O. C Marsh — Cretaceous Mammalia. 179 



on Plate VIII, figures 22-26, may be taken as the type. The 

 upper molars secured resemble the corresponding teeth of 

 Allodon, but the cones are more pointed, and there is no true 

 basal ridge. 



The tooth represented in figures 27-31 of the same Plate 

 may pertain to the same individual as the type. The premolar 

 with three cones, represented in figures 17-21, is referred to 

 the same species, and may belong with the type specimen. 



A second, very diminutive, species is represented by several 

 teeth but little larger than those of the Jurassic forms. The 

 type has a crown with four cones, nearly equal. This species 

 may be called AUacodon pumilus. All the known remains of 

 this genus were found in Wyoming, in the Laramie. 



These fossils evidently belong to the family named by the 

 writer, the Allodontidce, which includes the American genus 

 Allodon. and Bolodon from the Jurassic of England. 



i & j 



Halodon formosus, sp. nov. 



A diminutive species, apparently of this genus, is represented 

 by several specimens, of which the fourth premolar figured on 

 Plate YIII, figures 36-39, may be taken as the type. The 

 crown of this tooth is very low, and elongate. There .are 

 eight ridges on the outer side, and twelve tubercles along the 

 upper margin, the two anterior and the two posterior ones 

 having no ridges connected with them. 



The lower incisor represented on the same Plate, figures 32- 

 35 probably belongs to this species. 



The known specimens are from the Laramie of Wyoming. 



Didelphops (Didelphodon) . 



The name Didelphodon, proposed by the writer in Part I, 

 for a genus of mammals allied to the modern opossums, proved 

 to be essentially pre-occupied, and was changed in the errata to 

 Didelphops. A specimen of this genus, apparently belonging 

 to D. vorax, has one character that distinguishes it strongly from 

 the genus Didelphys. The palate is perforated by a large 

 median aperture. This is not due to imperfect ossification, as 

 seen in the palate of the opossum, and other existing marsupials, 

 but the opening has a well-defined thickened margin. This 

 orifice appears to have been functional, and may represent the 

 posterior nares. If functional, the vacuities in recent forms 

 are probably remnants of the same cavity. 



The allied genus Cimolestes may be distinguished from 

 Didelphys by the fact that the teeth in the lower jaw form 

 a continuous series, there being no diastema. 



