ANIMAL BEHAVIOUR 221 



same kind. The difference is not one of complexity, but of 

 kind. The starfish is not mechanically necessitated to act 

 as it does; it often checkmates mechanism. Even when we 

 say that it is coerced by its own brainless constitution we 

 must remember that it was itself in some degree an agent in 

 establishing that constitution. 



Similarly, an animal with a big brain, i.e., a well- 

 developed capacity for intelligent behaviour, is free com- 

 pared with a starfish. By careful study we may reduce 

 the experimental indeterminism and predict with some suc- 

 cess what our dog will do in a particular situation. But we 

 are likely to make a bigger mistake than we made with our 

 starfishes if we argue from our dog to our neighbour's. For 

 why, the individuality of the dog is so much greater than 

 that of the starfish. The details of its behaviour are deter- 

 mined much less by its general constitution and much more 

 by the character which it has itself been an agent in building 

 up. Thus we see in the realm of organisms a ladder of 

 emancipation the evolution of free-will. 



The impression which we get from the study of even star- 

 fishes seems to us to hold for the whole realm of Animate 

 Nature and for ourselves in it. We find neither systems of 

 absolute determinism nor ' miscellanies of miracles ', but 

 systems in which determinism and freedom are both illus- 

 trated, sometimes more of the one and sometimes more of the 

 other. The enregistrations within an organism limit its 

 actions within certain trammels; the reflexes, the tropisms, 

 the instincts, the hereditary appetites are all determinist 

 in effect; but our conception of the typical organism is not 

 complete unless we recognise its possibilities of initiative 

 and experiment, of trial and error, of choice and control. 

 Perhaps we may profitably continue into the realms of ethics 



