LIGHT OF DIFFERENT WAVE-LENGTHS BY FISH 61 



and red were again used. When the door was raised the 

 fish stood in the doorway three minutes, then went for- 

 ward an inch into the stimulus compartment and backed 

 away. It next turned straight toward the red plate and 

 darted forward to within 20 cm. of it, then drew back 

 and darted up to the blue stimulus patch. It is to be noted 

 that the waiting in the discrimination compartment of 

 the untrained minnows and their circling across the door- 

 way in the presence of the red-blue patches are exactly 

 the sort of behavior that the dace, habituated to blue-red 

 stimulus patches, gave when the blue-gray was presented 

 to them. When the matched blue and gray were presented 

 to the untrained minnows, all the fish tested swam promptly 

 to these plates. They went an equal number of times to 

 each plate. They often swam from one plate to the other. 

 I could detect no differential behavior toward either, but 

 these same untrained minnows, when presented with the 

 red and blue, often remained at a distance from the red 

 or turned from it or swam vigorously toward it. In other 

 words, the introduction of red in place of gray brought 

 about a differential response in those untrained fish. This 

 took place when no part of the apparatus was modified in 

 appearance except the stimulus patch. Without the evi- 

 dence of the long blue-red training series the tendency of 

 fish to change their behavior when the red plate is first 

 exposed might be attributed to brightness differences, 

 although the plates matched in brightness for human eyes. 

 But in the red-blue food association series with the variable 

 red intensity, the least discrimination was shown by ex- 

 perienced fish when the red and blue first approached 

 matched brightness for the human eye. Therefore, if 

 untrained fish showed a differential response when the 

 red plate was first exposed at this matched brightness, 

 this could scarcely be due to its brightness. 



When untrained fish were presented with blue-gray 

 patches that matched for the human dark-adapted eye, 

 they showed no preference for either plate and gave no 

 indication that the two patches were in any way different 

 for them. Therefore, it seems clear that a difference in 

 brightness was not the reason for the peculiar behavior 



