OSTEOLOGY OF CARNIVOROUS DINOSAURS. 81 



lation of the parietal wi^h the postorbital, postfrontal complex, posterior to the 

 supratemporal fossa, is not known so far as I can determine in any reptilian skull 

 either living or extinct. 



Frontal (/.). — The frontal hones are moderately broad with a convexly rounded 

 superior surface as compared with the slightly concave surface of Antrodemus. 

 The suture between the frontals and nasals is more strongly toothed than was rep- 

 resented in Marsh's superior view of the skull. The median suture between the 

 frontals of the two sides, as well as the suture between frontals and parietals is 

 obliterated. The frontal as in Antrodemus contributes slightly to the superior 

 border of the orbit, in a notch between the prefrontal and postfrontal + postorbital. 

 In the restoration of the skull by Marsh it has been incorrectly restored showing 

 the frontal as extending outward to complete the lateral boundary above the orbit 

 on a hne with the two elements just mentioned. 



Prefrontal (pf.). — Interposed between the lachrymal and the frontal on the 

 right side of the skull of Ceratosaurus there appears to be a small triangular element, 

 identified as the prefrontal. The outlines of this bone are obscure, but if what 

 are interpretated as sutures are such, the hinder end of the dorsal surface of this 

 bone is relatively wider transversely and considerably shorter antero-posteriorly 

 than in Antrodemus. It appears to reach the nasals although in Antrodemus the 

 inner anterior border is in contact with that element for some distance. In Ty- 

 rannosaurus the anterior ends of the prefrontal just reach the nasals though they 

 are entirely excluded from the lateral border by the lachrymals. In Ceratosaurus 

 their posterior end forms the greater portion of the anterior boundary of the notch 

 above the orbit (fig. 1, pi. 18). The lower or descending branch of the prefrontal 

 is entirely covered by matrix on the right side and its extent and relationships 

 are not to be observed, and on the left which has been more thoroughly prepared 

 1 am unable to differentiate it. Marsh and Hay did not recognize this bone, but 

 called prefrontal the element here designated lachrymal. 



Postfrontal +postorbital (po. /.). — The postfrontal postorbital complex is of the 

 usual triradiate form, its relation to the other elements of the skull is so similar 

 to Antrodemus that a detailed description is considered unnecessary here. 



This bone is well shown in two of its aspects in figures 1 and 2, plate 18. There 

 is no "rugosity" such as found by Osborn in Tyrannosaurus , and as mentioned 

 previously the union of this bone with the squamosal is not by a transverse suture 

 as represented by Marsh in his restoration of the skull, but a posteriorly directed 

 process laps along the N outer surfaces of the squamosal uniting by a long squamous 

 suture. 



LacTirymal (la.). — The element here identified as lachrymal was designated by 

 Marsh as prefrontal, and by Hay, the lower part was called lachrymal and the upper 

 prefrontal, the two being separated by a diagonal suture. Osborn in Allosaurus 

 considered it a complex of prefrontal and lachrymal. This suture as represented 

 by Hay (fig. 54) is certainly a fracture of the bone, and a critical examination of 

 these elements fails to disclose any indication of there being a suture at this point. 



The lachrymal as in Antrodemus and Tyrannosaurus has a descending process 

 which unites the jugal and maxillary above the large antiorbital fenestra. The 



