ansition from the reptile to the bird has been effected by such a
form as _Archaeopteryx_. And it is convenient to distinguish these
intermediate forms between two groups, which do not represent the actual
passage from the one group to the other, as _intercalary_ types, from
those _linear_ types which, more or less approximately, indicate the
nature of the steps by which the transition from one group to the other
was effected.

I conceive that such linear forms, constituting a series of natural
gradations between the reptile and the bird, and enabling us to
understand the manner in which the reptilian has been metamorphosed into
the bird type, are really to be found among a group of ancient and
extinct terrestrial reptiles known as the _Ornithoscelida_. The remains
of these animals occur throughout the series of Mesozoic formations,
from the Trias to the Chalk, and there are indications of their
existence even in the later Palaeozoic strata.

Most of these reptiles, at present known, are of great size, some having
attained a length of forty feet or perhaps more. The majority resembled
lizards and crocodiles in their general form, and many of them were,
like crocodiles, protected by an armour of heavy bony plates. But, in
others, the hind-limbs elongate and the fore-limbs shorten, until their
relative proportions approach those which are observed in the
short-winged, flightless, ostrich tribe among birds.

The skull is relatively light, and in some cases the jaws, though
bearing teeth, are beak-like at their extremities and appear to have
been enveloped in a horny sheath. In the part of the vertebral column
which lies between the haunch bones and is called the sacrum, a number
of vertebrae may unite together into one whole, and in this respect, as
in some details of its structure, the sacrum of these reptiles
approaches that of birds.

But it is in the structure of the pelvis and of the hind limb that some
of these ancient reptiles present the most remarkable approximation to
birds, and clearly indicate the way by which the most specialised and
characteristic features of the bird may have been evolved from the
corresponding parts in the reptile.

In Fig. 6, the pelvis and hind-limbs of a crocodile, a three-toed bird,
and an ornithoscelidan are represented side by side; and, for facility
of comparison, in corresponding positions; but it must be recollected
that, while the position of the bird's limb is natural, that