378 Bulletin American Museum of Natural History. [Vol. XXVII, 



dactyl manus and pes. The general proportions of the skull were about as 

 in Meniscotherinm, i. e., with face of medium length, large orbits, broad 

 interorbital region and deep lower jaw. The palate probably terminated 

 posteriorly in a pair of processes analogous to those in Hyrax. The tym- 

 panic very early became inflated and prolonged into a tubular meatus; 

 the squamoso-periotic region also became inflated. The upper molars, 

 at first brachyodont, with six cusps, and with prominent anterior and pos- 

 terior cingula, became hypsodont and compressed; the premolars tended to 

 become molariform; the anterior incisors became hypsodont, the lateral 

 incisors and canines either became like the adjacent teeth or tended to dis- 

 appear. The mandible originally had a moderately deep angle {cf. Menis- 

 cothcrium) a feature emphasized with the increasing hypsodonty of the teeth, 

 and procumbency of the incisors. To this factor may be ascribed: (1) the 

 progressive increase of the antorbital slip of the masseter {cf. Hyrax), (2) the 

 deepening and backward projection of the malar, (3) the broadening of the 

 palate. 



The cervical vertebrae had flattened articular faces on the centra. The 

 dorsolumbar vertebral formula (20 in Toxodon) originally approached the 

 Condylarth type (15 -\- 5) but secondarily rose to 15 + 7 {Interaiherium) 

 and 15 + 8 (Pachijrukhos). The tail was long and heavy (cf. Proti/po- 

 therium, Interatherium) and the true sacrals were three in number. The 

 carpus early lost the free centrale (which may have fused with the scaphoid) 

 and widened the lunar-unciform contact (p. 449). The tarsus in general 

 resembled the Condylarth type but very early developed the fibulo-calcaneal 

 contact and lost the astragalar foramen. Such a family of mammals would 

 fall under the Condylarthra as here understood (p. 383), Init would proba- 

 bly be separated from Meniscotherium. by the non crescentic character of the 

 protoconule, from Phenacodus and Euprotogania by the more advanced con- 

 dition of the molars. 



The Litopterna. 



Some of the smaller Litopterna such as Diadiaphorus and Proterotherium 

 resemble the three-toed and one-toed horses not only in the general form of 

 the skull and molar teeth but also in the predominance of the middle digit 

 (III) and reduction or disappearance of the lateral digits. But in spite of 

 these resemblances it seems highly probable that the Litopterns are only 

 very remotely related to the Perissodactyls, for the following reasons: (1) 

 the cervical vertebrte, in retaining flat terminal faces resemble the Typo- 

 there-Toxodont rather than the Perissodactyl type (Lydekker, 1893, p. 57); 

 (2) in the detailed relations of the carpals and tarsals (pp. 449, 379) the Litop- 



