ON THE STRUCTURE OF THE LABYRINTHODONTS. 211 



examples afford evidence of cartilaginous vertebral extremities, this evidence 

 consisting of the separation of the proximal ends of the ribs from the vertebral 

 column by a regular interval, and the hollowing-out of the ends as if for 

 junction with cartilage *. At this stage a transverse process may be seen to 

 project for a short distance from the lamina of the corresponding sujDcrior 

 arch. The junction is not completed by a true bony articulation until the 

 animal is nearly adult. 



Shoidder-rjinlle. — The shoulder-girdle of the Labyrinthodonts includes 

 three thoracic plates (which represent the clavicles andinterclavicles), one or 

 more scapular bones, and a coracoid. In form and arrangement these parts 

 differ much from the pectoral arch of any recent Amphibian, but correspond 

 generally with the structure which prevails in some Eeptilia, such as the 

 Lacertilia (e. g. Trachydosaurus, Monitor, Iguana) and the Ichthyosauria. 

 The resemblance between the shoulder-girdle of the Labyrinthodonts and 

 that of the Ichthj'osauria is close and striking. 



The thoracic plates are eminently characteristic of the true Labyrintho- 

 donts. They are three in number, a median and two lateral. The median 

 plate is elongated, and more or less rhomboidal ; it is placed lojigitudinally. 

 On each side it is overlapped by the lateral plates to a considerable degree, 

 especially upon the antero-external borders ; and frequently only the hinder 

 end is exposed. The free part ordinarily exhibits sculpturing. The lateral 

 plates have been compared as to form to the elytra of beetles. They are 

 often, but not always, triangular in form — the base, which is directed inwards, 

 being rounded, and the remaining sides set at an angle of 90 degrees or more. 

 A sculptured pattern is sometimes seen to radiate from the angle ; and this is 

 the thickest and strongest portion of the plate. 



The thoracic plates extend nearly from side to side, and may protect a 

 third, or even more, of the ventral surface of the trunk. They vary greatly 

 as to form and relative size. 



The median plate represents the interclavicle, and the lateral plates the 

 clavicles. All are dermal bones, forming no part of the true axial and ap- 

 pendicular endo-skeleton. 



Behind these (that is, nearer to the pelvic arch) and in a deeper plane are 

 the remains of the scapula and coracoid. These are most completely pre- 

 served in Archegosaurus, and much resemble the corresponding parts in the 

 recent Sir-en. 



The coracoid is ventrally situate, semilunar in form, having a concave 

 thickened posterior margin, a thickened postero-external angle, and a regu- 

 larly rounded anterior edge. There is no reason to suppose that this does 

 not retain, approximately, its natural position. On the outer side of the 

 coracoid there lies in an oblique position a long, narrow, flattisl) slip of bone ; 

 its posterior end, which is expanded and a little twisted, is adjacent to the 

 postero-external angle of the coracoid ; while the other or anterior end is 

 produced at great length forwards and inwards, generally passing beneath the 

 thoracic shield. Another bone, which may, however, be a detached part of 

 the same, is seen in several examples of Archetjosaurus. It lies somewhat in- 

 ternal to the last described bone, immediately behind the edge of the thoracic 

 plates, and has a slightly expanded end f. There can be little doubt that wo 

 have here a scapula, and probably a suprascapular bone also. The glenoid 

 cavity was probably cartilaginous in Avcliec/osaurus, and is not shown in the 



* Eeptilien &c., t. iv. fig. .5, and t. vi. fig. 10. 



t This end is directed backwards (J. e. towards the pelvis). The other extremity is not 

 shown. 



1873. R 



