UPON HYPERODAPEDON GORDONI. 



685 



Fig. 7. — Hyperodapedon Gordoni. Fig. 8. — Hyperodapedon Gordoni. 



Mr. Grant's specimen. Mr. Grant's specimen. 



Left side. Eight side. 



r.m. 



Dr. Woodward has afforded me the opportunity of examining a 

 very instructive, though fragmentary, specimen of the fore part of the 

 skull of a large Hyperodapedon, the property of Mr. Grant, of Lossie- 

 mouth. The premaxillary rostrum is nearly entire and in connexion 

 with the facial bones, as far as the anterior boundary of the orbit on 

 the left side. The anterior end of the left ramus of the mandible, with 

 its rostral process nearly entire, is almost in place (fig. 7), and on the 

 opposite side of the matrix there is the almost complete impression of 

 the right mandibular rostral process (fig. 8). Moreover, a cast of the 

 outer surface of the left rostral process of the premaxilla is preserved 

 in one of the halves of the matrix. The left rostral process itself is 

 quite free, and can be lifted out and examined on all sides. In- 

 cluding the anterior extremity, of which only the impression remains, 

 it is 57 millim. long, 23 millim. deep, and 13 millim. thick at its 

 proximal end. The anterior face is more strongly curved than the 

 posterior, and somewhat flattened, and it is rather obtusely pointed at 

 its free end. The inner face is flat, and was evidently closely applied 

 to that of the opposite rostral process in the greater part of its extent ; 

 but, in the course of fossilization, a thin layer of matrix has been 

 interposed between the two. About the posterior third, the inner 

 face of each rostral process slopes sharply outwards, and the upper 

 margin becomes rounded as it bounds the external nares. Both the 

 inner and the outer surfaces of the bony substance of the rostral 

 processes "are marked by fine close-set longitudinal grooves, and the 

 surfaces of the rostral processes of the mandible present similar 

 markings. 



The convex face of the left premaxillary rostral process fits closely 

 into the concavity of the cast, to which I have referred, and presents 

 fine longitudinal ridges which answer to the grooves on the surface 



