NO. 1660. SKULL OF TRWERATOPS—JIAY. 97 



Xevertheless, in the text the foramen magninn is said to be wholly 

 in the exoccipitals and the median expanding bar of bone is said to 

 articulate with the supraoccipital. Also, the supraoccipital is cor- 

 rectly represented on Plates 33 and 37 of The Ceratopsia. 



Of course, the question at once comes up regarding the comj)Osition 

 of the frill. It has always been interpreted as consisting of the squa- 

 mosals on the right and left borders and of the coalesced parietals in 

 the middle part. There seems to be no doubt that the bones called 

 squamosals are such. For the middle portion we must seek some other 

 bone or bones than the parietal. It seems to the writer that the re- 

 quired elements are to be found in the supra temporals, bones found 

 in many lizards and in some other reptiles. For those of the lizard, 

 .see Parker." Or, it seems possible that the middle bone of the frill 

 may have developed from the coalescence of nuchal bones such as are 

 found in the crocodiles. 



On each side of the frill of Tricemtop.s, between the squamosal and 

 the so-called parietal, there is an elongated excavation which termi- 

 nates farther in front in a foramen, and this excavation has been 

 called the supratemporal fossa. Xow, it is eas}^ to see that in the 

 alligator the hinder free borders of the parietal and squamosals might 

 grow backward over the animal's neck and make such a frill as we 

 have in Triceratops ; but in this case the supratemporal fossae would 

 be left in their original position. It is difficult to understand how 

 these bones became modified in such a way as to transfer the supra- 

 temporal fossa? behind the paroccipital processes of the exoccipitals. 



It appears to the writer that the supratemporal fossae have either 

 been abrogated or not yet recognized as such. Marsh described an 

 opening in the midline, slightly behind the great postorbital liorn 

 cores, and called it the pineal foramen; although he did not show 

 that it opened into the brain and did conclude that it opened into a 

 large sinus extending above the brain-case into the cavities of the 

 horn cores. Hatcher and Lull likewise say that this foramen com- 

 municates with large sinuses in the postfrontal liones and in the horn 

 cores. Furthermore, Lull '' writes that the sinus underlaying the horn 

 core can be explored through the so-called pineal, or postfrontal, fora- 

 men, and that the latter communicates with those of the horn cores 

 and with the space within the skull behind the orbit. By the latter 

 expression is understood by the present writer the space occupied by 

 the temporal muscles, the space called by Hatcher the temporal 

 fossa.^ Probably in all cases this postfrontal foramen divides below 



"Trans. Phil. Soc. London, CLXX, pp. .595-040. 

 6 Bull. Amer. Mus. Nat. Hist., XIX, p. 691. 

 ''The Ceratopsia, p. 125. 

 Proc. N. M. vol. XXXVI — 08 7 



