1899. J FROM THE LONDON CLAY OF SHEPPEY. 779 



Immediately in front of the rostral hinge the upper surface of 

 the beak is broad and slightly convex from side to side ; but as 

 it passes forward between the nares (n.) it becomes narrower and 

 more rounded, and continues to do so as far forward as preserved, 

 the tip being lost : between the nostrils a portion of the bone 

 is broken away showing that it is hollow. 



The nostrils («.) are very long and narrow, and extend backward 

 nearly to the rostral hinge. They are widest immediately in front 

 of the antorbital fossa, and thence run forward as a narrow cleft, 

 w])ich anteriorly probably becomes a groove {n.g.) : tlie anterior end 

 of the narial opening (or groove) is broken away with the tip of the 

 beak. Posteriorly the nostrils are separated from the antorbital 

 fossa by a bar of bone which no doubt is, as usual, formed by the 

 union of the downward process of the na^al with the maxilla. 

 The edge of the beak is formed by a bar of hone, which is narrow 

 anteriorly but widens out gradually from before backward, reaching 

 its greatest width just in front of the antorbital fossa. Behind 

 this it passes without interruption into the quadrato-jiigal bar {jug.). 

 The thickness of this bar seems to be constant throughout its 

 length, but the posterior extremity where it joined the quadrate 

 has been broken away. The interorbital septum (i.o.s.) is 

 incomplete, being perforated by a large quadrate fenestra (i.o.f.), 

 similar to that seen in Phaethon. It has been cleared of matrix 

 so that the thickened ventral edge (rostrum) is exposed, and it can 

 be seen that the anterior ends of the pterygoid.^ rest against it and 

 articulate with the posterior ends of the palatines as in most Cari- 

 nate birds. Unfortunately, it has not been possible to work out 

 the structure of the palate further. 



The quadrate (q.) is long and articulates with the skull by two 

 heads, the facets for which are situated very far back, immediately 

 within the rim formed by the lower edge of the paroccipital. On 

 the exposed side the outer half of the quadrate has been abraded, 

 and its distal articulation is still in its natural position with regard 

 to the mandible, so that its form cannot be observed. Eather 

 high up on the inner border of the bone there arises a fairly 

 large orbital process (o.p.) which projects upward and inward and 

 terminates in a blunt point. 



In the mandiUe the posterior end is so much injured that its form 

 cannot be determined, but there seems to have been a fairly promi- 

 nent postero-internal process. From its posterior end the ramus 

 increases in depth as far forward as about the middle of the orbit. 

 At this point there is a lateral vacuitj which is continued forward 

 as a shallow depression, most clearly defined above, and of which 

 the anterior end is opposite the nuddle of the nostril. In 

 front of this depression the mandible tapers towards its anterior 

 end. and its outer surface becomes rounded from above downward ; 

 the extreme tip has been broken away. It cannot be seen whether 

 there was a distinct coronoid process or not, because the upper 

 border is hidden behind the jugal bar. Beneath and behind the 

 lateral vacuity the suture between the dentary and the articulfir 

 region remains distinct. 



