SEVENTH ANNUAL MEETING 55 



from warm to cold. This was indicated by the fact that to 

 produce successive recurrences of the paroxysm, the temper- 

 ature change had to be made successively larger or the fish 

 left for a successively longer time in the warmer water. 



2. Reaction of Fishes to Temperature : Under this head 

 are included a number of experiments which were performed 

 by putting the fishes into a long narrow tank arranged so that 

 the water flowing in at one end was of a different temperature 

 from that flowing in at the other. The water of the two tem- 

 peratures flowed to the middle of the tank and thence out 

 through drains at top and bottom. The result was a mixing 

 of the water, especially of the center third so that a temper- 

 ature gradient was formed. This gradient was accurately de- 

 termined by testing with thermometers graduated to tenths of 

 a degree C. The fish was introduced at the center of the tank 

 and a graph (Fig. 1) of its movements was made according to 

 the method first used by Shelf ord and Allee ('13). A glance 

 at the figure, which is that of a typical graph, will make the 

 method and results clear. The graph is selected as typical of a 

 large number of similar ones ; it shows that the fish in the ex- 

 periment reacted very definitely to the temperature gradient, 

 selecting in this case the warmer end. Furthermore it will be 

 noted that most of the turnings back from the colder portions 

 of the gradient occurred some time before the fish had en- 

 countered the coldest water. Measurements showed that the 

 fish detected and reacted to variations in temperature of no 

 more than .1° C. All species tried were equally sensitive. 

 Other experiments showed that fishes tend to select an optimum 

 temperature (16°-19° C.) for they will turn back from warm 

 water when it is above this temperature. No attempt was 

 made to alter the optimum temperatures of the species used but 

 it is probable that the optimum varies with the physiological 

 state. 



Discussion : The above experiments suggest that fishes pos- 

 sess a temperature regulating mechanism of the most delicate 

 order. It has been shown that practically all cold blooded 

 animals maintain body temperatures above that of the sur- 

 rounding medium. Kidder (1879) showed that certain marine 

 fishes maintain temperatures above that of the surrounding 

 water (5°-20° C for different species). There must therefore 

 be present in these forms, just as in warm blooded animals, a 

 mechanism for temperature regulation. Furthermore, the 

 convulsive activity at high temperatures and upon sudden 



