250 THE INTELLIGENCE OF MAMMALS 
in my mind a general conviction that animals are capable 
of drawing simple conclusions, but there are alternate pos- 
sibilities of interpretation in many cases, and incompleteness 
of data in so many others that I am unable to present any 
number of instances in which inference of a very articulate 
kind is indubitably shown. The subject is one on which 
we need more experiments performed by investigators 
acquainted with the animal’s previous history and keenly 
alive to the various possible psychological interpretations 
which may be put upon an animal’s behavior. The diffi- 
culties and pitfalls of the subject are far beyond the realiza- 
tion of most of the contributors to our data on comparative 
psychology. There is a large amount of material too care- 
fully recorded to be cavalierly rejected as worthless, but too 
incomplete to be accepted as entirely conclusive on the sub- 
ject of animal inference; it will doubtless prove of great 
value in suggesting lines for future work. 
A case in point is the following account of two dogs, 
contributed by Mr. Stone to Romanes’ “Animal Intelli- 
gence.” “One of them, the larger, had a bone, and when he 
had left it the smaller dog went to take it, the larger one 
growled, and the other retired to a corner. Shortly after- 
ward the larger dog went out, but the other did not appear 
to notice this, and at any rate did not move. A few minutes 
later the large dog was heard to bark out of doors; the little 
dog then, without a moment’s hesitation, went straight to 
the bone and took it. It thus appears evident that she 
reasoned—'That dog is barking out of doors, therefore he 
is not in this room, therefore it is safe for me to take the 
bone.’ The action was so rapid as to be clearly a conse- 
quence of the other dog’s barking.” . 
The behavior described will not appear to anyone familiar 
with dogs as anything improbable. The doubtful feature 
