perating
upon more and more widely different antecedent conditions of the mineral
framework of the earth; until, at length, in place of that framework, he
would behold only a vast nebulous mass, representing the constituents of
the sun and of the planetary bodies. Preceding the forms of life which
now exist, our observer would see animals and plants not identical with
them, but like them; increasing their differences with their antiquity
and, at the same time, becoming simpler and simpler; until, finally, the
world of life would present nothing but that undifferentiated
protoplasmic matter which, so far as our present knowledge goes, is the
common foundation of all vital activity.

The hypothesis of evolution supposes that in all this vast progression
there would be no breach of continuity, no point at which we could say
"This a natural process," and "This is not a natural process;" but that
the whole might be compared to that wonderful process of development
which may be seen going on every day under our eyes, in virtue of which
there arises, out of the semi-fluid, comparatively homogeneous substance
which we call an egg, the complicated organization of one of the higher
animals. That, in a few words, is what is meant by the hypothesis of
evolution.

* * * * *

I have already suggested that in dealing with these three hypotheses, in
endeavouring to form a judgment as to which of them is the more worthy
of belief, or whether none is worthy of belief--in which case our
condition of mind should be that suspension of judgment which is so
difficult to all but trained intellects--we should be indifferent to all
_a priori_ considerations. The question is a question of historical
fact. The universe has come into existence somehow or other, and the
problem is, whether it came into existence in one fashion, or whether it
came into existence in another; and, as an essential preliminary to
further discussion, permit me to say two or three words as to the nature
and the kinds of historical evidence.

The evidence as to the occurrence of any event in past time may be
ranged under two heads which, for convenience' sake, I will speak of as
testimonial evidence and as circumstantial evidence. By testimonial
evidence I mean human testimony; and by circumstantial evidence I mean
evidence which is not human testimony. Let me illustrate by a familiar
example what I understand by these two kinds of evidenc