62 FREDERICK S. BREED 



were introduced into the size training. Table 20 shows the 

 course of modification of response to the color stimuli. Only 

 the errors appear in this summary. 



TABLE 20 



Errors in Black-Blue Reactions 



Chicks nos. 32, 33, 38, and 39. Hatched 4/13, '08 



Group 1 32 33 38 39 



Sex, F. Sex, M. Sex, M. Sex, F. 



Series ' Date W W WW 



A Apr. 25 5 5 5 5 



1 " 26 6 7 7 7 



2 " 27 3 3 5 4 



3 " 28 1 3 3 2 



4 " 29 3 2 4 



5 " 30 2 1 2 



6 May 10 12 



7 " 2 1 1 1 



8 " 3 1 



9 " 4 



10 " 5 3 



11 " (i 



] 2 " 7 



13 " 8 



These chicks really do not vary so much in the modifiability 

 of their behavior as the figures might indicate. No. 38 became 

 excited during the tenth series, which may explain its lapse. 

 The rapidity with which no. 32 finished the work is noticeable. 

 No chick before or afterwards in my experiments excelled this 

 record of 30 trials exclusive of preference and perfect series. 

 But it is sufficient for our present purpose simply to show clearly 

 the perfection of the training that preceded the formation of 

 the second habit, -without extended discussion of the other 

 characteristics of the modification. 



Following the completion of this training of Group I, both 

 groups were introduced into a size experiment, as described 

 above. Table 21 reveals the progressive development of the 

 second modification for each group in the column of average 

 daily number of errors. The total of these averages for Group 

 I amounts to 66.3 for the sixteen days training; for Group II, 

 70.1 during the same period. Curves I and II, fig. 10, exhibit 

 graphically the rate of decrease of error for the corresponding 

 groups. The record of each group in the preference tests is 



