130 THE CERATOPSIA 



This species cannot now be denned in terms of the criteria we have been using, for the skull 

 is absent. A careful comparison of the description and dimensions given by Brown fails to dis- 

 tinguish the vertebrae from those of Triceratops "ingens" (No. 1828 Y.P.M.) in so far as equivalent 

 bones are present in the two specimens, except that in certain dimensions cervical IV of maximus is a 

 trifle larger, and the three equally spaced foramina on the side of the centrum are represented by one 

 large one and others less clearly defined and not so regularly spaced. The other specific characters 

 mentioned are common to both. In the dorsal vertebrae of T. "ingens" the centra have either been 

 destroyed or are as yet hidden in the huge mass of matrix which envelops the specimen. 

 T. "ingens" is a gigantic representative of the horridus species, but features other than size which 

 might distinguish them are not now evident. 



Triceratops maximus may prove to be but another huge specimen of the same species. The 

 discovery of a skull sufficiently complete to show diagnostic characters, and associated with similar 

 vertebrae, is necessary for final proof either of uniqueness or synonymy with previously described 

 forms. 



Genus TOROSAURUS Marsh 94 



The known characters of Torosaurus lie in certain regions of the skull, for the skeleton, except 

 for a few limb bone fragments, is totally unknown. While material pertaining to the genus may 

 exist in our collections, it cannot be recognized. The skull characters have to do first with propor- 

 tions, for the portion anterior to the orbits is abbreviated, and the crest extended in width and 

 especially in length. As a consequence, the distinctive squamosal bone is extremely long and narrow, 

 recalling that of Chasmosaurus of the Belly River, and Pentaceratops of the Fruitland, both of which 

 may be in the line of its ancestry. Yet another generic character is the persistence of parietal 

 fenestrae, which are characteristic of all ceratopsians below the Lance, but absent in all known Lance 

 genera other than Torosaurus. In neither skull is the muzzle preserved, so that one cannot deter- 

 mine whether it was elongated, as in most long-crested forms, such as Chasmosaurus and Penta- 

 ceratops, or short, as in the average Triceratops. In the restoration of the Torosaurus gladius skull 

 (PI. XVI), the cranium and muzzle of which were modeled by the author a number of years ago, 

 Triceratops was used as a guide. Torosaurus certainly could not have had a deep, abbreviated 

 muzzle such as characterizes the short squamosal forms, Monoclonius and Centrosaurus. 



But two specimens of Torosaurus are known, both of which come from the very summit of the 

 Lance series of Niobrara County, Wyoming. 



Torosaurus latus Marsh 95 



PI. XV 



Holotype: 96 No. 1830 Y.P.M. ; skull without lower jaw. 



Horizon: Lance formation, No. 1 in the sequence. 



Locality: Niobrara County, Wyoming, near the summit of the bluff on the north side of Lightning Creek, about 



2 miles above its confluence with Lance Creek. 

 Collector: J. B. Hatcher, 1891. 



This skull is that of an old individual, about 7 feet 1 l / 2 inches over all, as now restored by the 

 author. The tntire cranium, including the left maxillary, the nasal horn, and the bases of the 

 brow horns, is preserved, together with both squamosals, and the anterior third of the fused 

 parietals. On the left side, a portion of the antero-external outline of the fenestra seems to be 

 present, but the form and extent of these distinctive features are highly conjectural. The dorsal 

 profile of the nasals is in line with that of the preserved part of the parietals. There is reason to 

 believe that beyond this the crest curved upward, balancing the upward curve of the nasals forward 

 into the nasal horn. 



94 Marsh, O. C, 1891, p. 266. 



95 Marsh, O. C, 1891, p. 266. 



90 Hatcher, Marsh, Lull, 1907, Fig. 118. 



