20 Bulletin American Museum of Natter al History. [Vol. XIV, 



Pterodon. M^ transverse reduced ; m^-^ triangular with well-developed pro- 

 tocone, /rt. and nie. connate, metastyle blade smaller; m^-g without met- 

 aconid but with small trenchant heel. Jaw short and deep. 



Sinopa. M^ transverse ; m^ "* sharply triangular with widely separated proto- 

 cone, pa. and me. connate to varying degree, metastyle and parastyle ex- 

 tended into small shearing blades. IMi-s with high triangular trigoii and 

 low basin heel, pa^ and me^ of equal size, pr^ much higher. 



Didelphodus. M^ subtrausverse unreduced. M^-* sharply triangular with 

 widely separate crescentic protocone, pa. and me. somewhat connate, para- 

 and metastyles moderately developed. Lower molars with triangular tri- 

 gonid of three equal cusps (pa^ internal), and large heel. 



Palaosinopa. M- ? subtrausverse ; m^~^ somewhat quadrate, external styles 

 not much developed. Lower molars with low trigonid of three equal cusps 

 (pa'? anterior) and large bicuspid basin heel. 



Quercytherium. Molars much as in Sinopa. Premolars distinguished by ex- 

 treme robustness and large size, p| being the largest. 



The genera Lim?iocyon Marsh, Stypolophiis Cope, and Prototomus 

 Cope are, as Scott holds, synonyms of Sinopa Leidy, except L. 

 riparius, subsequently (Marsh, '99) separated under the name 

 Telmatocyon. Triacodon Marsh is perhaps founded on incomplete 

 molars of either Sinopa or Viverravus. The species placed in it by 

 Cope in 1872 was afterwards correctly referred by him to Stypo- 

 lophus. Proviverra is very close to Sitiopa, but, though very im- 

 perfectly known, seems to differ from any of the species of Sinopa 

 in its short jaw and reduced premolar region, approximating 

 Prorhyzcena. Cynohyxnodon cayluxi is hardly distinguishable in 

 dentition from Sinopa, but seems to be somewhat more advanced 

 in skull characters. It is hardly worthy of generic separation. 

 C. minor., if correctly represented by the upper and lower jaws 

 referred to it in the Paris Museum, is more nearly allied to Quer- 

 cytherium., having the same proportions in the premolars, which, 

 however, are not nearly so robust. Hemipsalodon does not seem 

 separable from Pterodon; and Pseudoptei'odon, according to Scott, 

 is probably founded on milk teeth of Pterodon (Scott, '92). Thy- 

 lacomorphus is known only by the back of a skull described by 

 Prof. Gervais in 1876, but not figured. It was conjecturally re- 

 ferred by Schlosser to Cy?iohycBnodon. It is, however, not a Cre- 

 odont at all, but an Anoplothere, probably Diplobune quercyi. 



The accompanying table will show the conception which this 

 paper attempts to explain of the relationship of the genera and 

 limits of the families of Inadaptive Shear-toothed Creodonts, 

 the Oxysenidae, Oxyclaenidse, and Hyaenodontidse : 



