290 CAPT. D. M. S. WATSON ON SEYMOURIA, 



spondyli in lying within the nostril and not appearing in the 

 face. The condition of this bone is known in few other Coty- 

 losaurs. 



Loioer jaiu. — The mandible of Seymouria is typically Laby- 

 rinthodont in structuie, agreeing very closely with that of 

 Trimerorachis, in possessing a post-splenial and three coronoids. 



The presence of a patch of small granular teeth on the second 

 coronoid is also a point of resemblance to most Kachitomi. 



The single large tusk on the 3rd coronoid is unique amongst 

 Tetrapods, being unparalleled in either Amphibia or Reptilia, but 

 of constant occurrence in Osteolepid fish. 



The non-fusion of the articular with the surangular is a point 

 in which Seymouria differs from the Temnospondyls and agrees 

 with the majority of reptiles. 



Vertebral column. — The atlas of Seymouria is 

 lateral compression and great antero-posterior 

 neural elements. This feature comes out most 

 large spine composed of two apposed halves, 

 neural elements together much more resemble a normal dorsal 

 neural arch than in any other known Tetrapod, and are un- 

 doubtedly extraordinarily primitive. The odontoid is thoroughly 

 i^eptilian. * 



The axis of Seymouria in its complete unspecialisation, resem- 

 bling as it does the vertebra next behind it, is quite different 

 from that of any other reptile, whilst it agrees with that of 

 Temnospondyls, where the 2nd and 3rd vertebi'te have a similar 

 resemblance to one another. 



Prof. Williston in a recent paper has given an account of the 

 mode of evolution of a Seymouria vertebra from an Embolo- 

 merous type, with which I am in perfect agreement. He shows 

 that the Seymouria vertebra is more primitive than that of any 

 other known reptile in having a larger intercentrum, only 

 slightly reduced from the disk which represents that bone in 

 Embolomeri. 



The presence of distinct jirocesses for the head of the rib in 

 certain Seymouria intercentra is a condition known in many 

 Rachitomi and occurring in no other reptile. 



The presence of a single sacral vertebra only is an Amphibian 

 feature very rare in reptiles. The wide separation of the centra 

 of the caudal vertebrae suggesting the presence of a ring-shaped 

 intercentrum, is a very primitive feature. 



On the other hand, the massive neural arches of the pre-sacrals 

 with horizontally placed and widelj' separate zygapophyses are 

 tj'pically Cotylosaurian, and occur in no Labyrinthodont what- 

 soever. The caudal neural arches are also of typical reptilian 

 type. 



Rihs.-r-^Ti possessing double-headed ribs throughout the whole 

 pre-sacral part of the vertebral column, Seymouria differs from 

 all other known Cotylosaurs and agrees with the Embolomeri. 

 There can be no doubt that two-headed ribs are primitive, and 



