The differences between those vertebrate types of animals consist chiefly in the way in which their organisation is modified, by one strain of characters being eliminated so that another becomes predominant, while a distinctive set of structures is elaborated in each class of animals. The earlier geological history of the higher Vertebrata is very imperfectly known, but the evidence tends to the inference that the older representatives of the several classes approximate to each other more closely than do their surviving representatives, so that in still earlier ages of time the distinction between them had not become recognisable. The relation of the great groups of animals to each other, among Vertebrata, is essentially a parallel relation, like the colours of the solar spectrum, or the parallel digits of the hand. It was natural, when only the surviving life on the earth was known, to imagine that animals were connected in a continuous chain by successive descent, but Mammals have given no evidence of approximation to Birds; and Birds discover no evidence that their ancestors were Reptiles, in the sense in which that word is used to define animals which now exist on the earth. When the variation which animals attain in their maturity and exhibit in development from the egg was first realised, it was imagined that Nature, by slow summing up and accumulation of differences which were observed, would so modify one animal type that it would pass into another. There is little evidence to support belief that the changes between the types of life have been wrought in that way. The history of fossil animals has not shown transitions of this kind from the lower to higher Vertebrata, but only intermediate, parallel groups of animals, analogous to those which survive, and distinct from them in the same way as surviving groups are distinct from each other. The circumstance that Mammals, Birds, and Reptiles are all known low down in the Secondary epoch of geological time, is favourable to the idea of their history being parallel rather than successive. Such a conception is supported by the theory of elimination of characters from groups of animals as the basis of their differentiation. This loss appears always to be accompanied by a corresponding gain of characters, which is more remarkable in the soft, vital organs than in the skeleton. The gain in higher Vertebrates in the bones is chiefly in the perfection of joints at their extremities; but the gain in brain, lungs, heart, and other soft parts is an elaboration of those structures and an increase in amount of tissue.
The resemblances of Ornithosaurs to Mammals are the least conspicuous of their characters. Those seen in the upper arm bone and thigh bone are manifestly not derived from Mammals. They cannot be explained as adaptations of the bones to conditions of existence, because there is no community of habit to be inferred between Pterodactyles and Mammals, in which the bones are in any way comparable.
Other fossil animals show that a fundamentally Reptilian structure is capable of developing in the Mammalian direction in the skull, backbone, shoulder-girdle, hip-girdle, and limbs, so as to be uniformly Mammalian in its tendencies. This is proved by tracing the North American Texas fossils named Labyrinthodonts, through the South African Theriodonts, towards the Monotremata and other Mammalia. Just as those animals have obliterated all traces of the Bird from their skeletons, Birds have obliterated the distinctive characters of Mammals. The Ornithosaur has partially obliterated both. With a skull and backbone marked by typical characters of the Reptile, it combines the shoulder-girdle and hip-girdle of a Bird, with characters in the limbs which suggest both those types in combination with Mammals.
The bones have been compared in the skeleton of each order of existing Reptiles, and found to show side by side with their peculiar characters not only resemblances to the other Reptilia, but an appreciable number of Mammalian and Avian characters in their skeletons. The term "crocodile," for example, indicates an animal in which the skeleton is dominated by one set of peculiar characters. Crocodiles retain enough of the characteristics of several other orders of reptiles to show that an animal sprung from the old Crocodile stock might diverge widely from existing Crocodiles by intensifying what might be termed its dormant characters in the Crocodile skeleton. Comparing animals together bone by bone it is possible to value the modifications of form which they put on, and the resemblances between them, so as to separate the inherited wealth of an animal's affinities with ancestors or collateral groups, from the peculiar characters which have been acquired as an increase based upon its typical bony possessions or osteological capital. There is no part of the Pterodactyle skeleton which is more distinctly modified than the head of the upper arm bone, which fits into the socket between the coracoid bone and the shoulder-blade. The head of the humerus, as the articular part is named, is somewhat crescent-shaped, convex on its inner border, and a little concave on its outer border, and therefore unlike the ball-shaped head of the upper arm bone in Man and the higher Mammals. It is much more nearly paralleled in the little group of Monotremata allied to the living Ornithorhynchus. In that sense the head of the humerus in a Pterodactyle has some affinity with the lowest Mammalia, which approach nearest to Reptiles. The character might pass unregarded if it were not found in more striking development in fossil Reptiles from Cape Colony, which from having teeth like Mammals are named Theriodontia. In several of those South African reptiles the upper arm bone approaches closer to the humerus in Ornithosaurs than to Ornithorhynchus. Such coincidences of structure are sometimes dismissed from consideration and placed beyond investigation by being termed adaptive modifications; but there can be no hope of finding community of habit between the burrowing Monotreme, the short-limbed Theriodont, and the flying Pterodactyle which might have caused this articular part of the upper arm bone to acquire a form so similar in animals constructed so differently. If the resemblance in the humerus to Monotremes in this respect is not to be attributed to burrowing, neither can the crescent form of its upper articulation be attributed to flight; for in Birds the head of the bone is compressed, but always convex, and Bats fly without any approach to the Pterodactyle form in the head of the humerus. This apparently trivial character may from such comparisons be inferred to be something which the way of life of the animal does not sufficiently account for. These deepest-seated parts of the limbs are slow to adapt themselves to changing circumstances of existence, and retain their characters with moderate variation of the bones in each of the orders or classes of animals. It therefore is safer to regard Mammalian characters, as well as the resemblances which Pterodactyles show to other kinds of animals, as due to inheritance from a time when there was a common stock from which none of these animals which have been considered had been distinctly elaborated.
A few characters of Ornithosaurs are regarded as having been acquired, because they are not found in any other animals, or have been developed only in a portion of the group. The most obvious of these is the elongated wing finger; but in some genera, like Dimorphodon, there is also a less elongation of the fifth digit of the foot, and perhaps in all genera there is a backward development of the first digit of the hand, which is without a claw, and therefore unlike the clawed digit of a Bat. An acquired character of another kind, which is limited to the Cretaceous genera, is seen in the shoulder-blade being directed transversely outward, so that its truncated end articulates by a true joint with the early vertebræ of the back, and defended the cavity inclosed by the ribs by a strong bony external arch. And finally, as the animals later in time acquire short tails, and relatively longer limbs, the bones of the back of the hand, termed metacarpals, acquire greater and distinctive length, which is not seen in the long-tailed types like Rhamphorhynchus.
These and such-like acquired characters distinguish the class of animals from all groups with which it may be compared, and mark the possible limits of variation of the skeleton within the boundary of the order. But no further variation of these parts of the skeleton could make a transition to another order of animals, or explain how the Pterodactyles came into existence, because the characters which separate orders and classes of animals from each other differ in kind from those which separate smaller groups, named genera and species, of which the order is made up. The accumulation of the characters of genera will not sum up into the characters of an order or class.
In making the division of Vertebrate animals into classes the skeleton is often almost ignored. Its value is entirely empirical and based upon the observed association of the various forms of bones with the more important characters of the brain and other vital organs. What is understood as a Mammalian or Avian character in the skeleton is the form of bone which is found in association with the soft vital organs which constitute an animal a Mammal or a Bird.
The characters which theoretically define a Mammal appear to be the enormous overgrowth of the cerebral hemispheres of the brain by which the cerebrum comes into contact with the cerebellum, as among Birds. This character distinguishes both groups of animals from all Reptiles, recent and fossil. But in examining the mould of the interior of the brain case it is rare to have the bones fitting so closely to the brain as to prove that the lateral expansion below the cerebrum and cerebellum is formed by the optic lobes of the brain. Otherwise the brain of a Pterodactyle might be as like to the brain of Ornithorhynchus as it is like that of a Bird ([Fig. 19]). But it is precisely in this condition of arrangement of the parts of the brain that the specimens appear to be most clear. The lateral mass of brain in specimens of Ornithosaurs from the Lower Secondary rocks appears to be transversely divided into back and front parts, which may be thought to correspond to the structures in a Mammal brain named corpora quadrigemina, but to be placed as the optic lobes are placed in Birds, and to have relatively greater dimensions than in Mammals. No evidence has been observed of this transverse division of the optic lobes of the brain in Pterodactyles from the Chalk and Cretaceous rocks, and so far as the evidence goes this part of the brain was shaped as in birds, but rather smaller.
The brain is the only soft organ in which a Mammalian character could be evidenced. The uniformity in character of the brain throughout the group in Mammals is remarkable, in reference to the circumstance that the reproduction varies in type; the lowest, or Monotreme division, being oviparous. If there is no necessary connexion between the Mammalian brain and the prevalent condition under which the young are produced alive, it may be affirmed also that there is no necessary connexion between the form of the brain and the form of the bones, since the brain cavity in Theriodont reptiles shows no resemblance to that of a Mammal, while the bones are in so many respects only paralleled among Monotremata and Mammalia. The variety of forms which the existing Mammalian orders of animals assume, shows the astonishing range of structure of the skeleton which may coexist with the Mammalian brain. And therefore we are led to the conclusion that any other fundamental modification of brain—such as distinguishes the class of Birds—might also be associated with forms and structures of the skeleton which would vary in similar ways. In other words, if for convenience we define a Mammal by its form of brain, structure of the heart and lungs, and provision for nutrition of the young, without regard to the covering of the skin, which varies between the scales of a pangolin and the practically naked skin of the whale—a bird might be also defined by its peculiar conditions of brain and lungs, without reference to the feathered condition of the skin, though the feathered condition extends backward in time to the Upper Secondary rocks, as seen in the Archæopteryx.
The Avian characters of Pterodactyles are the predominant parts of their organisation, for the conditions of the brain and lungs shown by the moulds of the brain case and the thin hollow bones with conspicuous pneumatic foramina, give evidence of a community of vital structures with Birds, which is supported by characters of the skeleton. If any classificational value can be associated with the distribution of the pneumatic foramina as tending to establish membership of the same class for animals fashioned on the same plan of soft organs, the evidence is not weakened when a community of structures is found to extend among the bones to such distinctive parts of the skeleton as the sternum, shoulder-girdle, bones of the fore-arm and fore-leg; for in all these regions the Pterodactyle bones are practically indistinguishable from those of Birds. This is the more remarkable because other parts of the skeleton, such as the humerus and pelvis, show a partial resemblance to Birds, while the parts which are least Avian, like the neck bones, have no tendency to vary the number of the vertebræ, in the way which is common among Birds, following more closely the formula of the seven cervical vertebræ of Mammals.