68 MESSRS. HANCOCK AND ATTHEY ON 



slightly lengthened in the antero-posterior direction. In the 

 larger specimen it is |^th of an inch in length ; in the other it is a 

 little shorter. The occipital margin is concave, the epiotie bones 

 projecting boldly backwards, and each terminating in a slightly 

 arched, pointed, diverging horn or spine, about an inch long. 

 In neither specimen are these horns (which are very similar to 

 the " postero-internal cornua" of Keraterpeton^) perfect; but in 

 the smaller individual the upper surface only is wanting. 



The external surface of the cranial shield is strongly sculp- 

 tured in the usual manner observed in Labyrinthodonts ; that 

 is, it is covered with anastomosing ridges separating pits and 

 grooves. But this peculiar ornament is not equally distributed ; 

 it radiates from the centre of the shield, where it is almost ob- 

 literated, and is strongest at the margins. Here the pits and 

 grooves are deep and strongly defined. A rather wide rounded 

 groove extends along the outer margin of the frontals, resem- 

 bling the mucus grooves of the Labyrinthodonts. 



The larger of these shields is seven inches long, including the 

 posterior horns, and 3f inches wide. The other is 6 J inches 

 in length, and nearly three inches wide at the broadest part. 

 Prof. Huxley estimates the width of the skull of Loxomma All- 

 manni, including the lateral portions, which are entirely want- 

 ing in our specimens of Pteroplax cornuta, at lOf inches ; and 

 as it appears that the central portion, or that which corresponds 

 to the cranial shields above described, is about one-third the 

 entire width, we are enabled to form an approximate estimate of 

 the width of the skull of the new form, on the assumption that 

 it had similar lateral cranial expansions. On this basis our new 

 Labyrinthodont must have had a skull 11 J inches wide at the 

 posterior or widest part ; and, following up Prof. Huxley's esti- 

 mate, it could not be less than 15 inches in length. If the body, 

 therefore, was only seven times the length of the cranium, which 

 is about the proportion of these parts in Kerateipeton Galvcmi, a 

 comparatively short species, then Pterojdax must have been eight 

 or nine feet long. 



This new genus, though it seems, as above stated, to be nearly 



♦"Transactions" Royal Irish Academy, Vol. XXIV., "Science," p. 351, Plate XXI. 



