52 



THE CERATOPSIA 



These sternal plates have been figured and described by Brown. 24 They "are paired, elongate, 

 and symmetrical except where distorted by crushing. One end terminates in a point which is 

 restricted and enlarged with the rough portion most extensive on the visceral surface, and stands 

 away from the body of the plate. The opposite end is thickened and rugose at the outer angle but 

 thins rapidly toward the inner border which is very thin and nearly straight. The outer border is 

 thickened, smooth and slightly concave. The ventral surface or outside of the plate is convex trans- 

 versely while the dorsal or visceral surface is concave. Both are smooth. A deep incision on the 

 visceral surface near the broad end of each plate may represent a foramen or a deep ligamentous 

 attachment." Apparently "the pointed end extended forward with the enlarged point attached to 

 the terminal curved portion of the coracoid. When placed thus with the thin inner borders meet- 



a. 



Y.P.M. 2075 



Fig. 20. — Sternals of Monoclonius {Centrosaurus) flexus. Ventral view, 1/6 natural size, a, anterior; -p, posterior. 



ing in the median line the width of the body cavity is determined at this point and the position of 

 the coracoids is approximately fixed. ... In the Jurassic Dinosaurs, Brontosaurus, Morosaurus 

 and Di-plodocus, the sternal plates were firmly united in the median line by cartilage as indicated by 

 the long straight rugose inner border. But Triceratops presents an entirely different type of sternum 

 in which the median borders of the sternals as well as the body of the plate are extremely thin and not 

 coossified. . . . The deep anterior emargination of these plates indicates the presence of an inter- 

 clavicle." Subsequent, very complete ceratopsian skeletons have not, however, disclosed this ele- 

 ment. "The external third of the plate is thick and formed its chief strength, while the uniformly 

 smooth surface shows that there was no lateral attachment of cartilaginous ribs as in the Ratitae. 

 This rib attachment was clearly formed at the posterior end of the plate, probably direct with- 

 out the intervention of a xiphisternum, for the thick posterior end seems subdivided as for rib 

 attachments." 



The two sternal bones of Centrosaurus flexus (No. 2015 Y.P.M.) (Fig. 20) do not appear to 

 be mates, as they differ so markedly in contour, and in color and general appearance, doubtless as 

 a result of conditions of preservation. The left, which is much the shorter of the two, retains 

 apparently more of the original form, as there is little restoration on it. To what extent its outline 

 is eroded away can not be determined. It is in essential agreement with the sternal in Triceratops 

 and in C. nasicomus except for an emargination of the inner border, where the bone is extremely thin 

 and may therefore have lost some of its former extent. The right, however, is relatively longer and 

 narrower and lacks the emargination, so that in this latter respect it is in closer agreement with 

 Brown's figures. It resembles that of C. cutleri except that the distal end is not so wide. 



24 Brown, B., 1906, pp. 297-300 



