908 



EOCENE AND OLIGOCENE TITANOTHERES OF MONGOLIA 



It was from the Houldjin gravel, probably, that 

 Obruchey obtained the rhmocerid molar tooth {IBalu- 

 cMtherium) The Houldjin was regarded by Granger 

 and Berkey (1922, p. 4) as of more recent age than 

 the imderlying Irdin Manha formation and as equiva- 

 lent in age to the Hsanda Gol formation, because 

 both contain BalucMtTierium. 



In 1923 Matthew and Granger (December 18, pp. 

 1-6) described the fragmentary Houldjin fossils indi- 



Y- ToSairUsu 



Brontops qobiensis skull \^fe 



Fossiliferous escarpment 



A 



Ardyn Obo formation, CadurcotJierium ardynense 

 Osborn. The large Baluchitherium is comparable 

 with that from Turgai described by Borissiak. The 

 smaller rhinocerid is comparable in size with Caenopus 

 occidentalis, from the Oligocene of Nebraska, with the 

 Epiaceratherium turgaicum Borissiak of Turgai, and 

 with the Aceratherium filholi Osborn of the Phos- 

 phorites, Oligocene; its relationships remain to be 

 determined. 



\ 



Small Jlrtiodac1y/a,etc, \^ Pnmn 



Cadurcother/um ardynense quarry 



Obo 



M 

 W 



ToKalgan 



Ardyn Obo 



CROSS SECTION ATARDYN OBO 

 LOOKING WESTWARD 



Cadurcotherium ardynense cjuarr/ 



Brontops gobiensis 

 ' skull 



FiGX7HE 768. — Ardyn Obo formation (lower Oligocene), after field notes and sketches by 



Granger 



Upper part: Sketch map of Ardyn Obo formation, showing locstion of the principal finds, namely, type skull of Brcm- 

 tops gobiensis (Am. Mus. 20354J, Cadurcctkerium ardynense quarry, and small Artiodactyla. 



Lower part: Cross section of the Ardyn Obo formation, showing tho levels on which the remains of Cadurcotherium 

 ardynense, Brontops gobiensis, and Menodus mongoliemis were found. 



cated above in brackets, also those given in the 

 accompanying table. They concluded in general that 

 this fauna is of Oligocene age but can not be exactly 

 correlated until better known; it may be homotaxic 

 with the Hsanda Gol fauna, but this supposition 

 rests only on the doubtfully identified Baluchitherium; 

 the Houldjin and Hsanda Gol type localities are more 

 than a thousand miles apart, and the geologic character 

 of the two formations is quite different. The amyno- 

 dont 1 Cadurcotherium is comparable with that of the 



LOWER OIIGOCENE: ARDYN OBO FORMATION 



Berkey and Granger (May 25, 1923, p. 12) 



Type locality and notable features. — The type locality 

 of the Ardyn Obo formation on the main Kalgan- 

 Uliassutai trail is south of Sair Usu, where a great 

 escarpment, surmounted by an obo (the inevitable 

 guidepost of the Mongolian desert), stands 300 feet 

 above the general level of the plain over which the 

 Uliassutai trail passes, and where for many miles the 



