reduced to the alternative of the
impossibility of building or the modification of their habits, certain
animals modify them. Judging from this, how refuse them invention
altogether? This contradicts in no way the very just reservation of
Romanes. It is sufficient to remark that abstraction or dissociation has
stages, that the simplest are accessible to the animal intelligence. If,
in the absence of words, the logic of concepts is forbidden it, there
yet remains the logic of images,[37] which is sufficient for slight
innovations. In a word, animals can invent according to the extent that
they can dissociate.

In our opinion, if we may with any truthfulness attribute a creative
power to animals, we must seek it elsewhere. Generally speaking, we
attribute only a mediocre importance to a manifestation that might very
well be the proper form of animal fancy. It is purely motor, and
expresses itself through the various kinds of play.

Although play may be as old as mankind, its psychology dates only from
the nineteenth century. We have already seen that there are three
theories concerning its nature--it is "expenditure of superfluous
activity," "a mending, restoring of strength, a recuperation," "an
apprenticeship, a preliminary exercise for the active functions of life
and for the development of our natural gifts."[38] The last position,
due to Groos, does not rule out the other two; it holds the first valid
for the young, the second for adults; but it comprehends both in a more
general explanation.

Let us leave this doctrinal question in order to call attention to the
variety and richness of form of play in the animal world. In this
respect the aforementioned book of Groos is a rich mine of evidence to
which I would refer the reader. I limit myself to summing up his
classification. He distinguishes nine classes of play, viz.: (1) Those
that are at bottom experimental, consisting of trials at hazard without
immediate end, often giving the animal a certain knowledge of the
properties of the external world. This is the introduction to an
experimental physics, optics, and mechanics for the brood of animals.
(2) Movements or changes of place executed of their own accord--a very
general fact as is proven by the incessant movements of butterflies,
flies, birds, and even fishes, which often appear to play in the water
rather than to seek prey; the mad running of horses, dogs, etc., in free
space. (3) Mimicry of hunting, i.e.,