LIGHT OF DIFFERENT WAVE-LENGTHS BY FISH G7 



exercise discrimination in the choice of their food. Gold- 

 smith (1914) has shown that when the fish are very small 

 they do not chase their prey. The need for discrimination 

 grows as the fish develop and change their food habits. 



(2) The experiment in which the fish are found to show 

 no differential response toward red and blue when the 

 relative brightness of the two is suitably adjusted, is taken 

 as evidence for the lack of wave-length discrimination. 

 The fish are here not responding to a difference in wave- 

 lengths of light, but it does not follow that they are unable 

 to respond to such differences. The experiment shows only 

 that under the conditions they do not respond differentially 

 to the wave-length differences. Training experiments 

 might show a capacity not revealed by this method. It 

 is further possible that during the time required for ad- 

 justment of the blue and red lights the fish became accus- 

 tomed to the red so that they no longer avoided it. When 

 the adjustment was completed they might then distribute 

 themselves equally in lights of the two colors. My own 

 experiments contain much evidence of such rapid " getting 

 used " to red. In such work on response to light of different 

 wave-length it is the initial responses that have greatest 

 value. 



(3) Hess' experiments show that the fish do not gather 

 in the red of a spectral band, and this is taken to mean 

 that the red has little stimulating power. The experiment 

 could, however, be explained by supposing that the red 

 affords a stronger stimulus than other rays and is avoided. 

 My own observations as well as those of Reighard, Bauer, 

 Mast, and Goldsmith show an avoidance of red. 



(4) The experiments in which Hess determines the rela- 

 tive brightness of different parts of the spectrum for these 

 young fish in the dark is of interest but seems to me to 

 have no bearing on the question of wave-length discrim- 

 ination. That the brightness values of the different colors 

 are the same for fish as for the color-blind may be thus 

 established. Hess' work with the pupilloscope is confirma- 

 tory (Hess 1917). My own work adds to this evidence. 

 It does not necessarily follow that fish are color blind. Von 

 Kries (1896) shows that for the periphery of the human 



