x SOME EXPERIMENTAL RESULTS 239 



made preparations for an experiment, even though it was 

 of a new kind. This was very plainly marked in their 

 disappointment with the " control " experiment which I 

 tried without food. 



9. The normal process of learning. 



Whatever the basis of learning, whether accident, 

 imitation, or other, the normal process was clear enough. 

 In each case the central point was what I have called the 

 critical success. In nearly all cases this was also the first 

 success. Sometimes,indeed, success was arrived at in stages. 

 The animal would fairly attempt the right method without 

 succeeding ; or perhaps, though having once succeeded, it 

 would require a reminder before succeeding again. But 

 seldom would it fall back from partial success to complete 

 failure. In other cases success was heralded by a kind of 

 tentative action, as when the fox-terrier or the cat slightly 

 pawed the vase without seriously trying to overturn 

 it. Finally, after success, it was not uncommon to have 

 a break-down in one, and, more rarely, two trials of the 

 same series. Generally this happened when the series 

 had been prolonged, and it gave the impression that the 

 animal's powers of concentration had been exhausted by its 

 successful effort. There is no doubt that the process of 

 learning what to us appears so simple is to them attended 

 with great effort. It was seldom of any use to prolong a 

 lesson beyond a dozen, or, at most, twenty trials ; and the 

 dog and cat in particular would seem, at the end of such a 

 series, quite exhausted, and would stretch themselves out 

 for a good sleep, after half-an-hour of which they would 

 wake up, and do very much better. 



Many observers credit animals with apparent "memory " 

 of particular events, and, rightly understood, this view 

 coincides with the conception to which we have been led 

 of the learning process. We cannot, indeed, from any 

 external action infer that an animal makes a memory- 

 judgment in the strict sense, i.e., is aware of an event as 

 having taken place at a certain time in the past. In this 

 sense the possession of memory by animals is at least 

 unprovable. What we can say is that a single occurrence 

 often has a permanent effect upon the animal, as shown by 



