32 



VICTOR E. SHELFORD 



second scale at the time of the observation (Fig. 1) and those 

 presented here are original drafts. The different individuals 

 could not be distinguished in a few cases, but since this error 

 entered into all the experiments the results are constant for this 

 piece of work. (For further discussion of this point, see Shelf or d 

 and Allee, '14, p. 8). 



III. EXPERIMENTAL RESULTS 



To illustrate the method of obtaining and recording the data 

 used further on, we present three charts. In Chart I, Expt. 71, 

 the reaction of the red-backed salamander to increased evapo- 

 ration in the right hand third of the experimental tank, is shown. 



Figure I 



Showing the ground plan of the experimental cages in their relative positions, 

 the hood which covered and separated them is not indicated. E is the experimental 

 cage; W, the section used for wet air; M, for the air supplied directly from the 

 pump; D (dry), H (warm), and R (rapid flow), stand under the section where the 

 highest rate of evaporation was maintained. The crosses indicate the positions of 

 the 1 c.p. lights; the arrows the direction of the flow of air. The screen portions 

 of the cage are represented by the broken lines. C, is the control cage, similar to 

 the experimental in every way except the kind of air supplied. Below this is shown 

 the control record of an experiment during the first three minutes. The ruling 

 of the paper used corresponded to half minutes and the figures were written in at 

 the center. The graph is about two-thirds natural size and the cages about one- 

 seventh. 



The increased evaporation is due to a rapid flow of ordinary air. 

 The salamanders tried the region of highest evaporation re- 

 peatedly during the first fifteen minutes and then began to turn 

 back when the rapid flow was encountered. We note that one 

 individual turned back the first time it encountered the rapid 

 flow. The control individuals came to rest practically where they 

 were placed and moved only a little throughout the experiment. 



