do it.

Wundt, whilst differing from Frazer in his description of magic, is at
one with him in believing that before religion existed there was an age
of magic. But Wundt's view that marvels are magic when supposed to have
been done by man, but miracles when supposed to have been done by a god
or his priests, suggests the possibility that, as the belief in magic is
found usually, if not always, to exist side by side with the belief in
miracles, the two beliefs may from the beginning have co-existed, that
the age of magic is not prior in the course of evolution to the age of
religion. This possibility, it will be admitted, derives some colour at
least from the way in which the theory of evolution is employed to
account for the origin of species: different though reptiles are from
birds, the serpent from the dove, both are descended from a common
ancestor, the archaeopteryx. If this instance be taken as typical of the
process of evolution in general, then the course of evolution is not, so
to speak, linear or rectilinear, but--to use M. Bergson's
word--'dispersive'. To suppose that religion is descended from magic
would then be as erroneous as to suppose that birds are descended from
reptiles or man from the monkey. The true view will be that the course
of evolution is not linear, is not a line produced for ever in the same
direction, not a succession of stages, but is 'dispersive', that from a
common starting point many lines of evolution radiate in different
directions. The course of evolution is not unilinear but multilinear; it
runs on many lines which diverge, but all the diverging lines start from
the same point.

If we apply this conception of evolution in general to the evolution of
religion in particular--and Bergson, I should say, does not--then the
centre of dispersion, common to all religions, is the heart of man. The
forms of religion evolving, emanating and radiating from that common
centre are, let us say, fetishism, polytheism, and monotheism. If we
wish to avoid, in the theory of religious evolution, an error analogous
to that of supposing birds to be descended from reptiles, we must
decline to suppose that monotheism is simply polytheism evolved, or that
polytheism is descended from fetishism. We must consider that each of
these three forms of religion is terminal, in the sense that no one of
them leads on to, or passes into, either of the other two. All three
forms of religious life may, and indeed do,