HAVE FISHES MEMOKY. 393 



scarcely be held that pike exhibited with little fish restrain their appe- 

 tite only because their experiences in attempting to devour little fish 

 had apparently been forbidding. 



This exhausts the material offered by my correspondence. The 

 aggregate of the observations made by several hundred persons seems 

 extremely insignificant, but what there is is secured by the fact that 

 almost every item was observed at various times, by different persons, 

 in widely separated places, and often with regard to species of fish 

 not related to one another. 



The very limited capacity of fish appears when we realize that only 

 the following few facts could be deduced from the great amount of 

 material furnished me by observers : 1. The innate impulse toward 

 flight can })e lessened in fish by accustoming them to impressions as a 

 rule alarming ; but tameness so acquired is lost when new stimuli 

 supervene. The impulse toward flight may appear even in conse- 

 quence of stimuli never present before. The fish get shy. 2. The 

 optical or chemical irritation usually calling forth the act of taking 

 food may by long custom be replaced by some other, as, for instance, 

 the optical image of the attendant charged with the duty of feeding 

 the fish. 



In all ascertainable cases it is simply a question of a change of front 

 toward a definite irritation. Fish which as a rule swim up to their 

 food when not disturbed by irrelevant impressions learn to subordi- 

 nate them to such an extent that they approach food even in their 

 presence, or at least do not flee. They also learn to approach food 

 when irritations other than those proceeding directly from the food 

 acquaint them with its presence. They not only swim toward crumbs, 

 l)ut the sight of the person who usually scatters the crumbs attracts 

 them even when no food is in sight. 



Nothing prevents us from grouping these facts under the concept 

 of memory. We are, then, in position to say that vertebrates so low 

 as fish possess a sort of memory, widely differing, by many gradations, 

 from the memory of mammals, the only sort hitherto studied. Com- 

 pared with the latter, it is a very much simpler process, the peculiarity 

 of which lies in the close connection existing between the irritation 

 and the response. Not a single fact forces upon us the assumption 

 that these simple processes are accompanied or dominated ])y the 

 mental process of associating ideas. An ol)servation of Bateson's 

 offers a pregnant illustration of the difference in this respect between 

 fishes and other animals. Motella, as was told above, finds food only 

 by means of the sense of taste or smell. It does not become aware of 

 its presence through any other channel, even if the food enters its field 

 of vision. One of this genus was kept in a shallow reservoir in the 

 Brighton Aquarium, where it was often fed by people l)ending over 

 the sides. This individual learned to rise to the surface when it was 



