These semiaquatic reptiles never grew very large—not more than four or five feet in length; nor did they ever succeed in becoming fully at home in the water, certainly no more so than our modern alligators and crocodiles. They remained to the end of their comparatively brief existence essentially land animals, probably seeking their food in the water but fleeing to the land for protection and for the breeding of their young. Their chief water adaptations are seen in the elongate face and flattened swimming tail. Their legs remained essentially terrestrial, and could have been of but little use in the water for propulsion; the feet even were doubtfully webbed, or if so, not more than are those of the alligator. Singularly, like the proganosaurs, their ribs were heavy and stout, also suggesting bottom-crawling habits, like those of the living Galapagos lizards. The skull was lightly built, and the face was long and slender, like that of the gavials and proganosaurs; but, like those of the former and unlike those of the latter, the nostrils were situated at the extreme tip. The hind legs were firmly attached to the body by the sacrum; and no sclerotic bones of the eyes have been discovered. The neck was neither unusually long nor unusually short. The body was probably covered with horny scales. [Illustration: FIG. 88.—Restoration of _Champsosaurus_.] [Illustration: FIG. 89.—_Champsosaurus_; skull from above. (After Brown.)] To the student of paleontology these animals are of interest because of the retention of several primitive traits which had long disappeared in other known reptiles. While the vertebrae had ceased to be perforated by the notochord, as in the early reptiles, they were still shallowly biconcave. The first bone of the neck, the atlas, had changed but little from that of their very ancient forbears of Permian times, and the bones of the palate still retained numerous teeth scattered over it, like those of the same Paleozoic ancestors. Most primitive and old fashioned of all was the pelvis, which was so unlike that of all known contemporary or later reptiles that, were a paleontologist to see it without knowing whence it came, he would be almost sure to say that it belonged to a Paleozoic, or at least a Triassic, reptile, and not only to an early reptile but a very primitive one at that. This peculiarity consists in the absence of any opening between the ischium and pubis, which is characteristic of every living vertebrate with legs. And these and other old-fashioned characters could not possibly have been new developments; they must have existed in all the ancestors of the Choristodera from Paleozoic to early Tertiary times, though not a single other reptile is known to have possessed them, for the greater part of this time. Perhaps when Asia and northwestern America have been more thoroughly explored for vertebrate fossils, some of their ancestors which perished on their great migration from the western to