THE SKULL 37 



processes extending outward and backward from the occipital condyle, and bounded posteriorly by a 

 suture crossing from wing to wing about midway along the length of the latter. Just posterior to 

 the exoccipitals is a small bone called by Marsh and Hatcher the supraoccipital (sec Fig. 5). 

 Beyond this the parietal extends to the posterior end of the crest, at first broad and concave, but 

 flattening toward the rear. Brown interprets these structures differently, making the supraoccipital 

 a bone bounded by the foramen magnum, the wings of the exoccipitals, and a line drawn transversely 

 between the expansions of the exoccipitals. Behind this is his small parietal, with what he terms the 

 postfrontal (our parietal) forming the remainder of the middle portion of the crest. Our nomencla- 

 ture, however, is based on the following points: the apparent lack of a suture across the foramen 

 magnum defining an extra bone here; furthermore, it has always been held that the parietal should 

 form part of the roof of the brain case, which Brown's so-called parietal does not do. 



Lower Jaw 



(Fig. 4) 



The jaws in Centrosaurus are more slender and less massive than in Triceratops. Posteriorly, 

 there is a high coronoid process which curves forward toward its summit into a hook-like process more 

 marked in M. sphetweerus than in the Yale specimen. The main portion of the mandible, as usual, 

 is formed by the dentary, on the lower and inner surface of which is a large splenial extending to 

 the hinder end of the mandible. Externally, the angular forms a triangular plate of bone rising to 

 about half the depth of the mandible, exclusive of the coronoid, hence relatively larger than in 

 Triceratops prorsus. Above this lies the surangular, which forms the rear margin of the jaw except 

 where it meets the articular below. The articular is well developed although its external limits are 

 not very clearly defined. Externally, the modeling of the jaw is characteristic. There are two 

 longitudinal hollows, separated by a rounded horizontal ridge. Of these, the upper one is much the 

 deeper and runs within the coronoid process toward the rear of the jaw. Vascular impressions, well 

 shown along the inferior margin of the dentary and angular in the M. sphenocerus type, are only 

 faintly indicated in the Yale specimen, evidently that of a younger animal. This is in harmony with 

 the general surface of the skulls as well. The type of C. flexus does not possess a lower jaw. 



Predentary 



This bone in Centrosaurus resembles that of Triceratops except for one detail. The upper 

 surface which is defined by two longitudinal ridges is horizontal in Triceratops, but in Centrosaurus 

 is inclined outward at an angle of about 45°. Externally, the predentary resembles the rostral in its 

 vascular surface. 



Hyoid 



(Fig. 6) 



The final preparation of the wonderfully preserved type skull of Triceratops serratus Marsh in 

 the Peabody Museum of Yale University has brought to light a bone hitherto unknown among the 

 horned dinosaurs and rarely observed in the entire dinosaurian order. 



The hyoid, if such it is, lies nearly in situ alongside of the right ramus of the lower jaw, opposite 

 its hinder end. It is of very peculiar form, being sharply curved so as to form a nearly complete 

 semicircle, and having a thin, blade-like expansion set at an angle of about 35 with the main mass 

 of the bone The general form, seen from below, is shown in Figure 6, together with its position 

 relative to the adjacent jaw. The bone is evidently one of a pair, the articulating [surface being at 

 the end opposite the blade. This articular surface thins out to a moderately sharp edge but is slightly 

 roughened As it lies, the bone extends toward the left beyond the median line of the head. It 

 seems, however, to have been rotated through an angle of 90° rather than actually displaced. With 



