1905.] PRIMITIVE REPTILE PROCOLOPHON. 221 



below the postsquamosal above, and between the squamosal and 

 quadrate bones behind it and the postorbital in front, so that the 

 space between the bones, which might be occupied by the suj)ra- 

 temporal, has only a linear extension on the external surface, above 

 the malar. The internal suture which separates a supratemjjoral 

 from the squamosal is not clear in specimens of other species. 



The Postfrontcd.- — One of the most characteristic features of 

 Procoloplion is the small size of the postfrontal bone, which is a 

 narrow strip above the orbit external to the parietal and frontal, 

 contrasting with the relatively large size of the prefrontal bone. 

 There is an appearance of the prefi-ontal and postfrontal both 

 iinderlapping the frontal bones in P. trigoniceps^ but the pre- 

 servation is dissimilar on the right and left side of the head, and 

 the evidence is not conclusive that the postfrontal is larger than it 

 appears to be. This character is in mai-ked conti'ast to the 

 condition in Dicynodontia and Theriodontia, in which the post- 

 frontal not only contributes to the bar which divides the orbit 

 from the temporal vacuity, but is prolonged backward on the 

 temporal vacuity along the bevelled margin of the parietal bone. 



I conclude, from detailed comparison of these strvictures and 

 from measurements, that the type species are founded ujjon 

 characters which clearly distinguish them. Other evidence shows 

 unexpected variation in the skulls of Procolojihon. 



In 1878 I described additional material also from Donnybrook, 

 and discvissed the aiiinities of the genus with Hatteria and Ano- 

 modont reptiles. Three species ajDpeared to be indicated by as 

 many specimens, and were described under the names P. griersoni, 

 P. Iatice2)s, and P. cuneice2)s, and figured in pi. xxxii. Quartei'ly 

 Journal Geol. Soc. vol. xxxiv. The matrix was afterwards further 

 removed from these fossils, chiefly in the endeavour to elucidate 

 the back of the skull and the quadrate region. The jDublished 

 figures, which are somewhat rough, are chiefly directed to show 

 external variations of form, and the divided nares. Beyond cor- 

 recting the identification of the postfrontal bone in the way 

 already indicated in the evidence figured in 1889 (Phil. Trans. 

 Roy. Soc. B. pi. 19), and omitting the quadrato-jugal bone, I have 

 nothing to modify in those descrijations ; but better specimens 

 would be required to prove that the characters in which they 

 diflfer are constant. 



The Occipital Region. — Although all these types were developed 

 to display the occipital region, it was only found in Procolojjhon 

 laticeps. The transverse, slightly concave occipital border of the 

 roof of the skull, fonned by the paiietal bones and postsquamosal 

 bones at the outer angles (text-fig. 31), extends backward as a ledo-e 

 beyond the nearly vertical occipital aspect of the skull, which it 

 slightly overhangs. The ledge is inclined downward, and termi- 

 nates in a sharp edge, which at the outer angles curves down with 

 the postsquamosal to form an arch above the auditory notch 

 behind the squamosal bone (text-fig. 32). 



