tions of the moment. In other words, these actions resemble, in a
way, sneezing or hiccoughing in human kind; actions which are
stimulated by certain states of the body, but which are not at all
under the control of the will. Howling or bellowing doubtless
represents, in a measure, a state of mind as well as of body, but the
action is of a general and uncontrolled kind.

The effect of advancing culture upon a dog has been gradually to
decrease this ancient undifferentiated mode of expression afforded by
howling and yelping, and to replace it by the much more speech-like
bark. There is some doubt whether the dogs possessed by savages have the
power of uttering the sharp, specialized note which is so characteristic
of the civilized forms of their species. It is clear, however, that if
they have the capacity of thus expressing themselves, they use it but
rarely. On the other hand, our high-bred dogs have, to a great extent,
lost the habit of expressing themselves in the ancient way. Many of our
breeds appear to have become incapable of ululating. There is no doubt
but this change in the mode of expression greatly increases the capacity
of our dogs to set forth their states of mind. If we watch a high-bred
dog, one with a wide range of sensibilities, which we may find in breeds
which have long been closely associated with man, we may readily note
five or six varieties of sound in the bark, each of which is clearly
related to a certain state of mind. The bark of welcome, of fear, of
rage, of doubt, and of pure fun, are almost always perfectly distinct to
the educated ear, and this although the observer may not be acquainted
with the creature; if he knows him well, he may be able to distinguish
various other intonations--those which express impatience and even an
element of sorrow. This last note verges toward the howl.

It does not seem to me that we should regard barking as a new and
useful invention; there are, indeed, few such in the organic world. The
sound appears to me to have been derived from the primitive habit of
howling. If we hearken to this utterance we perceive that it is not an
unbroken sound, but is somewhat intermittent. At either end of the
prolonged sound we can often notice that it is divided into rather
distinct yelps more or less completely separated from the other notes.
The cries of a dog when beaten often exhibit the same peculiarity; so,
too, the puppy, before he has attained skill in barking, will