THE MAMMALIA OF THE UIl^TA FORMATION. 483 



latter hone was comparatively low, and the cuneiforms high. The distal end is en- 

 tirely occupied by the large facet for metacarpal iv, though what seems to be a small 

 facet for nit. v occuis upon the external side. The po.sterior hook-like process of the 

 cuboid is long, heavy and lecurved. 



The metatarsus apparently consists of two functional mombors, though grooves 

 and facets upon the sides of these indicate the presence of at least the proximal 

 jjortions of the lateral metatarsals. Mt. ii appears to have been less reduced than 

 mt. T. The median pair are more slender than the corresponding metacarpals, and 

 were i^robably more elongate; their shafts are flattened and the approximate surfaces 

 very closely applied, but show no tendency to coossify ; both have strong posterior 

 projections from the proximal ends. The head of nit. iii would .seem to indicate that 

 the nieso and ectocuneiforms were not coossified. Mt. iv covers the entire distal 

 surface of the cuboid, allowing to nit. v only a lateral connection with that bone. 

 The distal ends of the median metatarsals at once suggest the eharacteiistic cannon- 

 bone of the Tylopoda, and are i-emarkably like those of the AVhite Kiver representa- 

 tive of that group, Poehrotheriwn. The articular surface is but slightly marked off 

 from the shaft, and the keel, though piominent, is confined entirely to the palmar 

 side. When the ends of the two metatarsals are placed together they exhibit a slight 

 divergence, as is the case in a much more marked degree in the recent Tylopoda. 



The i)Jialan(/es of the proximal row are long and slender and curved outwards ; 

 the proximal surfaces are somewhat symmetrical, causing a slight divergence of 

 the toes. Tiie ungual phalanges are high, compressed and pointed, and somewhat 

 flattened on the internal (approximate) sides; they resemble the unguals of the ante- 

 lopes much more tlian those of recent Tylopoda. 



TJie Sydemalic Position of LeptulnKjulin^. 



In our preliminary account, drawn up IVoni the mandibular dentition only, we 

 inclined to the view that this genus was allied to J'rodreniotherium, and was, perhaps, 

 the forerunner of the White Kiver 2ycp<on<eri/a;. The new material iilcntilicd .since 

 that description was written, as well as direct comparison with original .speciniL'U.s of 

 J'ro(Iremothfriu7i), shows that our former conjecture is entirely untenable-. The teeth, 

 and still more strongly the skeletal fragments, prove that Lrp/otriiynliis i.s very clo.sely 

 allied to J'oehrdlhcriiaii, and is, theiefore, the earliest known iei)resentative of the 

 Tylopoda. If more perfect specimens should confirm the douhl I'lil indieatioii in tiie' 

 type of the loss of the first lower premolar, this species, at least, could not be jjlaced 



