20 



VICTOR E. SHELFORD AND W. C. ALLEE 



and 9 from top to bottom, we note that generally speaking the 

 less sensitive fish show the greatest number of trials of the modified 

 water before giving evidence of modification. In columns 10 

 and 11, we note that the percentage of individuals showing modi- 

 fication is greatest in the more sensitive species and that when all 

 are reduced to terms of 100 per cent, of modification the most 

 sensitive species show the smallest number of trials before turning. 

 The time in the two kinds of water is variable, but usually 

 greatest in the unavoided water. 



Reading the lines from left to right we note that the sensitive- 

 ness of the fishes is different for different stimuli. 



TABLE 3 



A list of the species used; the size of the individuals; physiological relations 

 to the factors used; number of turnings and time spent in each half before modi- 

 fication was apparent. 



