336 THE BIOLOGY OF BIRDS 



We suggest forming a group of these three kinds of 

 behaviour — compound reflex actions, tropisms, and en- 

 registered rhythms, because they grade into one another. 

 And about the same area on our incHned plane we may 

 rank non-intelhgent experimenting, experiential learning, 

 and associative learning. 



A careful naturalist has described the occasional combat 

 between a starfish and a small sea-urchin, in the course 

 of which the starfish persists in a rather tedious process of 

 disarming the sea-urchin, wrenching off' its scissor-like 

 snapping blades. As the starfish has no nerve-centres, 

 not to speak of brain, it is unwarrantable to use any big 

 psychological term. It seems to be the outcome of non- 

 intelligent experimentation. Let us take a possibly clearer 

 case, a starfish turned upside down. The animal rights 

 itself after a series of movements. When the experiment 

 is repeated day after day, the starfish learns (but not intelli- 

 gently) to right itself more quickly, with fewer useless 

 movements. 



But there are higher grades of " learning," as when 

 frog or tortoise becomes able to find its way quickly out of 

 a labyrinth or maze. It learns from its experience, but 

 how it does so remains obscure. That it discovers the secret 

 of the maze may be ruled out, and that it forms a mental 

 picture is also impossible. Slightly higher is associative 

 learning, which is very common among animals. At a 

 given signal certain fishes come to the bank for food ; when 

 a peculiar sound is heard the dog acts in a perfectly precise 

 way ; when a bird hears a certain voice it becomes excited. 

 In hundreds of cases a certain stimulus comes to be asso- 

 ciated in the individual lifetime with a particular experience 

 or a particular course of action. Much of the education 

 of young birds by their parents depends on the establish- 

 ment of associations. 



(7, 8) On a higher level on "the reaction side" are simple 

 instincts and chain-instincts. It is characteristic of instinc- 

 tive behaviour that it depends on inborn pre-arrangements 

 of certain nerve-cells and certain muscle-cells, that it does 



