136 SHOULDER-GIRDLE AND BREAST-BONE. 



4. That in these latter portions the alse originate at an earlier period than the prolongations which 

 run parallel to the long axis of the body, by means of which they subsequently coalesce, and by uniting 

 with the remaining symmetrical pieces, compose a ring. 



In order to fill up with bone, to a greater or less extent, the remarkably extensive spaces only 

 occupied by connective tissue, which in all Chelonians originally intervene in the middle line between 

 the bony pieces of the Plastron, in many species some of these bones increase in breadth in this 

 direction, sending in rays or outrunners, as Cuvier has already remarked in his ' Recherches' (vol. ix, 

 p. 403), which either remain in this condition, as occurs in the marine Chelonians; or, increasing con- 

 tinually in size, coalesce in their entire length, until at length a solid flat table is produced. I 

 remarked several of these rays in the young of Chelonia proceeding from the bones of the third pair } 

 in the young Platemys, from the second pair, in the young Trionyx ocellatus from those of the second 

 and third pair, in the embryo of Testudo, and in the young of Emys europosa, E. lutaria, and Terra- 

 pene tricarinata, from those of the second, third, and fourth pairs ; in the last-named species they were 

 the most numerous from each pair of bones. On the other hand, I have never observed them to pro- 

 ceed from the first pair of bones, nor from the asymmetrical pieces. 



XXIX. As regards the process of ossification the several constituent pieces of the Plastron 

 comport themselves similarly, though not quite identically with the plates of the Carapace. Ossifi- 

 cation begins, as I have observed in the embryo of Chelonia, at about the centre of each piece, and 

 indeed in their interior, like the supplementary plates of the carapace, and not on the surface of the carti- 

 lage 1 \_gristle~] of which they are primarily composed. When it is quite impregnated with calcareous mate- 

 rial medullary canals are conspicuous, running or radiating outwards from the centre, sometimes branching 

 at acute angles, containing medullary substance, lying in layers two to five deep, and, according to their 

 position and length, either running to the margin of the pieces or opening upon the inferior surface. 

 The canals of the lowest (nearest the cutaneous covering) layers are the shortest, those of the uppermost the 

 longest. Still later, as in the back plates and ribs on that surface which is turned towards the skin there 

 appear in these Plastron-plates medullary cells, seated apparently perpendicularly upon the canals, and 

 in the first instance wide open externally. These medullary cells are filled, not with medulla, but with 

 a loose connective tissue, and in their further development and multiplication (increase) behave them- 

 selves like other parts of the skeleton. A quite peculiar process is, however, found to take place in 

 those cases where the several pieces of the Plastron have sent out several rays, especially in Emys and 

 Terrapene. In the angle between each two rays a plate is formed consisting of osseous substance, 

 which is in the first instance delicate, and often cribriform, apparently proceeding from the rays them- 

 selves, but by no means presenting the appearance of a superimposed mass lying on them. These 

 plates gradually become thicker, become wider also as they approximate the tapering extremities of the 

 rays, and come more aud more to till up the spaces intervening between them. As they increase in 

 breadth there appear upon them, and on the rays connected by them, medullary cells of the same kind, 

 and in the same method as has been already described in the remaining part of the Plastron. These 

 cells are. seen on that surface of the plates and rays which is turned towards the skin. I found 

 scarcely any medullary cells, but only horizontally running medullary canals in the Plastron of the 

 young of Chelonia and Trionyx, although in the latter species similar bony cells had already formed 

 in abundance in the ribs and spinous processes. 



XXX. I found the pieces of the Plastron just described (in all the Tortoises examined in a state 



1 Rathke does not here speak of hyaline cartilage, but of a dense mat or web of nascent 

 connective tissue, such as that composing the cutis vera. All splint-bones are formed in a " stroma" 

 of this kind, and in studying Rathke's writings it is very necessary to ascertain whether he is speaking 

 of this latter tissue or of true hyaline cartilage. 



