332 MENTAL EVOLUTION IN ANIMALS. 



else than a rational one This, however, is just 



the process which we saw must arise whenever, from increas- 

 ing complexity and decreasing frequency, the automatic adjust- 

 ment of inner to outer relations becomes uncertain and 

 hesitating. Hence it is clear that the actions we call instinc- 

 tive pass gradually into the actions we call rational." 



Now in an earlier part of this treatise I have stated my 

 belief that consciousness arises when a nerve-centre is sub- 

 jected to a comparative turmoil of molecular forces, which 

 finds its physiological expression in delay of response, or, as 

 Mr. Spencer says, in "hesitation." But I do not believe that 

 in all such cases Eeason, as distinguished from Consciousness, 

 must arise. Therefore I should say that, although there 

 cannot be Eeason without such ganglionic friction, there 

 may be such ganglionic friction without Eeason. There may, 

 for example, be a large, and even a distressing amount of 

 such friction produced in the case of a conflict of instincts ; 

 there may in such cases be prolonged delay ending in " the 

 strongest group of antagonistic tendencies at length passing 

 into action ;" and yet no act of reason need arise. 



In what respect, then, do I differ from Mr. Spencer touch- 

 ing the genesis of Eeason ? I differ from him, firstly, in not 

 deeming an act of reason as such a constant or invariable 

 index of ganglionic disturbance greater than that which may 

 arise under other circumstances of psychical activity (and 

 therefore in not deeming that reason must necessarily arise out 

 of such disturbance) ; and, secondly, in not deeming that 

 Eeason can only arise out of Instinct. 



Taking these two points of difference separately, it will be 

 enough to say of the first that it has reference only to the 

 earliest origin of Eeason, or to acts of reason of the simplest 

 kind ; in the case of more elaborate processes of reasoning 

 I have no doubt that the ganglionic disturbance must be 

 great, and that without such disturbance these more elaborate 

 processes would not be possible. But this, of course, is a 

 widely different matter from concluding that wherever gan- 

 glionic disturbance reaches a certain degree of complexity, 

 leading to a consequent delay of response, there Eeason (as 

 distinguished from vividness of consciousness) must neces- 

 sarily arise. On the contrary, I hold that in the lower stages 

 of what I have defined as Eeason (and, a fortiori, in all the 

 stages of what I have defined as Inference), there may not be 



