MODIFICATION OF THE BEHAVIOR OF FISHES 17 



the fishes remained in one end for thirty minutes and then (see 

 below graph of Expt. 13) began to go back and forth more slowly 

 than in the experiment. Two turnings at the center are shown 

 by the control fishes and these are opposite in direction. This 

 shows that the fishes sometimes turn in the absence of solutes. 



A comparable result with boiled tap water is recorded in chart 

 2, Expt. 78 (control). Here the turnings are less prominent 

 and shortened stays in the boiled water were followed by resting 

 in the tap water. In the part of Expt. 78, where the amount of 

 oxygen was the same at each end of the gradient there is no evi- 

 dence of modification, but simple disturbance is indicated. This 

 shows that the fish sense the effects of boiling even when there is 

 a normal amount of oxygen in the boiled water. In chart 3, 

 Expt. 56, very rapid modification is indicated, and after the first 

 few trials of the modified water the fishes stayed in the tap water 

 end, except when driven out (indicated by arrows). In the 

 same chart, Expt. 66 shows an apparent modification due to 

 the introduction of an atmosphere of nitrogen and oxygen. This 

 atmosphere had an oily odor which may have affected the results 

 obtained. 



Expt. 13 of chart 1 and Expt. 83 of chart 2 show the type of 

 mild negative reaction with indications of slight modification 

 given by the sunfishes and basses tried. The graph of Expt. 53, 

 chart 2 shows the symmetrical type of control most common 

 with Notropis and a case of somewhat erratic action in very 

 strong carbon dioxide after it has time to seriously affect the 

 fishes. 



The graphs of Expts. 41 and 36 of chart 1 show the reaction 

 of Abramis to acetic acid and ammonia in boiled water. On 

 encountering the acetic acid the fishes often gave a definite 

 reflex — the " backing-starting " reaction to be described on page 

 24. The reaction to acetic acid in boiled water compares very 

 favorably with that to carbon dioxide in boiled water. Such a 

 reaction is shown in Expt. 84, chart 4. The ammonia experiment 

 shows the failure of the animals to react negatively to this factor, 

 although the backing-starting reaction occurred more often 

 than in the acid. The fishes died in the ammonia. 



The case of Abramis demands special attention as this is the 

 only species studied in which the groups of individuals developed 

 apparent preference for unknown reasons. In this species this 



