'666 



Fig. 6. 



FifiT. 7. 



Fig. 8. 



Fig. 6. Shoulder girdle of Melanerpeton spiniceps, Ceed. Restored from Crednfe. 

 Fig, 7. Shoulder girdle of Palaeohatteria longicaudata Cred. Restored from Credner. 

 Fig. 8. Scapula and Coracoid of a young Sphenodon punctatum, Gray. 

 cl clavicle, icl interclavicle. clei cleithrum. sc scapula, c coracoid. 



distally somewhat broadened. The element called coracoid by Credner 

 and scapula by Zittel is doubtless the scapula ; the coracoid remain- 

 ing unossified in all of the Stegocephali, Sclerocephalus possibly 

 excepted. The scapula of the Stegocephali has the same general form, 

 as that of Palaeohatteria of the Proganosauria where the coracoid is 

 ossified as a small round disc, which doubtless was surrounded by 

 cartilage. The cleithrum disappears in the Amniota, but it seems to be 

 present in Pareiasaurus Baini (epiclavicle Seeley). The clavicles 

 of Polypterus, which form triangular plates, meeting in the middle 

 line, are not unlike the clavicles of some of the Stegocephali. There 

 is, it is true, no interclavicle in any of the Teleostomi; but there is 

 little doubt that in the crossopterygian ancestors of the Stegocephali 

 this element was present. This we may expect with some certainty 

 even; since the jugular plates in the Crossopterygia are either re- 

 presented by a single pair (Holoptychidae) or by a pair and an anterior 

 median piece (Rhizodontidae) ; or the median unpaired plate may be 

 present or absent (Osteolepidae). Just as a median element may 

 be developed between the jugulars, so it may also be developed 

 between the clavicles. 



I have already stated above that the paroccipital plates (supra- 

 temporals, Teleostomi) were probably connected by ligament with the 



