ORGANIC MOVEMENTS 39 



to find an upper one. All instincts are separated from the 

 next higher group of motion, which we propose to call 

 " actions " in the widest sense of the word, by being com- 

 plete in their specificity from the very first time they occur. 

 There may be some improvement in consequence of their 

 being repeated, but this improvement never affects their 

 specificity as such. Perhaps it will be more correct to say 

 that we shall not apply the term " instinct " to any animal 

 movement that shows an improvement with regard to its 

 specificity. 



Instincts are often said to be " purposeful " with regard 

 to their performer. We prefer to say, at present, that they 

 possess some regulative character ; that they tend to " nor- 

 mality " with regard to the whole life of the organism which 

 performs them. Here the limit between instincts and 

 other classes of motions is not always very clearly marked : 

 almost all typically combined motions, be they pure chain- 

 reflexes or be they of a more complicated type, are alike in 

 possessing a regulative character. And it is impossible to 

 draw a sharp boundary here, if one has renounced the 

 question of " consciousness " as illegitimate. In fact, all 

 instincts are chains of single nervous acts concerned in 

 movements, just as are real chain-reflexes and many other 

 combined motions : it is only the degree of chaining that 

 comes into account. 



But what is the meaning of the word " degree " in this 

 connexion ? Does it apply only to different states of com- 

 plication of the same invariable general type ? It is here 

 that our analytical problems begin. 



