No. 2304. REPTILIAN CHARACTERS IN MAMMALS— WORTAIAN. 49 



the arrangement of the conditions seen in the reptile. From the 

 structure and relations of the temporal area as exhibited in Rhyncho- 

 cyan, to such a type as Tupaia, Galeoptenis, or Procavia, the transition 

 is easy and accomplished with very slight modifications. In the 

 same way slight further modifications would produce the conditions 

 found in the Primates, Cheiroptera, Rodents, and Myrmecohius, in 

 which the point of origin above of the postorbital bar is located 

 slightly in advance of the parieto-frontal union. If these conclusions 

 are well founded, then it becomes a matter of considerable doubt 

 whether the postorbital bar of the reptilian stage has ever been bro- 

 ken through and destroyed in such forms as Tupaia and the Primates, 

 in which it is complete, but still lingers as a heritage from their rep- 

 tilian ancestors. If, moreover, this type of mammalian skull is to be 

 traced to a reptilian source of this nature, then it would appear that 

 there must have been two or more types of reptilian skull from which 

 the mammal was derived, since it is inconceivable that the form dis- 

 pla3^ed by Khynchocyon could have arisen from such a type as that 

 sho'\\Ti by the carnivore or carnivorous Marsupial in which the rela- 

 tions of the temporal area and its surroundings are so fundamentally 

 different. 



]f the Cynodonts have had anything to do with the ancestry of the 

 Mammalia — and it is undeniable that they exhibit many decided 

 approaches to the mammalian condition in their structure — then we 

 must assume (1) either there existed an unknown group of them 

 with a skull form not very different from that of ProcolopJion from 

 which all mammals were derived, or (2) that the mammals arose 

 from more than one type, or (3) that the mammxalian resemblances of 

 the Cynodonts are purely accidental and without any special signifi- 

 cance as far as direct ancestry is concerned. These alternatives are 

 suggested by the fact that the Cynodont skull as we at present know 

 it could have given rise to the carnivorous type of skull alone, and 

 it is unthinkable that the other type of mammalian skull as exempli- 

 fied by RhyncJiocyon could have been derived from this form. We can 

 readily understand, on the other hand, how the compressed narrow 

 type of skull could have arisen from the Rliynchocyon type, since we 

 have the examples of Cercodenus and Petrodromus before us, which are 

 more or less transitional between the two and afford an explanation 

 o"^ how the other Insectivora may have reached their present condi- 

 tion. In these two genera there is no bony postorbital process above 

 or below, but along the upper margin of the orbit and as far back as the 

 parieto-frontal suture there is in the recent state a dense, thickened, 

 more or less triangular fibrous pad, wliich is attached to a roughened 

 area on the eye margin, which undoubtedly represents and is the 

 remains of the postorbital process. This fibrous pad is intimately 

 connected with the fascia covering the temporal muscle behind, the 

 144382— 20— Proc.N.M. vol.57 4 



