1 9© I.] Matthew, Additional Observations on the Creodonta. 13 



A. {Mioclce?ms, Clcenodon) protogonioides Cope,' Basal Eocene of 

 New Mexico. 



The last genus is distinguished only by the smaller quadrate 

 teeth — simpler than those of Cicenodon if the reference of Mioclce- 

 nus protogonioides be correct. The type species is little known. 



If the foot bones figured by Osborn ('go, p. 60) are correctly 

 referred to Arctocyon, there are important differences in foot- 

 structure between it and Cicenodon, the former having no fibular 

 facet on the calcaneum, while the ungual phalanx is much less 

 compressed. The figure given by Lemoine ('78, pi. IV) of a cal- 

 caneum of A. gervaisi, is much more like those of Cicenodon ferox 

 and C. corrugatus and appears to indicate a fibular facet. I think it 

 probable that the bones figured by Osborn are those of some 

 other Creodont. The skeletal structure of Anacodon is unknown. 



The dental distinctions are : 



Arctocyon. 

 Upper premolars consid- 

 erably reduced, first 

 one-rooted, second 

 and third two-rooted, 

 fourth large with 

 strong deuterocone. 



Anacodo7i. 

 Upper premolars much 

 reduced, the first and 

 second minute or ab- 

 sent, third two-rooted, 

 fourth small with a 

 strong deuterocone 

 and rudimentary third 

 and fourth cusps. 



Upper tnolars as in CI(T- 

 nodon. 



Molar cusps obsolete. 



Cicenodon . 



Upper premolars mod- 

 erately reduced, the 

 first one-rooted, 

 second two-rooted, 

 third and fourth three- 

 rooted with high tri- 

 hedral protocone and 

 weak cingular cusps 

 at the bases of the 

 three solid angles. 



Upper molars with three 

 low subequal cusps, 

 strong metaconule, 

 somewhat weaker hyp- 

 ocone and very small 

 paraconule. Metacon- 

 ule weak and hypo- 

 cone absent on m^. 



These distinctions are based on the typical species of each 

 genus. It is probable that the other species referred to Arctocyon 

 are closer to Cicenodon, but the present evidence seems to warrant 

 holding the three genera separate. Arctocyonides if correctly 

 represented by the American species A. {MioclcEntes, Cicenodon) 

 protogonioides (Cope) is the most primitive of the family, but 

 hardly deserves more than subgeneric separation from Cicenodon. 



