TRICERATOPS FLABELLATUS 121 



horn is somewhat elliptical, but the section becomes more nearly circular, distally. The crest 

 proportions are, at first sight, unusual, for it seems much wider in proportion to its length, than 

 in prorsus for instance. This is due, however, to the transverse curvature of the crest, which 

 differs in each case, that of serratus being flattened, while the prorsus crest is highly convex. As a 

 matter of fact, the relative measurements, taken oxer the curve of the bone, are much the same. 

 This is true also of the undulations of the midline, although the name serratus was based upon the 

 "series of bony projections on the median line of the parietal crest." The name could, with equal 

 propriety, be applied to prorsus and brevicortHts, on this detail alone. The very slight elevation of 

 the midline of the crest toward the rear agrees with fabellatus and no other species. To what 

 extent this may be due, in each specimen, to immaturity is not clear. The number of vertical rows 

 of maxillary and mandibular teeth is about 28 in each instance, which is the average number for 

 Triceratops species. The number of epoccipitals is 1 7. 



The mandible, which is somewhat pathologic in the right ramus, is long and slender, in 

 harmony with the facial portion of the skull. The predentary bears the same proportions, while the 

 coronoid process is only moderately expanded fore and aft at the summit. 



Triceratops flabellatus Marsh 72 



Holotvpe:"' 1 No. 1821 Y.P.M.; nearly complete skull associated with several vertebrae, a few limb bones, etc. 

 Horizon: Lance formation, No. 10 in the sequence. 

 Locality: Niobrara Countv, Wyoming. 

 Collector: J. B. Hatcher, 1889. 



The type is a very young, partly disarticulated skull, of which the length is 6 feet 2 inches. 

 The depth and relative shortness, which seem to be juvenile proportions, indicate that had the 

 animal lived, it would have attained a fairly large size, probably above the average. These general 

 proportions would presumably have changed with age, and hence are not considered specific 

 characters. 



The dorsal profile is much straighter than usual, that of the face and crest being nearly in 

 line. The nasal profile is a gentle curve, the degree of which remains the same throughout, 

 instead of varying as is usual. The nasal horn is missing, as in the serratus type. Marsh's restora- 

 tion shows, in outline, a truncated horn with an anterior margin which slopes somewhat backward. 

 It might, however, have been restored with equal propriety, to resemble that of brevicoruus. In 

 other words, this essential specific feature is lacking. The brow horns rise more erectly than in any 

 other species, except Triceratops (Dkeratops) hatcheri, but their forward curvature above the base 

 is somewhat greater than in the latter. The horns are laterally compressed near their base; distally, 

 they are more nearly circular. The orbit is an elongated ellipse of slightly irregular outline, with 

 the long axis inclined at 45° from the perpendicular, the maximum inclination for Triceratops. The 

 descending limb of the jugal is quite broad, and lacks a median ridge or keel. It resembles that of 

 serratus more nearly than that of any other. The jugal notch is not deep, but the infratemporal 

 fossa is a broad oval with but one angle, the antero-inferior one. The face is deep, but of moderate 

 length, while the muzzle is very short. The rostral and predentary are both missing. The crest, 

 while wide for its length as measured over the curve, appears narrower, due to its curvature, than in 

 either serratus or calicornis, wherein the crest is comparatively flat, when viewed from above. It is 

 smooth both above and below, with no trace of vascular impressions, except immediately in the 

 wake of the supratemporal fossa. The epoccipitals are 19 in number, which is near the average, 

 while the number of maxillary teeth, apparently about 35, seems excessively great, especially as the 

 number of mandibular teeth, apparently 30, is near the average and one assumes that they would 

 approximately correspond. 



"Marsh, O. C, 1889, B, p. 174. 



73 Hatcher, Marsh, Lull, 1907, Pis. XLIV-XLV. 



