PROGRESSIVE EVOLUTIONARY CHANGES AND PHYLOGENY 



(Fig. 3) 



The recorded evolutionary history of the Ceratopsoidea opens abruptly, with several highly 

 interesting early chapters not yet revealed, for when these grotesque animals appear in our records 

 they are already ceratopsians with the essential features established. Not only so, but their differen- 

 tiation into at least three phyletic lines has already been brought about. 



The Protoceratopsidae end where they begin, for the two isolated genera, Protoceratops and 

 Leptoceratops, each with its single known species, tell all we know of the group. With them there 

 is no recorded phylogeny. 



The Ceratopsidae, on the other hand, while established as such at their initial appearance, do 

 nevertheless show a very considerable series of changes, some of which are common to all phyla, 

 others peculiar to one or more. It is necessary, therefore, to give a "synopsis of preceding chap- 

 ters," the details of which must be derived from comparative morphology and the known evolu- 

 tionary trend of later forms, for the actual record of these chapters is as yet undiscovered. I should 

 derive the Ceratopsia from some unarmored stock of ornithischian dinosaurs, probably bipedal in 

 gait, although among those already discovered we know of none which could be considered 

 ancestral, even in the structural sense. 



The unrecorded changes have to do with the secondary assumption of the quadrupedal posture 

 and with the increasing power of the fore limbs to carry the growing weight of the anterior part of 

 the body, although the fore limbs never assume the rectigrade posture of the hind limbs. The 

 increase in size of the head is correlated with a shortening of the neck and loss of motion, due to 

 the condition of coossified immobility on the part of the three anterior cervical vertebrae. The 

 postpubis shortens to a vestige and, on the part of the Ceratopsidae, the elongated ischia become 

 curved downward, and the upper border of the ilium reflected so that the blade becomes horizontal 

 as in certain of the Stegosauria. 



The skull becomes more or less wedge-shaped when viewed from above, and in front of the 

 laterally compressed premaxillaries there appears an additional bone, the rostral, which forms the 

 opposing element to the predentary and, like the latter, is sheathed in a prehensile beak. The 

 ancestral form was provided with premaxillary teeth, which persist in the Protoceratopsidae, but are 

 lost in the Ceratopsidae. 



The crest is composed of the squamosals and a median element which is variously interpreted, 

 but is called here the parietal, and which is fenestrated on either side of the midline. The manner 

 of origin of the fenestrae is clearly indicated in the genus Chasmosaurus, although not so evident in 

 other contemporary genera, and it may possibly have arisen independently in more than one way. 

 In Chasmosaurus, the squamosals are elongated and the median element consists of a keel-like, 

 longitudinal portion, at the hinder end of which a transverse bar is extended to meet the end of 

 either squamosal. Ancestrally, therefore, the outer limit of the fenestra must have been the 

 squamosal. At the anterior end of the fenestra, behind the supratemporal openings, a plate of 

 bone extends laterally to join the squamosal near its anterior end, and from these transverse por- 

 tions of the parietal, branches developed forward and backward, as the case may be, extending along 

 the squamosal until they met in an overlapping suture which, at any rate in Chasmosaurus, failed to 

 fuse and in one case failed to meet. This now excludes the fenestra from contact with the squamosal 

 and the aperture thus formed is persistent throughout the recorded history of the phylum. The 

 other type of aperture, which disappears in evolution, seems to be merely a fenestration or perfora- 

 tion of the broad, plate-like parietals, and is associated with squamosals of short or intermediate 

 length, for in the first records of this type, in Protoceratofs, Monoclonius, or Centrosaurus, the 

 fenestrae are some distance from the squamosals (toward the midline), with thin margins, some- 

 what variable in extent, even between the two of the same crest. There is no sign of sutural union 

 in the bone forming the outer margin. 



