INTEGRATION OF MOVEMENTS IN LEARNING IN THE RAT 479 



rats when learning the inclined-plane problem, has led many to 

 conclude that the plunging of the plane in the first few trials is 

 accidental. The plane may for 3 or 4 trials be ascended or 

 stepped on without any evidence of a quickening of responses. 

 Such a view of the accidental character of the early trials is quite 

 in harmony with the one that all undirected movements are 

 " unsuccessful/' "useless," and have no direct or specifically 

 stated effect upon learning. The movements that are sup- 

 posed to be " accidental" in plunging the plane are, it must be 

 admitted, as undirected as the so-called " useless" and the lt un- 

 successful" movements, and no specific difference can be said to 

 exist in these movements. They must be regarded as prac- 

 tically the same in any scheme of "trial and error" that expects 

 the appearance of the "successful" movement. The records of 

 the first two rats presented do not admit the existence of "use- 

 less" movements or the "accidental" character of the first solu- 

 tion of the inclined-plane problem, but the "accidental" char- 

 acter of the first solution may, uncritically, be judged to be 

 present in the records of the other rats. However, if the move- 

 ments of these last rats are "accidental," when and how do they 

 become "non-accidental," or "useful," or "successful"? Evi- 

 dently definite changes must take place in some rats before 

 movements can at all be effectively produced. 



Whatever may be the interpretation of the early trials of 

 learning in the inclined-plane problem, there are features in the 

 solving or in the learning of other problems that are comparable 

 to it. In many instances in solving the latch-box problem, 

 several daily trials of 30 minutes each were required before the 

 problem was solved. This prolongation of the time in rats for 

 the solution of the problem solved occurred when reflex excit- 

 ability was low or hyponormal; with the inclined-plane problem, 

 which is invariably solved, there exists, when reflex excitability 

 is hyponormal, a prolongation of the time when effective re- 

 sponses can be made to facilitate directive integration, and not 

 for the solution of the problem, as is the case with the latch- 

 box problem. Different situations in the two problems, one at 

 the time of solving the problem, and the other in the early 



