1910.] The Litoptenia. 379 



terns disagree radically with the Perissodactyls; (.3) the premolars are 

 simpler than in the contemporary hippoids and the molars retain clearer 

 indications of derivation from the trigonal type. 



The Litopterns appear to be separated subordinally from the Toxodont- 

 Typothere group: (1) by the detailed characters of the carpals and tarsals; 

 (2) by the characters of the molars, which are of hi{)poi(l rather than 

 rhinocerotoid type and have a prominent mesostyle; (3) by the lack of 

 inflation in the squamoso-periotic region. 



More in detail, the evidence for the preceding conclusions in regard to 

 the relationships of the I>itopterna is as follows: (1) Deniition. In the most 

 jii'imitive dentition, that of Proterotherium sp. (figured l)y Osborn, 1907, 

 p. 189) the molars have the familiar elements of the Ili/mrofherium. molar, 

 but differ from those of Eocene Perissotlactyls in significant details, espe- 

 cially the lack of union of the proto- and metaconules with the ectoloph. The 

 protocone also is centro-internal, as in the Periptychidse. The Miocene 

 Litopterns were thus less progressive in molar evolution than the Lower 

 Eocene Hippoids. (2) Feet. The feet of the Litopterns, as figured by Ame- 

 ghino (1898, pp. 162, 165) parallel those of the Equidje in the reduction 

 of the digits, first to three and then to one; but ordinal relationship with the 

 Perissodactyla is excluded by the following detailed characters of the carpus 

 antl tarsus: In the tridactyl manus of TJieosodun {op. rit., p. 162) the carpus 

 is serial, but of a peculiar type, inasmuch as the cuneiform touches the 

 magnum, the lunar being thus separated from the unciform; all the bones 

 are relatively wide and flat. In the tarsus the astragalus is widely separated 

 from the small cuboid and the calcaneum has a larae ridge for the fibula, 

 both of which characters are seen also in the Typotheres. In the mono- 

 dactyl Thoatherium (op. cit., p. 165) one of the typical Litoptern "jjseudo- 

 horses," all the peculiarities of the carpus and tarsus of Theosodon are 

 emphasized. The lunar is now widely separated from the unciform, while 

 in the tarsus the navicular is widely separated from the cuboid, which is 

 greatly reduced in size; the fibular facet on the calcaneum is pronounced. 

 In all these features the " pseudo-horses " contrast with the Equidfe. Thus 

 the characters of the carpus and tarsus in no case appear to indicate direct 

 relationshij> with the Perissodactyla, and they are without doubt of higher 

 phylogenetic significance (p. 110) than resemblances either in the number 

 of digits and axis of symmetry or in the dentition. 



The preceding conclusions might be criticised on the ground that while 

 the "serial" tarsus and carpus of the Litopterns is held to disprove rela- 

 tionship with the Perissodactyls, the interlocking carpus of the early Typo- 

 theres is yet held to be consistent with derivation from the serial Condylarths. 

 But this apparent inconsistency seems to be necessitated by the evidence. 



