ral form common to
the lower Eastern monkeys, the anthropoid apes, and man, on the other.
But considerations similar to those which showed it to be impossible
that man should have developed from an ancestor common to him and the
monkeys, yet outside of and parallel with these, may be urged also
against the likelihood of a parallel evolution of the lower Eastern
monkeys, the anthropoid apes, and man. The anthropoid apes have in
common with man many characters which are not present in the lower
Old World monkeys. These characters must therefore have been present
in the ancestral form common to the three groups. But here, again, it
is difficult to understand why the lower Eastern monkeys should not
also have inherited these characters. As this is not the case, there
remains no alternative but to assume divergent evolution from an
indifferent form. The lower Eastern monkeys are carrying on the
evolution in one direction--I might almost say towards a blind
alley--while anthropoids and men have struck out a progressive path,
at first in common, which explains the many points of resemblance
between them, without regarding man as derived directly from the
anthropoids. Their many striking points of agreement indicate a common
descent, and cannot be explained as phenomena of convergence.

I believe I have shown in the above sketch that a theory which derives
man directly from lower forms without regarding apes as
transition-types leads _ad absurdum_. The close structural
relationship between man and monkeys can only be understood if both
are brought into the same line of evolution. To trace man's line of
descent directly back to the old Eocene mammals, alongside of, but
with no relation to these very similar forms, is to abandon the method
of exact comparison, which, as Darwin rightly recognised, alone
justifies us in drawing up genealogical trees on the basis of
resemblances and differences. The farther down we go the more does the
ground slip from beneath our feet. Even the Lemuridae show very
numerous divergent conditions, much more so the Eocene mammals
(Creodonta, Condylarthra), the chief resemblance of which to man
consists in the possession of pentadactylous hands and feet! Thus the
farther course of the line of descent disappears in the darkness of
the ancestry of the mammals. With just as much reason we might pass by
the Vertebrates altogether, and go back to the lower Invertebrates,
but in that case it would be much ea