lOlU.J The TUlodnntia. 293 



in(lei)endently in several different orders, Weber (1904, p. oU) seems ([uite 

 justified in coiieluding that the supposed connection with the Rodentia is 

 very problematicaL 



The evidence for the view that the Tillodontia are not related to the 

 Rodentia is as follows: 



(1) The dentition, except the incisors, is wholly unlike that of any 

 known Rodent. 'I'he most ])rimitive dentition in the group is that of 

 E.sthonij.r, in which the tritubercular molars (figured by Osborn, 1907, p. 

 151) are of a peculiar type, which is entirely unlike that of the oldest Eocene 

 Rodents of the family Ischyromyidiie (ef. p. 327.) 



(2) The skull as figured by Marsh (1875) shows many significant 

 diflferences from the Rodent type. In TiUotheriiim the mandibular condyle 

 and its glenoid facet are transversely instead of antero-posteriorly elongate, 

 the post-glenoid process of the scjuamosal is well developed, so that the 

 motion of the jaw in gnawing must have been largely vertical. The basi- 

 cranial region is very unlike that of the oldest known Rodents, the Eocene 

 Ischyromyid» and is rather of the Creodbnt type. The bulla instead of 

 being large and inflated as in all Rodents appears to have been uninflated. 

 There is no suggestion of the i)eculiar Rodent foramina (p. 329), nor does 

 the angle of the jaw foreshadow any of the various Rodent peculiarities. 

 The large facial exposure of the lachrymal is another Creodont character 

 which tends to separate TiUotJierium from the Rodents. 



On the other hand the Tillodontia retain many primitive features besides 

 those mentioned above, pointing to derivation from the Tnsectivore-Creodont 

 stock, as follows: 



(1) The most primitive member of the group is K.s-thonij.r burmei.steri 

 from the Lower Eocene (Wasatch), as shown by Wortman (189G). This 

 species retains a dental formula of I.|, C.i, P.-, M.g, which is an immediate 

 derivative of the primitive Insectivore-Creodont formvda; and the material 

 proves that the pair of incisors which is destined to become scalpriform 

 and rootless is i|. A i)rogressive hypertrophy of the anterior lower incisors 

 seems frequently to be a result of insectivorous tliet (Bensley, 1903, p. 124; 

 cf. CwnoJestes, Di.stcechurus, Erinaceuft, Solenodon, Chiroinys-, etc.) and the 

 molars of Es-thoni/.r might also readilv have been derived from the sup- 

 posedly insectivorous-carnivorous type of the Oxyclfenidse. 



(2) Tillotherium, even though highly modified in certain respects, also 

 resembles the Insectivore-Creodont type in the following characters of the 

 skull: muzzle very heavy, long narrow mid-craniiun, low sagittal crest, 

 small brain case, rather slender zygomata, large orbit broadly continuous 

 with temporal fossa, lachrymal foramen marginal, facial exposure of lachry- 

 mal considerable, two anterior dental foramina; characters of the base of 

 the cranium of Creodont type. 



