ittle game of rapping "1!"--the meaning
of which, she had once informed me, was "_I won't tell!_" and the
sequel to which I generally found to be that she would put me off with
any word that might just happen to come into her head. But why had this
remark occurred to her yesterday? I wanted to get to the bottom of it,
so returning to the attack, said: "Why wouldn't you tell me yesterday
what water is good for?" "I thought of ear!" "What has water to do with
'ear'?" "water in ear horrid!" Here, then, was the reason! In her very
fear she had not been able to bring forth her true answer--for, owing
to me, the water had got into her ears--and made this lasting and
unpleasant impression--when she was being bathed--or when I threw her
into a stream! The reader may already have noticed other instances
where a direct connexion of ideas has occurred. I have purposely
abstained from pointing to the obvious in each case, believing that
anyone who is keenly interested will do so quickly enough for himself,
and I am loth to weary my Public by needless repetitions.

SPONTANEOUS REPLIES

Spontaneous replies provide a special proof of this ability to form
independent thoughts, and is found both among horses and dogs. Such a
reply is indeed the sudden and evident utterance of some thought, and
of a thought which--to it--transcends all other thoughts at the moment:
one which regardless of all other questions which may at the time be
put to it, looms largest, and the animal will therefore utter this
remark, asked or unasked--and quite independently of any question, but
more after the manner of "making an observation." Such a thought may
have nothing to do with the subject in hand, and persons who are
participating in this conversation _a deux_, can only arrive at the
inference of ideas after having carefully thought the matter over--it
may also be that they will fail to see any association of ideas at all.
Now, it is indisputable that such replies belong to the most important
category--for they may serve as proofs to those who themselves have not
worked with animals for any length of time, and who, therefore, cannot
become sincerely convinced as to the truth of the matter by travelling
the longer road of personal test and experience. The teacher of any
horse or dog of good parts does not need this proof: there are
thousands of small instances which in their sum total prove
important--trivial and uncertain though each one may be,