56 WALTER S. HUNTER AND JOS. U. YARBROUGH 



follows they were brought back to standard on handclaps: No. 

 7, 9 days; No. 15, 38 days; and No. 23, 47 days. They were 

 now retrained on going to the left for the buzzer and to the 

 right for silence. The intention was to train them upon this 

 for 60 days, unless the habit was established sooner, and then 

 test their retention of h. c. to the right. Rat No. 7 learned in 

 54 days, 540 trials; No. 23 learned in 35 days, 350 trials; and 

 No. 15 learned in 45 days, 450 trials. If we add to this only 

 the 300 trials which they had previously had on the same prob- 

 lem in the 30-day test, No. 7 learned in 840 trials; No. 15, in 

 750 trials; and No. 23, in 650 trials. 



At the close of the 40 trials at the standard percentage for 

 rats 7, 15 and 23 as just noted, they were retested on h. c. to 

 the right. No one of the three fell below 80% for 30 trials. 

 In other words, there was perfect retention. When given con- 

 trol 1 — tests made without the auditory stimulus — the per- 

 centages ranged between 30 and 50. On one day and with only 

 one rat did it go as high as 70%. So there could be no doubt 

 that the rats were dependent upon the auditory stimulus. Here 

 we have a case where two opposite habits are present simul- 

 taneously in the organism although the respective stimuli were 

 not originally differentiated. The process of the differentiation 

 has been a successive formation of habits and not a simulta- 

 neous one as is usual in discrimination tests. And the inter- 

 esting thing is that the formation of the second (and opposite) habit 

 has not interfered with the retention of the first habit. A second 

 automatism has arisen gradually and independently of the first. 



Further tests were made upon rat No. 7 to determine the 

 nature of the difference between the buzzer and the handclaps. 

 These results will be published in a separate paper. 



V 



Ninety-day Rats. — Three untrained rats, Nos. a, b and c, 

 were trained to go right for handclaps and left for silence. The 

 number of trials required in learning is shown in table 1. 

 At the close of this series, control 1 was alternated with 

 normal for three days in order to be sure that the animals were 

 not depending upon extra-auditory cues. The percentages were 

 all around 50. These three rats were then given a period of 

 idleness for 90 days. During this period, they remained in 



