with those of the other there is no general agreement and there can be
none. The usual practice, however, is to call the Iranian _Yima,
haoma_, etc., to one's aid if they subserve one's own view of _Yama,
soma,_ and other Hindu parallels, and to discard analogous features as
an independent growth if they do not. This procedure is based as well
on the conditions of the problem as on the conditions of human
judgment, and must not be criticized too severely; for in fact the two
religions here and there touch each other so nearly that to deny a
relation between them is impossible, while in detail they diverge so
widely that it is always questionable whether a coincidence of ritual
or belief be accidental or imply historical connection.

It is scarcely advisable in a concise review of several religions to
enter upon detailed criticism of the methods of interpretation that
affect for the most part only the earliest of them. But on one point,
the reciprocal relations between the Vedic and Brahmanic periods, it
is necessary to say a few words. Why is it that well-informed Vedic
scholars differ so widely in regard to the ritualistic share in the
making of the Veda? Because the extremists on either side in
formulating the principles of their system forget a fact that probably
no one of them if questioned would fail to acknowledge. The Rig Veda
is not a homogeneous whole. It is a work which successive generations
have produced, and in which are represented different views, of local
or sectarian origin; while the hymns from a literary point of view are
of varying value. The latter is a fact which has been ignored
frequently, but it is more important than any other. For one has
almost no criteria, with which to discover whether the hymns precede
or follow the ritual, other than the linguistic posteriority of the
ritualistic literature, and the knowledge that there were priests with
a ritual when some of the hymns were composed. The bare fact that