300 ROYAL SOCIETY OF CANADA 



extended series of experiments if these are to be of the nature of wide 

 generalizations. 



The above observations are presented simply for what they may 

 be worth to each reader. I know of none of a similar character on 

 birds of these species. They are in no respect ideal but may prove 

 at least suggestive. 



It was difficult for me even within the narrow range of observ- 

 ation afforded by the few experiments made on the crow not to realize 

 how greatly different a so-called association may be in one species of 

 bird, even one individual, as compared with another. 



An " association " in the case of a gull and a crow, are, I am 

 convinced very different things. They have elements in common, but 

 what is not common may be the greater part. The crow seemed to 

 show real intelligence ; he had learned a good deal by a few experi- 

 ences, and "association" in the narrower sense does not wholly cover this. 

 One sees who looks without prejudice much that cannot be put [into 

 words, but which nevertheless impresses, and must influence his judg- 

 ment, and it is for this reason that I have always laid so much stress 

 on that general knowledge of animals that must modify all conclusions 

 derived from experiments which in the nature of the case can give 

 but a very limited amount of real knowledge of the psychic processes 

 of any animal. 



Let us have all the experiments possible — let us gather facts — ' 

 let the facts be much more numerous than the theories, and let us be 

 very cautious in drawing conclusions, especially such as seem to be 

 radically destructive in tendency. 



I am as thoroughly convinced as ever that the simple explanations 

 some offer are not justified for the higher mammals or even the higher 

 vertebrates. 



