50 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 57. 



deep fascial lining of the eye socket, and sends down a strong band to 

 the zygomatic arch to complete the posterior boundary of the orbit. 

 The orbital edge of the parietal, moreover, is produced and angulated, 

 furnishing a condition ver}^ similar to that seen in Tupaia, with the 

 exception that the bony parts representing the post frontal and the 

 postorbital have been reduced to the fibrous structure alread}^ 

 mentioned. The general shape of the skull is more compressed and 

 elongated than in RhyncJiocyon, but the temporal area is small, the 

 temporal muscle is weak and the mandibular coronoid is little devel- 

 oped. The extreme development of these modifications is to be seen 

 in Gentetes among the modern Insectivora in which the skull is long, 

 narrov/, and compressed, the temporal area greatly enlarged, with 

 unusually high crests, and with complete absence of a postorbital 

 process. Directly correlated with these latter characters is the loss 

 of the zygomatic arch and the great enlargement of the mandibular 

 coronoid. Thus the various steps may be conceived of just how the 

 elongated and compressed type of skull could have arisen in the 

 Insectivora at least. 



Myrmecoiius among the Marsupials occupies nearly the same 

 position in this group with reference to these characters that 

 Bhyncliocyon does among the Insectivora, and if the Mammalia are 

 of monophyletic origin then, in my judgment, we must look for the 

 precursor of this t3^pe of skull in their reptilian ancestry rather than 

 the carnivorous type. If, on the other hand, the various orders of 

 mammals arose from more than one group of reptiles independently, 

 which upon the whole appears to me not at all impossible or improb- 

 able, then it would be only reasonable to suppose that there were 

 various forms of skulls developed while yet in the reptilian stage 

 which passed into the mammal. 



Mammalian interparietals. — Another matter of importance in 

 connection with the temporal area in the mammalian skull is the 

 presence of an extra bone often seen intercalated between the pari- 

 etals and the occipital wliich is always referred to as the interparietal. 

 Careful investigation of the development of this seemingly unpaired 

 element reveals the fact that it is always developed from at least 

 two centers, and there may be as many as four. I have not been 

 able to fuid any ossification corresponding to an interparietal in any 

 of the carnivorous Marsupials except the Bandicoots and Myrme- 

 cobius. In the former of these species there are two elongated 

 ossicles surmounting the lambdoidal crest which without much doubt 

 represent paired interparietals. In Myrmecohiiis these bones con- 

 sist of a pair of sizable ossicles lying upon either side of the median 

 line just in advance of the occipital crest at the junction of the 

 occipital with the parietal. It is very commonly present in the 

 adult skulls of the kangaroos and phalangers, and when sufficiently 



