EVOLUTIONARY SUMMARY 29 



Arrhinoceratops and the Triceratops species obtUSUS and hatcher), in which the nasal horn is reduced. 

 The closing member of this phylum is Torosanrus, which again fulfills the requirements in being just 

 what one would be led to expect from the evolutionary trend established by its predecessor . It i 

 strange that this genus should be so rare, with but two recognizable specimens found within a mile of 

 each other at the very summit of the Niobrara County Lance. The contemporary Triceratops, on 

 the other hand, is represented by hundreds of individuals' of which comparatively few were 

 collectable. 



The initial short-crested form is Monoclonius (Centrosaurus), from the Judith River and Belly 

 River formations, which may be the beginning of the entire line culminating in Triceratops, but the 

 sub-genus Centrosaurus itself, with the high degree of ornamentation at the rear of the crest, seems 

 too specialized. It may well be that the true ancestry lies in the ill-known forms of Monoclonius 

 or even Ceratops of the Judith River which seem to be more conservative in this regard. The con- 

 tinuity through the Edmonton is not good, for A rrhinoceratops is the only known possibility and the 

 objections to its inclusion in this phylum have already been given. Styracosaurus, with the riotous 

 processes around the rear of the crest, is clearly derivable from the Monoclonius stock but not from 

 any of the known forms, all of which are apparently too nearly contemporaneous. Triceratops is 

 evidently a polyphyletic genus, and I have shown at least three descent lines, or rather groupings, 

 within the genus. Of these, perhaps the most conservative is the prorsus-brevkornus-horridus 

 phylum. The sequence, as I have shown it, is not actual, for the type of horridus is next to the 

 lowermost of the Niobrara County series, although, if I am correct in referring the giant skull some- 

 times known as Triceratops "ingens" (No. 1828 Y.P.M.) to this species, it persisted until near the 

 close of the Lance record. As the skulls are arranged, they show progressive increase in size, 

 and reduction of the nasal horn. Of the elatus-calicomis group, the elatus type is somewhat the 

 older, but here again elatus has a long vertical distribution and the single ca/icoruis specimen may 

 be merely a variant from the type, but actually conspecific. The great brow horns and curiously- 

 reduced nasal horn are the most outstanding features of this group. T. obtusus, with the nose horn 

 nearly vestigial, is at the very beginning of the Niobrara County sequence; T. (Diceratops) hatcheri 

 is two-thirds of the way up. Whether obtusus was actually ancestral to hatcheri is questionable, but 

 except for smaller size, the latter seems to be the embodiment of the obtusus evolutionary trend. 



Brown reports 500 in the Hell Creek region alone. 1933, A, p. 2. 



