ANTHONY, FOSSIL MAM3IALS FROM PORTO RICO 201 



the convexity being pointed inward and forward. The molars probably 

 grew throughout life, as they are rootless and of practically the same 

 cross-section at the bottom as the top. 



Measurements: Total length (approximate), 125 mm.; width, back of 

 zygomatic process of maxillary (approximate), 46 mm.; length of nasals 

 (estimate), 40 mm. ; length of maxillary tooth row, 32 mm. ; length dias- 

 tema, 30 mm. ; dimensions of m^, 8X8 mm. ; length diastema of lower 

 jaw, 19 mm. ; dimensions of pm4, 6X8 mm. long. 



Lwih bones and trunk skeleton. — The skull of Elasmodontomys suffi- 

 ciently establishes the distinctness of this type, and a description of the 

 other bones found must be deferred to a later paper, making mention, 

 however, of the fact that the limb bones are in a normal proportion to 

 the size of the skull and are in most respects very similar to those of any 

 hystricomorph of this size, Myocastor for example. 



EEMARKS 



Elasmodontomys seems to occupy a position of its own among the hys- 

 tricomorphs. At the present writing none of the accepted families of 

 this section appear to have very strong claims upon it, and this conclu- 

 sion is not a hasty one, but the result of a careful comparison with many 

 fossil types and all of the recent forms. The molars of Elasmodontomys 

 may be matched approximately in several families and among the sciuro- 

 morphs as well as the hystricomorphs, showing that to this character 

 undue importance may not be attached, as it is a parallel evolution in 

 these different groups. But when we couple this character, the laminate 

 structure of the molars, to such others as are recognized hystricomorph 

 characters, the presence of anteorbital fenestrse and four molars, for 

 example, the search for an including family must be restricted to the 

 following families,^ the Cavidse, the Chinchillidas, the Dasyproctida\ the 

 Erethizontidse and the Octodontidse, excluding for obviou.s reasons the 

 Old AVorld hystricomorphs, and troublesome characters arise in the case 

 of each family. Examination of the accompanying plates Avill demon- 

 strate these points better than a written statement. It is not unlikely 

 that more extended comparisons and reflections and additional material 

 may warrant the erection of a separate family for this most interesting 

 rodent. 



As the bones of this animal were found quite well bedded in the red 

 stalactite formation, the deeper layers of the cave deposit, its age may be 

 assumed as at least as old as the late Pleistocene, and it is plausible to 

 consider Elasmodontomys as a contemporary of Acratocnus. 



® Palmer's classification : N. Amer. Fauna, No. 23, p. 782. 1904. 



