252 'Journal of Comparative ISI eurology and Psychology. 



one order predominantly in the morning, and the reverse in 

 the afternoon of the same day, and in general with two fastenings 

 the two orders appeared alternately. Consequently I should say 

 that the monkeys are superior to the raccoons as habit-formers. 

 The raccoons operated almost as complex mechanisms, but they 

 could not reduce the performance to routine. Finally, if the 

 natural history books are to be believed, trappers, m order to 

 ensnare the raccoon depend not on his habits but on his instincts. 

 The trap is not put where he habitually enters the stream, for he 

 enters all along it; instead a bright swinging object is hung over 

 the trap so that in reaching ior it he steps on the trigger. Finally: 

 a corollary of the proposition that there are two types of learning, 

 namely, learning by trial and error and learning by means of ideas, 

 should be that there are two types of forgetting, distinguished 

 especially by their time intervals. This, our records seem to show 

 when compared with those for dogs and cats. 



Variability. — In addition to having no fixed order for groups 

 of fastenings, the raccoon changes his method of reacting to a 

 single fastening. I have shown that which paw he uses depends 

 on his position with regard to the latch to be unfastened [cf. p. 225). 

 As already stated, No. 2, who had learned perfectly to lift the 

 horizontal hook with his mouth finally changed to lifting it with 

 his nose. Finally, raccoons which have done two acts separately 

 hundreds of times may suddenly come to do them simultaneously. 

 No. 3 and No. 4 depressed the platform and pulled loop i in Boxes 

 12 and 13 at the same time. Others occasionally worked the 

 thumb-latch and a button at once, and sometimes two buttons 

 w^ere depressed simultaneously. 



Association by Similarity. — If a latch similar to another be 

 added to a group of fastenings, but in a different place, it may be 

 attacked and worked first. I cannot say certainly that this is 

 untrue of a dissimilar fastening for, while it was not the fact with 

 the horizontal hook, the wooden plug or the platform as a plat- 

 form, it would probably have occurred with the thumb-latch had 

 I not first used it singly. The vertical cord leading down to the 

 platform was jumped at directly and vigorously pulled by No. i 

 as soon as he saw it, as if he thought it another loop; later he 

 learned to jump upon the platform. He also worked a second 

 barn-door latch before the first one, with which he was familiar. 

 The second was two inches above the first. All the animals would 



