26 THE BEHAVIOR OF LOWER ORGANISMS. 



Mendelssohn (1902, a, p. 406) misunderstands my ground for this 

 statement. He supposes that I hold that the Paramecia do not react 

 to differences in temperature less than that existing in a certain illus- 

 trative experiment, where one end of the slide was resting on ice, while 

 the other was heated to 40°. This experiment was purely for the 

 purpose of bringing the phenomena of thermotaxis concretely before 

 the attention of the reader ; its details had no special significance. I 

 have not the slightest reason for doubting the entire accuracy of the 

 quantitative experimental results set forth by Mendelssohn, and consider 

 them a most valuable addition to our stock of exact data. But the 

 calculation of the sensitiveness of the organisms concerned, from these 

 experimental results, involves a certain interpretation as to the reaction 

 method, and it was this interpretation that I called in question. Men- 

 delssohn, in accordance with his general theory, holds that the reaction 

 is due to the difference in temperature between the two 

 ends of the organism, and he calculates that this difference 

 in temperature could amount, in the case of Paramecia, 

 to but 0.01° C. According to the reaction method which 

 I have described above, however, it is not the difference 

 in temperature between the two ends of the same indi- 

 vidual that causes the reaction. Consider a slide cooled 

 below the optimum at the end a ; above the optimum at 

 the end b (Fig. 10), the optimum temperature for the 

 Paramecia being between the lines x and y. The animal 

 „ ^ ^ may swim a considerable distance from a position y^ at 

 one side of the optimum, to a position x^ at the other side 

 of the optimum, before it reacts (by backing and turning, etc.) at all. 

 We have no ground for maintaining then that it perceives any less differ- 

 erences in temperature than that between the lines x and jj/, and this 

 difference will be much greater than that between the two ends of the 

 animal. A similar diagram could be made for the case where the tem- 

 perature is raised or lowered only at one end of the slide. It seems to 

 me correct, therefore, that the sensitiveness to temperature differences 

 has probably been much overestimated. The only way that it could 

 be estimated would be by observation of individuals to determine the 

 extent of the stretch x-y over which they pass before reacting, and to 

 calculate the difference in temperature between the ends of this stretch. 

 It would of course be very difficult to do this with accuracy. 



Mendelssohn's view that it is the difference in temperature between 

 the two ends of the same individual that determines the reaction is not 



* Fig. 10. — Diagram illustrating conditions necessary for determining the sen- 

 sitiveness of Paramecia to differences in temperature. See text. 



