42 VICTOR E. SHELFORD 



to turn back. In the ninth column, we note that the average 

 number of such trials varies from 1 to 3.2 It will be noted, also, 

 in the eighth column that the percent of the individuals tried, 

 which showed modification, varied from 33 to 100. In most of 

 the cases where no modification was indicated, the animals 

 reacted to the higher rate of evaporation from the first or remained 

 in* the air of low evaporating power without encountering the 

 higher; hence they do not indicate that the capacity for modifi- 

 cation is not present under favorable conditions. 



It will be further noted from the tenth and eleventh columns 

 that the length of time spent in the air of low evaporating power 

 is nearly always greater than the time spent in the highest 

 evaporating power (the time spent in the central third is omitted 

 and is usually small). 



The experiments were not planned to test the ability of the 

 animals to form associations. In the main, the work was con- 

 ducted in a manner intended to prevent the formation of asso- 

 ciations. There was a considerable stock of most of the animals 

 and in nearly all cases the individuals used in a given experiment 

 were not used again for some days. Still, after the gradient 

 had been encountered a number of times, turning before it was 

 encountered may indicate that the stimulation by the air was 

 associated with the approach to the light or with the screen- 

 covered side on the right (Fig. 1). In the case of Plethodon 

 cinereus there were ten more turnings before the gradient was 

 encountered than in corresponding positions in controls and 

 with the exception of Pterostichus, the number of turnings in 

 the experiment before entering the gradient exceeded that for 

 the corresponding positions in the controls. 



To test the possibility of association formation, four readings 

 (A, B, C and D, Experiments 42 and 43, Chart IV) of the 

 behavior of three toads were made. "A," (Chart IV) is a typical 

 symmetrical control which was observed first. The same toads 

 were then transferred to the experimental cage, and reading 

 begun within five minutes. Here they began showing some 

 avoidance of the air of the high evaporating power at the end of 

 six to eight minutes. At the end of a twenty-minute observation, 

 they were removed for a few (less than five) moments and then 

 returned to the same cage (C) where they showed avoidance of 

 the dry air in one and two minutes. At the end of the twenty- 



