INTEGRATION OF MOVEMENTS IN LEARNING IN THE RAT 403 



cent, on the fourth day in 9.3 per cent, and on the fifth and sixth 

 days in 3 per cent of rats. This per cent estimate includes cases 

 in which the latch was raised by the teeth instead of by the 

 neck reflex. Only in 3 per cent of rats were the teeth used. 

 Learning was possible in nearly all cases when the neck reflex 

 movement w r as used, with the exception of a solution occurring 

 on the fifth day. Though learning was possible when solution 

 in one rat occurred on the sixth day, many slight imperfect re- 

 sponses occurred before the latch, which were not considered in 

 the general estimate when learning was complete. From the 

 first to the sixth day, there was a decrease in the per cent of so- 

 lutions, and with the rats that solved the problem between the 

 first and the sixth day, there was generally an increase in the 

 number of trials to complete learning. 



The descriptive account of the observations made on rats and 

 the results presented in table 2 do not determine what conditions 

 both the observations and the results obtained. However, 

 something does seem conclusive from the observations of the 

 behavior of rats, namely, that reflex excitability is an essential 

 qualification in establishing an interaction of body parts for pos- 

 turing under the latch to produce the neck reflex movement ; and 

 when posturing can not be produced, the mechanism for the 

 production of reflex excitability appears to be undeveloped. 

 But this does not mean that posturing of the body can not at 

 any time take place when this manifestation is not in evidence, 

 for ineffective posturing may occur with no evidence of reflex 

 excitability manifesting itself before or after posturing. The 

 effectiveness of posturing and every response is markedly in- 

 creased with the presence of excitability. Another condition, 

 however, determines the rapidity of learning, and this is the 

 number of imperfect responses made. There exists the greatest 

 difference in rats in the number of such responses made. When 

 learning is completed in 19 to 25 trials, few if any imperfect 

 responses are made. From the third or the fourth trial pro- 

 gression is then direct to the latch and the neck reflex to raise the 

 latch is easily produced. When in one rat 19 trials were re- 

 quired to complete learning, direct progression to the latch and 



