443 
of a gun is somewhat different. The fear of guns is not innafe 
in the rook, and it does not fly from the gun until it has 
learned the dangerous properties of that weapon by experience ; 
and on a superficial view the action might seem to* be a rational 
one. Arguing, however, from the automatic nature of the 
human memory, we can explain the action of the rook without 
assuming the intervention of a rational process. A friend of 
mine once informed me that in middle life he settled in the 
same city with his former schoolmaster, and that he never met 
the old gentleman, who was then in an advanced stage of 
decrepitude, without experiencing an unconquerable sensation 
of terror : instances of this kind are of such common occur- 
rence that one will be quite sufficient for the purpose of 
illustration. The terror was produced by a process precisely 
analogous to that which awakens the instinct of fear in the 
rook. There was no innate fear in my friend’s mind of a 
person presenting the particular aspect of his schoolmaster ; 
but the appearance of the schoolmaster having been once 
associated with the idea of danger, a sensation of fear w r asever 
afterwards excited by his presence ; although the slightest 
exercise of the reason was sufficient to show that the fear was 
absurd. No one will dispute the fact that my friend’s feeling 
was instinctive, and that it was checked by the reason before it 
passed into action. But we must consider the action of the 
rook to be instinctive also. The automatic association of im- 
pressions in the memory, which we learn from our own con- 
sciousness, sufficiently explains such instances and enables us 
to establish the principle, that the causes which awaken the 
desires are capable of extension without the operation of a 
rational process. 
6. The third point is more open to objection ; and yet it is 
indispensable to a true apprehension of the distinction between 
man and brute. It is, that what are called intellectual processes 
are in themselves automatic, just as memory is automatic. I 
mean that generalization and abstraction, for instance, are, in 
their simpler forms, merely a part of the psychological mechanism 
of the animal, by means of which the impressions received by 
the senses are duly modified before they act upon the desires, 
and by which the motive power of the desires is directed into 
its proper channels. To requote an instance adduced in a 
former paper on this subject, the bull who is irritated by a red 
colour really abstracts the colour from the object of which it is 
an attribute. The dog who singles out his master from a crowd 
of indifferent persons abstracts likewise ; but the process is 
mechanical, and is another way of expressing the fact that the 
