28 CORA D. REEVES 



At 1:30 P. M. the curtains were drawn to exclude all 

 extraneous light from the experiment aquarium, and the 

 ceiling light was turned on. At 2 P. M. the Nernst glowers 

 were lighted. By the method already described the fish 

 were brought into the discrimination compartment. The 

 slit was not used, so that the brilliant red patch and the 

 saturated blue were presented. The slide door was raised 

 a short distance. The two dace were swimming about 

 the top of compartment 2, figure 2, and did not appear 

 to note the opening. At 3:21 P. M. the slide door was 

 again opened. H went into the blue lighted area, and 

 although the food-bar was released and he remained near 

 the end of the bar with its piece of worm, the food was 

 not touched. His path is shown in figure 11. The ap- 

 proach to the red and then the straight path across in front 

 of that plate appear to show an inhibition of further ap- 

 proach to the bright red area. At 4:20 P. M. the door 

 was again raised and both dace swam out in the wavering 

 paths characteristic of exploring fish. One went to the 

 blue lighted area and fed. H was then closed into the 

 retention compartment, while the fish Md was tested. 



A typical record (fig. 12) follows: 



P. M. 



4:53 Slide door opened. 



4:54 Swimming back and forth in the discrimination compartment; pauses fre- 

 quently with head in midline. 



4:58 Pauses on either side of the midline. 



5:00 Went out to blue area and up into the direct blue light; stayed there a time; 

 fed. 



In this preliminary work H was next tested. The first 

 thirteen out of fifteen responses were to the blue. When 

 Md was again tested thirty-eight out of forty responses 

 were to the blue. 



For the next ten trials of H the slit was used, so that 

 the light intensity of the red was reduced to 66, 37, and 

 20 per cent of its maximum value. The fish went to the 

 blue area nine out of ten times with this shifting intensity 

 of red. It appeared that there was a decided tendency 

 to approach the blue rather than the red. It was now 

 evident that this was true even when the bright red was 

 reduced in intensity. 



