MONOCLONIUS DAWSOM 85 



do not possess a crest ( Monoclonius), or the specific distinctions seem to lie in other parts of the skull, 

 the crests alone being spedficall) indistinguishable from the genoholotype, or from one another 

 Utrosmmu). Lambe's original description stressed the curved, hook-like processes at the posterior 

 end of the saddle-shaped frill, hence Centrosaurus. He did not, however, recognize the forwardly 

 projecting process as such, for it was detached and hence mistaken for a rather peculiar nasal horn. 

 Later, it became evident that this horn fitted accurately on a fractured portion of the posterior parietal 

 bar, and it was then cemented into place. The right one only is present, however, there being no 

 sign that the animal ever had possessed the left one. That was later found to be of occasional 

 occurrence (cf. Monoclonius (Centrosaurus) flexus type). 



The sub-genus Centrosaurus is now known in its entirety, even to the hide (PI. Ill, B), and the 

 description of it appears in detail in the morphological portion of this memoir (pp. 30-66). We have- 

 already stressed the sub-generic characters in distinguishing Centrosaurus from Monoclonius ; but 

 the specific distinctions separating the various centrosaurs are not so clear, and seem to lie mainly 

 in the character of the nasal horn, whether straight or forwardly curved. The named species are: 

 Centrosaurus a perl us Lambe, genoholotype; C. flexus Brown; C. nasicornus Brown; and C. cutleri 

 Brown. Of these, C. nasicornus is based upon an entire skeleton with skull and jaws, all of a single- 

 individual, and is, therefore, entirely adequate. Of C. flexus, on the other hand, the type consists 

 of a skull only, without the lower jaws, or other skeletal elements; while C. cutleri is based upon a 

 partial skeleton with the merest fragments of skull which cannot be compared with the flexus or any 

 other skull. In other words, there is no assurance that these two species are distinct, as the types 

 contain no homologous parts for comparison. Furthermore, the Yale specimen, again a single 

 individual, possesses a flexus skull while the skeleton is more suggestive of cutleri, as the descrip- 

 tions show. I am inclined, therefore, to consider C. cutleri as possibly invalid, which reduces the 

 species to three, or possibly two, since neither flexus nor nasicornus can be distinguished from apertus 

 on the basis of the crest alone. 



Characters, shown by the skulls, which may be contrasted and are hence of specific value, are 

 as follows: 



The nasal horn is always well developed, much more so than in Chasmosaurus, and is either 

 erect as in nasicornus, or curved forward as in flexus. The only variant known to me is in skull 

 No. 4519, at Toronto, in which it is erect but curves slightly backward toward the tip. 



The brow horns are never large in the Belly River forms, except in Chasmosaurus kaiseni 

 (PI. V, A). In Centrosaurus, they have the same degree of development as in the other chasmo- 

 saurs, and vary from mere rugosities over the orbits, as in nasicornus, to pointed, trihedral horns, of 

 which the height about equals the fore and aft diameter at the base, as in the Toronto specimen, 

 No. 4519, mentioned above. Strangely enough, in both the flexus type and that at Yale, as well as 

 in a curious skull, No. 348, at Ottawa, one horn is developed to a much greater extent than is the 

 other and, through coincidence or otherwise, in each of the three it is the left which is the larger. 



The forwardly directed processes at the rear of the crest may be straight or have varying degrees 

 of curvature. An extreme instance is shown in the skull No. 348 referred to, in which the curvature 

 of both the nasal horn and the processes reaches a maximum. The posterior processes of the crest 

 also vary in their degree of curvature, as well as in their distance apart at the base. The crest is 

 always saddle-shaped, but varies in the degree of transverse curvature, from rather flat in the apertus 

 type to the extreme curve of the greater number of skulls. This may be due in part to the orienta- 

 tion of the specimens in their original position in the strata. The apertus crest was, I believe, 

 detached and lying flat, whereas most of the specimens lie on their side when found, which, in either 

 case, must influence the ultimate form of the fossil due to crushing. It does not seem as though this 

 feature can be of specific significance. 



Finally, the muzzle, while always short and deep as compared with that in Chasmosaurus, 

 nevertheless varies, especially in profile and in the degree of rugosity of the part immediately above 

 the rostrum. F.\treme curvature and rugosity is seen in the type of nasicornus, and also in specimen 

 No. 8897 U.S.N.M., which agrees with the type in other details as well. 



