spiritual or extra-natural agencies. Some learned men say that we had
better drop the word "religion," lest we be misunderstood. They would
rather use "daimonism," or "supernaturalism," or other such new term; but
none of these seems to me so wide and so exactly significant of what I
mean as "religion."

All now agree that in this very broad sense some kind of religion exists
in every human community.[1]

[Footnote 1: I suppose I am not going too far in saying "all agree;" for I
think that the latest study of this subject, by Gustav Roskoff, disposes
of Sir John Lubbock's doubts, as well as the crude statements of the
author of _Kraft und Stoff_, and such like compilations. Gustav Roskoff,
_Das Religionswesen der Rohesten Naturvölker_, Leipzig, 1880.]

The attempt has often been made to classify these various faiths under
some few general headings. The scheme of Auguste Comte still has
supporters. He taught that man begins with fetichism, advances to
polytheism, and at last rises to monotheism. More in vogue at present is
the theory that the simplest and lowest form of religion is individual;
above it are the national religions; and at the summit the universal or
world religions.

Comte's scheme has not borne examination. It is artificial and sterile.
Look at Christianity. It is the highest of all religions, but it is not
monotheism. Look at Buddhism. In its pure form it is not even theism. The
second classification is more fruitful for historical purposes.

The psychologist, however, inquires as to the essence, the real purpose of
religions. This has been differently defined by the two great schools of