LIGHT OF DIFFERENT WAVE-LENGTHS BY FISH 79 



inhibition by other stimuli (delayed response in sunfish 

 p. 50). For these reasons a negative conclusion based wholly 

 on evidence from the method of response seems unwar- 

 ranted. The method may answer for lower organisms, 

 but animals as highly organized as fish may be quite capable 

 of responses that do not appear under it. Those who use 

 it need to take particular pains that they are not relying 

 upon innate responses which have disappeared through 

 adaptation or which are inhibited by uncontrolled stimuli. 



If on the other hand, when positive results are obtained 

 by the response method (Reighard, Bauer), they appear to 

 have the greater value. Here, at least, it is certain that 

 response is not lacking through adaptation to the stimulus 

 or through inhibition. Positive results are obtained when 

 the fish show differential response toward light-stimuli 

 of different quality at all intensities of the variable used. 

 Owing to the evanescence of the response (adaptation) 

 it is not practicable to repeat the experiments at many 

 intensities of the variable. It is, therefore, possible that 

 matched brightness for the fish is neier secured by this 

 method. But if, at various intensities of the variable, 

 the fish continue to show a differential behavior toward 

 the two stimuli of different quality and do not show this 

 toward stimuli of the same quality at presumably like in- 

 tensities, then we have evidence of color vision. Such 

 evidence we get, for instance, from the peculiar behavior 

 of fish toward red obtained by Reighard, Bauer and myself. 

 Hess' failure to obtain similar differential behavior is merely 

 la,ck of evidence, and as already pointed out, permits of 

 other explanation than inability of fish to respond differ- 

 entially to red. 



b. The method of training gets rid of the difficulty in- 

 volved in evanescence under the response method, for it 

 insures response over long periods of time and makes it 

 possible to use the variable at many intensities in the 

 effort to secure matched brightness At the same time 

 the fact that the fish take food during experimentation 

 shows that their responses are not inhibited. In work that 

 has been done hitherto on color vision of fish by training 

 methods, three procedures have been used in the attempt 



