the most remote and unsophisticated
barbarians, and as the educated Platonists of the fourth century
after Christ, and that many of the other phenomena should be
identical in each case, is certainly noteworthy. This kind of
folklore is the most persistent, the most apt to revive, and the
most uniform. We have to decide between the theories of independent
invention; of transmission, borrowing, and secular tradition; and of
a substratum of actual fact.

Thus, either the rite of binding the sorcerer was invented, for no
obvious reason, in a given place, and thence reached the Australian
blacks, the Eskimo, the Dene Hareskins, the Davenport Brothers, and
the Neoplatonists; or it was independently evolved in each of
several remote regions; or it was found to have some actual effect--
what we cannot guess--on persons entranced. We are hampered by not
knowing, in our comparatively rational state of development, what
strange things it is natural for a savage to invent. That spirits
should knock and rap seems to us about as improbable an idea as
could well occur to the fancy. Were we inventing a form for a
spirit's manifestations to take, we never should invent _that_. But
what a savage might think an appropriate invention we do not know.
Meanwhile we have the mediaeval and later tales of rapping, some of
which, to be frank, have never been satisfactorily accounted for on
any theory. But, on the other hand, each of us might readily invent
another common 'manifestation'--the _wind_ which is said to
accompany the spirit.

The very word spiritus suggests air in motion, and the very idea of
abnormal power suggests the trembling and shaking of the place
wherein it is present. Yet, on the other side, the 'cold non-
natural wind' of seances, of Swedenborg, and of a hundred stories,
old or new, is undeniably felt by some sceptical observers, even on
occasions where no professional charlatan is engaged. As to the
trembling and shaking of the house or hut, where the spirit is
alleged to be, we shall examine some curious evidence, ancient and
modern, savage and civilised. So of the other phenomena. Some seem
to be of easy natural invention, others not so; and, in the latter
case, independent evolution of an idea not obvious is a difficult
hypothesis, while transmission from the Pole to Australia, though
conceivable, is apt to give rise to doubt.

Meanwhile, one phenomenon, which is usually said to accompany others
much more