INTERFERENCE OF HABITS IN THE WHITE RAT 51 



the distribution curve. All but six of the rats had mastered 

 the problem within 500 trials. I am inclined to attribute the 

 irregularities largely to position habits which appeared during 

 the learning and which had to be overcome. Fear caused by- 

 punishment retarded the last part of the learning in rats 25-29. 

 No sex differences appear. The form of the learning curve will 

 be shown in section VIII. 



TABLE 1 

 Number of Trials per Rat in Learning First Habit 



Each rat of set A, when it had completed its four days with 

 a general average of 87 J%, was given three controls, each of 

 which in general alternated with a day on the normal stimulus 

 of handclaps. Control 1 — no stimulus was given. The reac- 

 tion was counted correct if it agreed with the series of presenta- 

 tions. This control was to test the animal's dependence upon 

 extra-auditory cues. Control 2 — an electric buzzer was sounded 

 in place of the handclaps. This control was given to each rat 

 on each of four successive days (40 trials). Control 3 — a tuning 

 fork, 256 d. v., placed over the apparatus as described in pre- 

 vious papers, was sounded by striking with a felt hammer. This 

 tone was substituted by the experimenter for the handclaps. 

 The necessity for the first control needs no comment. Con- 

 trols 2 and 3 were used in order to determine the relation of the 

 respective stimuli to the habit just established. It is very im- 

 portant, if interference effects are to be studied, that the mutual 

 relations of the stimuli (i.e., their transfer relations) be known, — 

 in the present case, whether the buzzer and the tuning fork 

 would be substituted readily for the handclapping in this par- 

 ticular co-ordination. 



The results of control 1 indicate that the animals were depend- 

 ing upon auditory cues. Only one rat's (No. 17) record was 



