96 CORA D. REEVES 



b. Blue-red 



21. Four untrained but tame horned dace 

 were tested before a brilliant red and 

 duller blue plate in order to learn 

 whether these plates were effective 

 stimuli. There was an initial avoid- 

 ance of the bright red plate but this 

 soon ceased and the behavior of the 

 fish showed that both plates were 

 effective stimuli. (p. 26.') 



22. Four horned dace were given 420 trials 

 before blue and red plates with the 

 red varied in intensity so as to include 

 an intensity equal to the blue. They 

 were fed before the blue. The results 

 are plotted in the graphs, figure 13, 

 p. 31, in average percentage of correct 

 choices at different intensities of red. 

 At an intensity of the red obtained 

 by a slit-width of about 1 mm. the 

 percentage of blue choices diminishes 

 and the curve falls to near the 50% 

 level. This is believed to be the 

 region of matched brightness of red 

 and blue for the fish. As the trials 

 are continued with the red held at this 

 brightness, the percentage of blue 

 choices again increases as is shown by 

 the rise in the curve. With subsequent 

 lessening of the intensity of the red 

 the percentage of correct choices con- 

 tinues to be high. The fish then dis- 

 criminates the red and blue patches 

 at all intensities of the red. (p. 30 

 also figs. 13 and 14.) 



23. Three sunfish were given 680 trials 

 before a blue in contrast to a red plate 

 as in the preceding experiment; they 

 were fed before the blue. The graphs 

 showing the percentage of correct 



