292 Bulletin American Museum of Natural History. [Ynl XXMl, 



(19) Entocarotid entering tympanic chamber from the rear and giving 

 rise to three branches (p. 247). 



(20) Auditory ossicles in process of acepiiriny tlie Insectivore character- 

 istics. Stapes becoming widely open, pierced by the large stapedial branch 

 of the entocarotid; malleus with a distinct head, wide lamina, long processus 

 gracilis, the latter fitted to the ring shaped tympanic. 



(21) Skull retaining a large number of primitive Marsu pio-Placental 

 characters in addition to numbers 6-19 mentioned above. The preceding 

 studies of the skull of the Marsupials (p. 217), of Solenodon (pp. 242, 253) 

 and of Tupaia (p. 279) reveals a very large number of untierlying characters 

 in common, especially in the arrangement of the majority of the cranial for- 

 amina, structure of the ethmoid, vomer, relations of the pterygoid, alisphe- 

 noid, short basioccipital, etc. 



(22) General characters of the skeleton much as in Tupaia, especially 

 the vertebral formula and characters of the shoulder girdle, humerus, femur, 

 pelvis. In the ancestral Lipotyphla the carpus and tarsus probably fore- 

 shadowed those of Ericulus (p. 251) and thus approximated the primitive 

 Creodont type ; but in the ancestral Menotyphla the carpus and tarsus sug- 

 gested those of Tupaia and thus remotely foreshadowed the Primate type. 

 In both manus and pes digits II, III, IV were subequal, digit III being 

 slightly the longest; the poUex and hallux were slightly divergent; the claws 

 were compressed and well developed. 



(23) Testes remaining abdominal (cf. Centetes, Macroscelides). 



(24) Uterus bicornis or duplex. 



As thus conceived the family of ancestral Insectivorcs would be ordinallv 

 related to the contemporary ancestors of the Creodonts. It would be dis- 

 tino'uished from them chiefiv bv smaller size and less robust structure, more 

 elongate snout and premaxillaries, smaller canines, smaller and more sharply 

 cusped molars, and perhaps by habits which were more arboreal, less preda- 

 tory and more insectivorous frugivorous. 



The Tillodontia. 



It is customary to place the Tillodontia in the neighborhood of the Ro- 

 dentia and to regard the order as ancestral to the latter; but an examination 

 of the skull and skeletal remains of Tillotherium- JNIarsh (1875) and of its 

 ally Esthonyx as described by Wortman (1S9G), shows that apart from the 

 pair of enlarged rootless incisors and the concomitant reduction of the other 

 incisors and canines, there is not a single decisively Rodent character in the 

 whole skull and skeleton. As scalpriform incisors have been evolved quite 



