170 Bulletin American Museum of Natural History. [Vol. X, 



lower canines of the Uintatheres, but no known species leads 

 directly into Uintatherium. 



Second. — It is now demonstrable that the Periptychidse belong 

 in this group as a bunodont division., which probably arose and 

 diverged in the Cretaceous period. 



Third. — The nearly complete skeleton of Pantolambda bath- 

 modofi, the most archaic type of Ungulate known, places almost 

 upon the firm basis of fact Cope's hypothesis, that the Ungulates 

 sprang from the Creodont division of the Unguiculates. 



Fourth. — A fairly complete systematic revision of the entire 

 group is rendered possible. 



The writer desires to express at the outset his indebtedness to 

 Dr. J. L. Wortman for his energetic and intelligent supervision in 

 the field of these various expeditions, assisted by Mr. Walter 

 Granger, Mr. Barnum Brown and others ; to Dr. W. D. Matthew 

 for the skill and care with which the very fragmentary skeletons 

 of Pantolambda have been put together, and for critical assist- 

 ance in the diagnosis of specific types. Also to the various pre- 

 parators and draughtsmen of the department. 



I. — Origin of the Amblypoda. 



A. The Cretaceous Trituberculate Molar. 



Hypothetical Upper Cretaceous Ancestors. — In describing the 

 Laramie mammals in 1893 the writer' directed attention to the 

 resemblance which certain isolated upper and lower molars bore 

 to the teeth of the Periptychidae, especially to Ectoconus and 

 Haploconus. The figures of these teeth are here reproduced. 



In C {Synconodon), the crowns are laterally compressed, thus 

 bringing the primary proto-, para- and metacones, both of the 

 trigon and trigonid, very close together. This compression of 

 the primary cones is also true of F {Fctoconodon), in which type 

 the outer wall is reinforced as in Ectoconus by accessory tubercles. 



' Osborn, 'Upper Cretaceous Mammals,' Bull. Am. Mus. Nat. Hist., Vol. V, pp. 325-329. 



