COLOR VISION IN FISHES 485 



still higher and remained high. She felt justified in concluding that the 

 intensity of red at which confusion at first occurred was one at which 

 the brightnesses of the red and blue lights were equated for the dace, 

 and that the confusion was due to a re-learning, the fish switching its 

 attention from the brightnesses to the hues and making the association 

 'blue = food' which she had supposed to have been made in the first 

 place. Another dace showed no such temporary confusion at any inten- 

 sity of the red, indicating that it had been attending to hue from the 

 first, rather than to brightness. Of course there were two other possi- 

 bilities — that the fish, from the outset, were not going to blue so much 

 as avoiding red, with individual differences in this red-avoidance; or, 

 that the confusion was due to getting used to red and ceasing to avoid it, 

 the 'blue = food' association being not yet established. But, interpreted 

 in any of these ways, the experiment had demonstrated hue-discrimin- 

 ation, for there could be no redness-fear in the first place unless there 

 were redness-perception. 



The sunfish (Lepomis) proved to be more sensitive to intensity than 

 the dace, tending to lurk and hide in dark corners rather than come out 

 and face white lights to which the dace had readily gone. In keeping 

 with this. Reeves found a greater capacity for discriminating intensities, 

 the ratio needing to be but 1 : 2+, which however was still far short of 

 the 1 : 1.23 difference claimed by Hess — and yet was far better than the 

 capacity elicited from any other species by any other investigator before 

 or since. When offered blue versus red, the sunfish was extremely slow 

 to build up the 'blue = food' association; and again it was the larger 

 specimen which showed red-shyness, the younger specimen lacking fear 

 of red as also of many other things which would startle the older animal. 

 After successful training positive to blue, the gradual dimming of the 

 red stimulus to a certain value caused the same sudden confusion, at 

 that critical intensity, which had been manifested by the dace — the 

 intensity this time being, by coincidence certainly, the very one at which 

 the red and blue were identical to Reeves' own dark-adapted eye (the 

 fishes were Ugh t-a.dsipted by an initial white illumination before each 

 trial). The sunfish recovered from the confusion at matched bright- 

 nesses much quicker than had the dace, however. Both species readily 

 discriminated the training-blue from 'gray' light (produced with several 

 layers of photographic negatives as a filter) which matched it in bright- 

 ness for the human; but they behaved very differently from what they 

 did when a blue and a red, also of equal brightness to the human, were 



