336 DR JOHN BERRY HAYCRAFT ON THE 



My own work on this subject is followed out very much on the same lines ; but I 

 have endeavoured to fill in certain gaps left in his description. I am obliged to differ 

 from him in what is to my mind his most important conclusion he has arrived at. I 

 shall endeavour to show that there is no true periosteum (perichondrium) around either 

 the primitive cartilaginous rib, or the primitive cartilaginous vertebrae, and that the 

 costal and neural plates do not therefore arise outside the periosteum as he affirms. 

 Whereas in the growing rib or vertebra, say of a crocodile, we find an investing perios- 

 teum, the inner surface of which is alone osteogenetic, the periosteum of a turtle's rib 

 consists only of a few fibres of the osteogenetic membrane, of which the carapace is com- 

 posed, arranged loosely around the bone. In the case of the crocodile's rib, the deposition of 

 bone, being on the inner surface of a cylindrical osteogenetic tube, leads to the formation 

 of a cylindrical rib, whose girth is limited by the growth of the periosteal tube. In the 

 case of the turtle there is no such limiting membrane, and the bone forming first on the 

 surface of the cartilage spreads outwards in all directions, gradually involving the whole 

 carapace. This is a short, and as I believe an accurate description of the formation 

 of the neural and costal plates. I shall endeavour to amplify this statement in the fol- 

 lowing pages. When the facts of the developmental process are thoroughly grasped, 

 it will be possible to discuss with advantage the relation which the costal and neural 

 plates bear to the ribs and vertebrse. 



Thanks to the embryo turtles kindly given me by Professor Moseley, I have been 

 able to investigate the development in its earliest stages. The subsequent steps have 

 been worked at from a very nearly complete series of fresh- water tortoises. The carapace 

 was hardened either in alcohol or picric acid. A very convenient method when dealing 

 with animals that have already begun to develop osseous tissue is to place the carapace 

 in picric acid, to which a few drops of hydrochloric acid have been added, when the 

 carapace will soften in a few days, and sections may at onoe be prepared. 



Study of a Series of Sections Transverse to the Length of the Carapace (Plate, 



figs, 7-10). 



The youngest embryo that I have been able to obtain was a green turtle, measuring 

 about 1 centimetre in length. The vertebral cartilage in this embryo is entirely 

 cartilaginous, and so are the ribs. The cartilage is in all situations surrounded by 

 embryonic tissue, such as is found in the rest of the carapace. There is no special 

 periosteum, for one can hardly speak of the tissue next the cartilage, which has some- 

 what of a set around it, as a periosteum. Dorsal muscle plates are evident in this 

 embryo. 



Each vertebra is composed of three pieces. The first encloses the notochord, forming a 

 complete ring around it. The other two encircle the neural canal laterally and posteriorly. 

 The cartilaginous ribs are attached to the vertebrae by fibrous tissue. 



In a somewhat more advanced embryo (fig. 7), the lateral pieces of cartilage have 



