THE EVOLUTION OF INTELLIGENCE in 



rational process a piece of behavior which may be partly or 

 wholly instinctive, or which may have been acquired by wholly 

 unknown means at some earlier period of the animal's life. 



When due allowance is made for such modifications as have 

 been already suggested, the general upshot of the observations 

 of the experimentalist may perhaps be summarized in this 

 way. The very simple forms, like the low marine orders, 

 exhibit a relatively limited group of reactions, some of them 

 tropistic in character, others, however, seemingly somewhat 

 variable, but few of them suggesting any exercise of intelligent 

 discrimination. A little further up the scale, animal forms are 

 encountered which are equipped with fairly definite instincts 

 and which evince the ability to learn to accommodate them- 

 selves to changes of environment, but only by the slowest and 

 most tedious processes, giving practically no indication of any 

 trait which one would naturally designate intelligent. Thus 

 the frog, for example, by earlier observers supposed to be 

 entirely incapable of taking on any modified reactions by the 

 process of learning, has now, through the most patient obser- 

 vations, been found to be capable, after many, many trials, of 

 slightly improving his reactions. Granted, for example, two 

 pathways, one of which always leads to food and the other 

 not, he can, if the patience of the experimenter is sufficient, 

 ultimately learn to choose the correct pathway. Still further 

 up the scale, as will be illustrated in some of the lower mam- 

 malian forms, we get more positive evidence of the beginnings 

 of real memory and of a more definite ability to modify re- 

 actions in a beneficial manner, although this capacity with 

 some of the mammals, such as the guinea pig, is almost at the 

 zero point. With others, however, and particularly with the 

 primates, there is much greater adaptability, and in a few 

 seemingly well-authenticated instances there is evidence for 

 something at least suggesting the human forms of inference 



