integument, while the edge of the hand carries the principal
quill feathers. In the _Archaeopteryx_, the upper-arm bone is like that
of a bird; and the two bones of the fore-arm are more or less like those
of a bird, but the fingers are not bound together--they are free. What
their number may have been is uncertain; but several, if not all, of
them were terminated by strong curved claws, not like such as are
sometimes found in birds, but such as reptiles possess; so that, in the
_Archaeopteryx_, we have an animal which, to a certain extent, occupies a
midway place between a bird and a reptile. It is a bird so far as its
foot and sundry other parts of its skeleton are concerned; it is
essentially and thoroughly a bird by its feathers; but it is much more
properly a reptile in the fact that the region which represents the hand
has separate bones, with claws resembling those which terminate the
fore-limb of a reptile. Moreover, it had a long reptile-like tail with a
fringe of feathers on each side; while, in all true birds hitherto
known, the tail is relatively short, and the vertebrae which constitute
its skeleton are generally peculiarly modified.

[Illustration: FIG. 5.--ICHTHYORNIS DISPAR (Marsh).

(Side and upper views of half the lower jaw; and side and end views of a
vertebra.)]

Like the _Anoplotherium_ and the _Palaeotherium_, therefore,
_Archaopteryx_ tends to fill up the interval between groups which, in
the existing world, are widely separated, and to destroy the value of
the definitions of zoological groups based upon our knowledge of
existing forms. And such cases as these constitute evidence in favour of
evolution, in so far as they prove that, in former periods of the
world's history, there were animals which overstepped the bounds of
existing groups, and tended to merge them into larger assemblages. They
show that animal organisation is more flexible than our knowledge of
recent forms might have led us to believe; and that many structural
permutations and combinations, of which the present world gives us no
indication, may nevertheless have existed.

But it by no means follows, because the _Palaeotherium_ has much in
common with the horse, on the one hand, and with the rhinoceros on the
other, that it is the intermediate form through which rhinoceroses have
passed to become horses, or _vice versa_; on the contrary, any such
supposition would certainly be erroneous. Nor do I think it likely that
the tr