80 CORA D. REEVES 



to determine whether the fish were guided by brightness 

 differences rather than by quality of light. 



In the first of these procedures (Washburn and Bentley) 

 no attempt is made to secure matched brightness of the 

 two colored objects presented to the fish as stimuli. One 

 of them is kept constant in brightness value and is pre- 

 sented to the fish in contrast to the other colored stimulus, 

 which is used at two values, one brighter and one duller 

 than the first. The relative brightness of the two stimuli is 

 judged by the human eye either light- or dark-adapted. 

 The method is probably adequate to insure the end sought, 

 namely, that the variable stimulus be used both distinctly 

 brighter and distinctly duller than the constant. When, 

 under these conditions, the fish learns to avoid the con- 

 stant stimulus in contrast to both a brighter and a duller 

 variable, he chooses in the one case the duller object and 

 in the other case the brighter, and it has been assumed 

 that the discrimination is, therefore, made by means of 

 a quality difference between the two stimuli. Since con- 

 siderable brightness differences for the human eye always 

 exist under this procedure, it is possible that they exist for 

 the fish eye also, and it follows that the fish may discrim- 

 inate the constant stimulus from the variable by brightness 

 difference only. A totally color blind person might so dis- 

 criminate them by avoiding the stimulus of definite con- 

 stant brightness, with preference for one either brighter 

 or duller. 



The second training procedure referred to above attempts 

 to make the two stimuli of equal brightness for the fish. 

 To this end a colored stimulus object is matched with 

 another colored object or with a gray by means of the 

 human dark-adapted eye. The two objects may be iden- 

 tical except for color and may be offered to fish as food, 

 or one may be offered as food and the other form a back- 

 ground against which the food is seen by the fish. If the 

 fish learns to distinguish the two objects assumed to be 

 matched in brightness for it, the conclusion is drawn that 

 quality differences in the light are the basis of discrimina- 

 tion. However probable it may be that colors matched 

 in brightness for the human dark-adapted eye are also 



