In the Turtles, and in the Trionyees, or Soft Tortoises, these radiating expansions do not unite throughout ; and even when the four pieces on each side unite together and the odd piece ig joined to those of the first pair, there remains in the middle, between them all and on each side between them and the dorsal buckler, great spaces which are filled up by cartilage only.
Vertebrx.—The atlas is composed of four pieces. The first two, united above in a slight spinous prominence, after having surrounded the vertebral canal, and each having given backwards its articular apophysis, concur with a third very small one in the formation of a ring for the reception of the condyle of the head : Cuvier calls it a ring, because in the skeleton this fosset is open, and its bottom filled by a fourth piece, which is a true body of a vertebra without the annular portion, and which, presenting an anterior convex surface in the space here noticed, is articulated behind by a concave surface on the body of the axis. This piece, analogous to what we have already seen in the crocodile, represents, he observes, the odontoid apophysis of the axis of mammals. At theirjunction, there is besides, attached below, a small bone formed nearly like a patella (rotule).
The axis and the succeeding vertebra) are composed of a nearly rectangular body, carinated below, concave in front, convex behind, and of an annular portion, which remains distinct from the body throughout life, by means of two sutures, is elevated above by a crest in lieu of a spinet's apophysis, and whose anterior articular apophyses, placed at first under the posterior portions of the preceding vertebra, raise themselves obliquely to embrace them slightly up to tho sixth, and nearly resume their horizontal position in the two succeeding ones. At the anterior angle of each side of the body is a small facet, common to the body and the annular portion. • The vertebra) adhering to the dorsal buckler have their body wide and feebly carinated in the Marine and Fresh-Water Tortoises : in these last it is even flattened in the anterior ones. It is also wide and with but little convexity in Trionyx, and Chigs has it wide and elevated longitudinally into a small crest. But there are Land Tortoises (Testudo yeometrica and T. radiata) in which it is exces sively compressed, and does not even join itself throughout, except by a membranous partition, to the pieces of the middle row of the buckler, these pieces only affording each two narrow laminte, and descending on each articulation of the two bodies. It is in a fold of the lower portion of this membrane, between these vertical lamiuie, and in a semicanal hollowed at the upper part of the bodies, that the spinal marrow goes.
In the other sub-genera the pieces of the longitudinal series of the dorsal buckler afford more complete vertical partitions, which form with the bodies a continuous bony canal, the nerves of which go out throegh holes which remain between the laminae.
The sacral and caudal vertehrie are each composed of a body, concave before and convex behind, of an annular portion, squarely flattened, and without a spine above, the anterior articular apophyses of which obliquely embrace below the posterior apophyses of the preceding vertebra, and of two transverse short apophyees, articulated on each side on the suture, which joins the body to the annular ring.
Cuvier counted 23 caudal vertebrae in Testae& Greece, T Indica, and other Land-Tortoises, and as many as 27 in Testado radiata. He states that there were only IS in the Fresh-Water and Marine Tortoises which he examined.
Bones of the Extremities.—The bone which goes from the dorsal buckler to the sternum is suspended by a ligament under the dilata tion of the second rib, but in front of the first, which, AA we have seen, consists only of a head articulated under the second ; so that in some respects this bona is outside the thorax. There is sometimes in the ligament by which it is attached one, and even two, peculiar bones. This berm is at first nearly cylindrical : it proceeds forwards, and after having afforded on its external snrfaee a portion of the articular facet which receives the head of the humerus, it goes with a more or less strong inward bend to attach its other extremity to the internal surface of the sternum, towards the lateral angle of the odd piece. The rest of the facet for the articulation of the humerus is furnished by another bone, which is directed more or less obliquely backwards and towards the menial line, widening into a fan-shape, and which thus lies nearly parallel to the sternum. The osseous branch which comes from the bony buckler, is, according to Cuvier'e self-corrected opinion, the shoulder-blade, and the part which it offers beyond the articular foaset is its acromion. The flattened bone which is directed backwards is, ho adds, incontestably the coracoid bone : and ho further remarks that all the muscles which proceed from these bones to go to the arm are respectively the same am in birds, whatever changes they have underrne in their position relatively to the horizon In their size and in their figure. Cuvier considers that it ramie* to be known whether there is a clavicle or not.