and on each side one is abdominal, two are thoracic, one clavicular, and one cervical, which last is at the base of the neck. The clavicular and abdominal air cells are perhaps the most interesting. The air cell termed clavicular sends a process outward towards the arm, along with the blood vessels which supply the arm. Thus this air cell, entering the region of the axilla or arm-pit, enters the upper arm bone usually on its under side, close to the articular head of the humerus, and in the same way the air may pass from bone to bone through every bone in the fore limb. The hind limbs similarly receive air from the abdominal air cell, which supplies the femur and other bones of the leg, the sacrum, and the tail. But the joints of the backbone in front of the sacrum receive their air from the cervical air sac. The air cells are not limited to the bones, but ramify through the body, and in some cases extend among the muscles. A bird may be said to breathe not only with its lungs, but with its whole body. And it is even affirmed that respiration has been carried on through a broken arm bone when the throat was closed, and the bird under water. Birds differ greatly in the extent to which the aircell system prolonged from the lungs is developed, some having the air absent from every bone, while others, like the Swift, are reputed to have air in every bone of the body. Comparison shows that in so far as the bones are the same in Bird and Ornithosaur, the evidence of the air cells entering them extends to resemblance, if not coincidence, in every detail. No living group of animals except birds has pneumatic limb bones, in relation to the lungs; so that it is reasonable to conclude that the identical structures in the bones were due to the same cause in both the living and extinct groups of animals. It is impossible to say that the lungs were identical in Birds and Pterodactyles, but so far as evidence goes, there is no ground for supposing them to have been different. THE LUNGS OF REPTILES [Illustration: FIG. 18. THE SIDE OF THE BODY OF A CHAMELEON Ribs removed to show the sacculate branched form of the lung] There is nothing comparable to birds, either in the lungs of living reptiles or in their relation to the bones. The Chameleon is remarkable in that the lung is not a simple bladder prolonged through the whole length of the body cavity, as in a serpent, but it develops a number of large lateral branches visible when the body is laid open. Except near the trachæa, where the tissue has the usual density of a lizard lung, the air cell is scarcely more complicated than the air bladder of a fish, and does not enter into any bone of the skeleton. And although many fishes like the Loach have the swim bladder surrounded by bone connected with the head, it offers no analogy to the pneumatic condition of the bones in the Pterodactyle. THE FORM OF THE BRAIN CAVITY [Illustration: FIG. 19. THE FORM OF THE BRAIN]