs) theVe is temporary 



