INSTINCT AND CONSCIOUSNESS OF FINALITY. 33 



consciously? In the natural sciences it is a very doubtful 

 proceeding to admit into a definition any mark which 

 cannot be judged upon empirically." 



On the contrary, we must insist on a principle well 

 explained by Romanes in his "Animal Intelligence." 

 "Taking it for granted," he says, "that the external 

 indications of mental processes which we observe in 

 animals are trustworthy, so that we are justified in in- 

 ferring particular mental states from particular bodily 

 actions, it follows that in consistency we must every- 

 where apply the same criteria. . . . It is, of course, per- 

 fectly true that the less the resemblance, the less is the 

 value of any analogy built upon the resemblance, and 

 therefore that the inference of an ant or a bee feeling 

 sympathy or rage is not so valid as is the similar in- 

 ference in the case of a dog or a monkey. Still it is 

 an inference, and, so far as it goes, a valid one being, 

 in fact, the only inference available. That is to say, 

 if we observe an ant or a bee apparently exhibiting 

 sympathy or rage, we must either conclude that some 

 psychological state resembling that of sympathy or 

 rage is present, or else refuse to think about the sub- 

 ject at all; from the observable facts there is no other 

 inference open. ' ' Romanes adds that the analogy from 

 human to brute psychology becomes weaker and weak- 

 er as we recede through the animal kingdom down- 

 wards from man; but he insists that "it is the only 

 analogy available" and "that when we get down as 

 low as the insects I think the most we can confidently 

 assert is that the known facts of human psychology 

 furnish the best available pattern of the probable facts 

 of insect psychology. " * ) 



') P. 8. 



