1894.] NATURAIj sciences of PiriLADEI^PHIA. 227 



Whatever the fiitui'e may show the exact genetic interrelations of 

 these animals to be, it is evident that Ptyssophorus is the more primi- 

 tive type; Tretomijs and Hod()mij>< seem to represent more advanced 

 stages in the evolution of the group, while Xenomys and Neotoma 

 are more specialized. Xeno)nys retains more primitive characters 

 than Neotoma, and consequently must be looked upon as nearer 

 Tretomys, with which it agrees closely in dental characters; on the 

 other hand, it is a far more specialized type than Neotoma and can in 

 noAvise be regarded as holding a place in the direct line of descent 

 between Ptyssophoms and Neotoma. 



The discovery of complete skulls of PtyssopJwri(s and Tretomys may 

 show that the line of generic separation should be drawn between 

 Pfys.-<ophon(S, on the one hand, and 'Tretomys and Xenomys on the 

 other; and believers in comprehensive genera may unite the two 

 latter as subgenera of a single genus, the relation of which to 

 Neotoma would be that of a specialized ancestral type to a modern- 

 ized type. 



The fact that of the living genera only a single species of Xenomys 

 and two of Hodomys have been discovered, while nearly thirty species 

 of Neotoma are known, is strong evidence that Hodomys and Xowmys 

 are survivors of the past, bridging over the gap between Ptyssophonis 

 and Neotoma, and that the latter genus is now at or near the height 

 of its development. 



Irrespective of the interrelations of these animals, it is evident 

 that, collectively, they form an important though not highly special- 

 ized subdivision of the Murine series, standing somewhat apart from 

 the others. Wliile they resemble the Arvicolinaj in some respects, 

 they differ in numerous important characters and cannot be regarded 

 as intermediate between the Arvicoliuic and Cricetiuce. On the con- 

 trary, the Neotomime and Arvicolinieseem to be independent oHshoots 

 from the half-tuberculate crowned Cricetines. Among the many ex- 

 cellent characters that serve to distinguish these two groups, the fol- 

 lowing are sufficient for present purposes: — 



Subfamily ARVICOLINiE. 



Cranium abruptly and strongly constricted immediately in front 

 of brain case, which is (juadrangular, projecting squarely into 

 orl)it ; orbital and temporal fossje well difi'erentiated ; jugal forming 

 half, or more than half, of outer side of zvgoma and always reach- 



