Jlylodon liarlani, from Rock Creek, Texas. 371 



The proximal two-thirds of the fifth metacarpal of the right 

 man us is present, but too imperfect to characterize in detail. 

 It is remarkable, however, for the obliquity of its proximal 

 facets. 



Unguals. — The two ungues present are not distinctively 

 characteristic, though with the exception of the basal sheath in 

 one, they are admirably preserved. I have compared them with 

 Cope's description of Jlylodon sodalis, which is based upon 

 a single ungual from Oregon, with the following results. 



Fig. 12. 



Fig. 12. Ungual phalanx of Mylodon harlani, No. 10266, Yale collection. 

 One-fourth nat. size, a, lateral aspect, b, palmar aspect. 



The sodalis has the basal sheath developed on one side only. 

 Texan specimen No. 10265 has none, it having been entirely 

 broken away. Texan specimen No. 10266 has it well developed 

 all around, though slightly fractured on the right side. The rest 

 of Cope's description could apply to either Texas claw. The 

 proportions of the Texas claws vary in that the somewhat 

 smaller one, No. 10265, is more highly arched on its dorsal 

 outline and hence actually and relatively a little deeper. It is 

 also more compressed. Herein it agrees with Cope's descrip- 

 tion of 21. sodalis, which he says is more compressed than 

 21. harlani. As 21. harlani type includes only a jaw, this 

 statement, based doubtless upon subsequently referred mate- 

 rial, should be taken for what it is worth. 



We have therefore in the 21. sodalis type a larger claw of 

 much the same proportions as the Texas specimens, standing 

 between the two in the degree of preservation of the basal 

 sheath. I do not believe, therefore, that Cope's description 

 brings out characters other than those of individual variation 

 or degree of preservation and hence it is entirely inadequate 

 as the characterization of a valid species. 



