Structure of the T/ierocephallan Skull. 77 



Seeley, Phil. Trans, vol. 185 B, pp. 987-10i8, pi. Ixxxviii. 

 fig, 1) has a lower surface which, except for the fact that it 

 is not known to have any secondary |Kilate, is exf'.raordinarily 

 like that of Ci/nognat/ius or J)iadeinodon, especially in the 

 fact that the interptery<foiJ vacuity is entirely suppressed, 

 and the suborbital vacuities not developed. 



Fiff. 7. 



Figs. 6 and 7. — Dorsal and rifrht lateral views of the skull of Arcto- 

 gnathus curvimola (Owen), x f, restored from the type and only 

 known specimen. Areas with uniform shading- are completely 

 missing on both sides of the specimen. These tigiires will render 

 the lithographic drawings of the skull published by Owen more 

 intelliu-ible. 



Tills type is remarkable in that it has a broad parietal 

 region, so badly preserved, however, that its structure js not 

 known. It shows no features wiiich are inconsistent witii a 

 close connection with the Cynognathidje, using tliat term to 

 include Cynocjnathus, Diad'emodon, and Trirachodon, for 

 the broad cranial roof could easily be reduced to the narrow 

 ridge of the later type. 



At the same time this broad roof, in which presumably the 

 postorbitals extend back to the squamosals, suggests it 

 connection with the Gorgonopsids. 1 see nothing in the 

 palate of the two types to negative such an affinity, but the 

 extreme shortness oi' the preorbital part of the Arctognathus 

 skull is rather opposed to it. It must, however, be remem- 

 bered that in carnivorous animals after the establishment of 

 a carnivorous dentition there is often a tendency towards an 

 abbreviation of the face, for example, in Felis, Lutra, Patrio- 

 feh's, and amongst reptiles Tyranosaurus. 



