often
prolong each utterance in a way which makes its relation to the ancient
mode of expression tolerably clear. At the risk of being deemed
fanciful, I venture to suggest that the bark is in effect a division
of the howl into clearly separated notes, the change having come about
as a similar alteration is effected in our own speech, by the increase
in the intelligence which the creature is called upon to express. I
conceive that while the primitive and massive emotions found
satisfying utterance in the long-drawn notes, the more divided state
of mind of the humanized successor has led to a change in its
utterances. Although these modifications of speech, if such we may
term them, have probably been developed on the basis of the dog's
human relations, there is, it seems to me, good reason to believe that
the diversities in note have come to have a distinct conventional
value between the individuals of all the different breeds. Any one who
closely observes these animals must have noticed the fact that the
degree of attention they give to the utterances of their kindred
varies in a way which indicates that they have great varieties of
denotations. Some of the shades of the meaning which a dog's bark has
to others of his species probably escape our less fine ears.

The creation of something like a language among our civilized dogs
has naturally been accompanied by the development of an understanding
of human speech. Although we cannot attach much importance to the
mass of anecdote on this point, there is enough which is well
attested--sufficient, indeed, which has come within the limits of my
own observation--to make it clear that dogs, even without deliberate
teaching, frequently acquire a tolerably clear understanding of a
number of words and even of short phrases. They will catch these not
only when given in distinct command, but when uttered in an ordinary
tone, without any sign that they relate to their affairs. It is true
that these understood words generally relate to some action which the
dog is accustomed to perform, yet there are instances so well
attested that they deserve credit, which seem to show that the
creatures can get some sense of the drift of conversation even when
it is carried on by persons with whom they are not familiar and does
not clearly relate to their own affairs.

It should be observed that within the narrow limits of this essay little
or no effort has been made to interpret the state of m