96 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxvi. 



The writer will first consider that impaired bone which was re- 

 garded by Marsh and Hatcher as the supraoccipital.^ This question 

 needs to be answered : How can that bone be the supraoccipital 

 which overlies, as this one does, the midbrain instead of the 

 medulla oblongata and the cerebellum? Furthermore, it appears 

 quite imi:)robable that the supraoccipital alone of the bones of 

 the occipital segment would fail to coossify with the others. How- 

 ever, what seems to be an unanswerable argument against the identifi- 

 cation of this bone as the supraoccipital is the fact that the latter 

 bone in reptiles takes an essential part in the formation of the internal 

 ear, including on each side, as it does, always some part of the pos- 

 terior semicircular canal. It may be said that its right and left bor- 

 ders are thereby firmly anchored to those other bones that enclose por- 

 tions of the semicircular canals, the opisthotic and the prootic. Now, 

 if the bone called supraoccipital by jNlarsh and Hatcher is such, the 

 semicircular canals would have to make a looj:) about 100 mm. long in 

 order to reach the supraoccipital. 



The fact appears to be tliat this bone has been wrongly identified. 

 The true supraoccipital is that bone which forms the roof over the 

 medulla oblongata, and which in the specimen here studied and in all 

 others known is ankylosed to the exoccipitals on each side. AVliether 

 or not the latter join each other over the foramen magnum can not 

 now be determined. In C ayn ptosanrus the supraoccipital forms a 

 considerable part in the boundary' of the foramen magnum, but in 

 crocodiles no part. As shown by Hatcher's fig, 8, the lower border 

 of the bone called supraocci})ital rests on the side walls of the brain- 

 case, his alisphenoid, as far forward as the "alisphenoid buttress for 

 the postfrontal;"' but these are almost exactly the relations that the 

 parietal has in the alligator. Unfortunately, in the specimen before 

 us the bone regarded by the writer as the parietal is missing, except 

 a part of the left side, whose broken edge is seen in fig. 1, Plate 2, jki. 

 From No. 4286, T. sulcatuH^ described below, it appears that the su- 

 ture between the parietal and the alisphenoid would run along the 

 lower border of the broken surface just referred to, but it would prob- 

 ably strike the outer surface of the brain-case much below the upper 

 border of the broken surface shown in the figure mentioned. 



It would appear that the editor of The Ceratopsia himself, who 

 lettered the figures of Hatcher's work, was now and then either in 

 doubt regarding the identity of the supraoccipital and the parietal or 

 was led instinctively to their correct determination. Fig. 107, on 

 p. 121, represents as supraoccipital a portion of the bone called exoc- 

 cipital in fig. G; while that bone which in the former figure is denom- 

 inated parietal is represented in fig. 6 as being the supraoccipital. 



'^ Marsh, Dinosaurs of North America, p. 210; Hatcher, The Ceratopsia, pp, 

 16, 17, figs. 7, 8. 



