HABIT FORMATION AND FEELING QUALITIES 305 



and to the 5 of C in the other. The movement from the J of H 

 was confused by some with the movement from the J of C. The 

 subject found that these confusion points could be overcome, 

 provided the movement was recognized in its proper setting and 

 tune relations. Until the distribution was well mechanized, 

 skill in passing the critical points depended upon the readiness 

 to anticipate and discriminate between the rival movements. 



4. Feeling tone accompanying the distribution, a. Artificially 

 created. The first reports of the subjects, contrary to anticipa- 

 tion, uniformly show a comparatively slight disturbance from 

 the stimulus. Even the unpleasant stimulus was mildly amus- 

 ing, to be forgotten as soon as the distribution began. Such 

 reports as "The stimulus did not amount to any thing/ ' "I for- 

 got about the stimulus," or "It does not disturb me in any way," 

 were often made. To sum up the matter for stimulus effects : 



(1) P stimulus is for the most part neutral, though there is 

 some evidence that it lessened the number of errors, of dropped 

 cards and of hold-ups. 



(2) The stimulus effects were more noticeable in periods of 

 high records, and in the case of the U stimulus increased the 

 number of hold-ups. 



(3) The U stimulus served both to increase and decrease the 

 distributing rates, the former by adding momentum to the 

 smoothly running movements and the latter by prolonging the 

 relapsed condition of a movement or of the place memory for a 

 box. "If I stumble the stimulus tends to keep me down, but 

 as long as I keep up and going it increases the speed." 



(4) The subjects who took stimuli, whether P or U, made 

 more errors hi the non-stimulus periods than in those in which 

 stimulus was given, and the subjects taking P stimuli dropped 

 more cards and had more hold-ups in non-stimulus periods, 

 while subjects taking U stimuli dropped more cards and had 

 more hold-ups during stimulus periods. This shows for the 

 most part that the subjects did better during the stimulus periods, 

 a slight exception being with subjects taking U stimuli. See 

 table 5 in this connection. 



