LIGHT OF DIFFERENT WAVE-LENGTHS BY FISH 53 



delayed response clearly showed that the change from red 

 to gray light of presumably equal brightness was effective 

 in modifying behavior. The dace Md showed the same 

 modifications of behavior, but the dace H did not. Just 

 before these tests were made, this fish had been presented 

 once with a dull red and gray. After going up to the gray, 

 he shied off from it as if frightened. It was plain that the 

 gray stimulated. The lack of uniformity in the response 

 made by the different individuals at any one time and 

 shown by these fish is characteristic of fish. 



F. Tests with Energy Equated. — -It has been held (Laur- 

 ens 1911) that animals are able to differentiate between 

 various light stimuli, not by difference in luminosity or 

 subjective value as measured by the eye, but by difference 

 in radiant energy, i. e., by mechanical value as measured 

 by physical instruments. The possibility, therefore, that 

 the radiant energy differences of the stimuli might account 

 for the discrimination already described, must be tested. 

 If the radiant energy reaching the two stimulus plates 

 were made equal, and the fish, especially the sunfish, could 

 still discriminate, then the possibility that they discriminate 

 by radiant energy would be eliminated. In order to equate 

 the energies on the stimulus plates a bismuth-silver thermopile 

 was used (Coblentz 191vS). It was placed in the center of 

 a heavy glass container, fitted into a cardboard support. 

 This made it possible to place the thermopile beneath 

 the lenses above the edges of the stimulus plates. 

 The thermopile was connected with a sensitive galvano- 

 meter. On one side were placed a water-cell, a slit, and 

 a new blue filter; on the other was the red filter and a solu- 

 tion of potassium sulfo-cyanide and ferric chloride. The 

 energy passing through the red filter was altered by ad- 

 justing the depth of the solution above it until the gal- 

 vanometer deflections were the same when the hi-ag ther- 

 mopile was shifted from one side to the other. The energy 

 of the two plates was also equated by means of a slit on 

 the red side, while a water cell and slit were used with the 

 blue. The relative brightness of the energy-equated plates 

 varied for the human eye according to the method used. 

 When the red solution was used, the red plate was some- 



