OSTEOLOGY OF THE SKULL OF THE DIMETRODON 309 



premaxillary suture the prevomers are free from the side walls of the 

 skull, leaving the elongate openings of the posterior nares. The 

 sides of the posterior nares are marked on the prevomers by rugose 

 ridges. The prevomers are united on the lower surface, but the 

 upper portion is divergent, and receives anteriorly the lower edges 

 of two vertical plates that seemingly originate from the inner edges 



Fig. 5. — Section of the same skull showing septal bones. Letters as in Fig. 2. 



of the pterygoids and extend directly upward in the skull. Owing 

 Xo the somewhat crushed condition of these very slender plates, 

 it is impossible to tell exactly the point of their connection with the 

 prevomers below, but apparently they lie between them, and there 

 was either a squamous contact, or the bones were free in life and have 

 been crushed together in fossilization. The origin of the two ver- 

 tical plates is somewhat obscure. They occupy the position of 

 vomers behind the prevomers, but the true vomer is a single median 

 bone, and, moreover, is accounted for. Broom^ has described in 

 Proterosuchus two slender vertical plates rising from the inner edges 

 of the pterygoids, with a single median vomer between. It seems 

 that these plates must be the same sort of a structure. They occupy 



I R. Broom, "On a New Reptile {Proterosuchus fergusi) from the Karoo Beds of 

 Tarkastad, South Africa," Annals of South African Museum, Vol. IV (1903), Art. 7. 



