yet a certain spell-power may attach even to the highest
types, for we find not infrequently the conception that not only the
power of the worshipper, but the power of the deity also is nourished
and strengthened by prayer, and the prayer itself is usually accompanied
by a potent act (such as that of sacrifice). "May our prayers increase
Agni": "The prayers fill thee with power and strengthen thee, like great
rivers the Sindhu."[385]

I must now turn to the form and manner of Roman prayers, in order to
gain further light on the question as to the mental attitude of the
worshipper towards the deity invoked. Of late years there has been a
strong tendency to find the origin of prayer in spell; or, in other
words, to discover a bridge between that mental attitude which believes
that a deity can be forced into a certain course of action by magical
formulae, and the humble attitude of the petitioner in prayer, which
assumes that the power of the deity altogether transcends that of his
worshipper. The evidence of Roman prayers is, I think, of considerable
value in dealing with this question; but it needs to be carefully
studied and handled. The general impression conveyed by those who have
written on the subject is that Roman prayers were dull, dry formulae,
which were believed to have a constraining influence on the deity simply
as formulae, if they were repeated with perfect precision the right
number of times. Dr. Westermarck, for example, has no shadow of a doubt
about this; quoting Renan, he says that "in the Roman, as in the
majority of the old Italian cults, prayer is a magic formula, producing
its effect by its own inherent quality." And again, he writes that the
Romans were much more addicted to magic than to religion; "they wanted
to compel the gods rather than to be compelled by them. Their _religio_
was probably near akin to the Greek [Greek: katadesmos], which meant not
only an ordinary tie, but also a magic tie or knot or a bewitching
thereby."[386] I need not stop to point out the misconception of the
word _religio_ which suggested the whole of this passage; the supposed
derivation from _ligare_ was quite enough to suggest magic to those who
are on the trail of it.[387] Let us go on to examine the prayers
themselves; I think we shall find that though there is much truth in the
common view of them, it is not quite the whole truth.

The oldest Roman prayers we possess are usually called hymns, because
the Latin