56 CORA D. REEVES 



failure to discriminate with return of discrimination as 

 the tests are continued at this value. This temporary 

 lapse of discrimination and its subsequent recovery are 

 held to be evidence of matched brightness of the stimulus 

 patches. 



The evidence for discrimination of the stimulus patches 

 by a quality difference between them is quite uniform and 

 is most readily appreciated in the graphs figs. 13 and 15. 

 A fuller discussion of the whole question is reserved for 

 the final section of this paper — but attention is here called 

 to the case of failure to discriminate at matched brightness 

 that appeared in the training experiments. In this case 

 the slit was removed from above the red filter and was 

 replaced by solutions which reduced the intensity of the 

 red to that of the blue. The result was a change in the 

 distribution of the light on the red plate. The dace failed 

 to discriminate. (Table I.) Two explanations of this 

 failure are suggested. Either with matched brightness of 

 red and blue discrimination must be maintained by con- 

 stant use when the red has become familiar, and this had 

 not been done, or the fish in this case were tending to go to 

 the modified red plate, changed enough in appearance by 

 the solutions to awaken their " curiosity." As Bauer 

 (1910) has shown, they often go toward an unusual object. 

 Of course, there are those who will be apt to call attention 

 to the fact that here a psychic factor is resorted to in order 

 to account for what will appear to them to be proof of 

 inability to distinguish differences in color as shown in 

 response of a particular fish. It appears, however, that a 

 psychic factor is continuously resorted to in this problem; 

 namely, brightness discrimination. The change in the 

 appearance of the red plate by the use of solutions modified 

 response, in that after the first responses another psychic 

 factor dominated behavior. Bi;t Table I shows that all 

 first responses, even with a modified red plate, except that 

 of 552, were to the blue. After a bite or two, hunger did 

 not inhibit the tendency of dace YP and Yl to go fre- 

 quently to the slightly modified red plate. In view of the 

 uniformly positive results obtained in other experiments, 

 little importance is attached to this single failure and it 



