pterodactyles to birds ends here, unless I may add the entire absence of
teeth which characterises the great pterodactyles (_Pteranodon_)
discovered by Professor Marsh. All other known pterodactyles have teeth
lodged in sockets. In the vertebral column and the hind-limbs there are
no special resemblances to birds, and when we turn to the wings they are
found to be constructed on a totally different principle from those of
birds.

There are four fingers. These four fingers are large, and three of them,
those which answer to the thumb and two following fingers in my
hand--are terminated by claws, while the fourth is enormously prolonged
and converted into a great jointed style. You see at once, from what I
have stated about a bird's wing, that there could be nothing less like a
bird's wing than this is. It was concluded by general reasoning that
this finger had the office of supporting a web which extended between it
and the body. An existing specimen proves that such was really the case,
and that the pterodactyles were devoid of feathers, but that the fingers
supported a vast web like that of a bat's wing; in fact, there can be no
doubt that this ancient reptile flew after the fashion of a bat.

Thus, though the pterodactyle is a reptile which has become modified in
such a manner as to enable it to fly, and therefore, as might be
expected, presents some points of resemblance to other animals which
fly; it has, so to speak, gone off the line which leads directly from
reptiles to birds, and has become disqualified for the changes which
lead to the characteristic organisation of the latter class. Therefore,
viewed in relation to the classes of reptiles and birds, the
pterodactyles appear to me to be, in a limited sense, intercalary forms;
but they are not even approximately linear, in the sense of exemplifying
those modifications of structure through which the passage from the