ORDER UNGULATA. 1 357 



several other forms from the Middle and Upper Eocene of the 

 Continent ; Propalazotherium (in which Lophiotherium may be in- 

 cluded) does not appear generically separable. 



Family Pal^eotheriid.e. — With the Palceotheriidce we enter 

 upon another extinct family of this suborder, the type genus of which 

 has long been known from the classic labours of Cuvier. In this 

 family the upper premolars may be either simpler or quite as com- 

 plex as the true molars ; the lower molars have crescentoid crowns, 

 and in the last tooth of this series the third lobe may be either well 

 developed or almost wanting ; all the cheek-teeth are brachydont, 

 and when cement is present it does not fill their valleys. There 

 are always three digits in each foot. The type genus Palazotherium, 

 in which may be included Paloplotherium (or Plagiolophus), ranges 

 from -the Middle Eocene of the Paris basin to the Lower Miocene 

 of Ronzon, but is especially characteristic of the Parisian stage. 



The dental formula is I. % C. -, Pm. i^zf}, M. 3 - ; and the last 



3 i (3-4) 3 



lower molar has a third lobe. In the more typical species (fig. 1229) 

 the four premolars are present in both jaws, the fourth upper pre- 



Fig. 1229. — The right upper cheek-dentition of Palceotherium crasstem; from the 

 Upper Eocene of Paris. Reduced. 



molar is as complex as the first true molar, and the third lower pre- 

 molar as the fourth lower premolar ; while the diastema is compara- 

 tively short, and the canines are not large. The upper true molars 

 of all the species exhibit an expansion of the inner extremities of the 

 transverse ridges, foreshadowing the structure of the teeth of some 

 of the EquidcE. The cranium is Tapiroid in character, especially in 

 the prominence of the nasal bones ; from which it is deduced with 

 great probability that the nose possessed a short movable proboscis. 

 The general form may also be supposed to have been like that of 

 the Tapirs, and the restoration of P. magnum given by Cuvier (fig. 

 1230) exhibits to us an animal closely similar to the existing Tapir. 

 In this particular instance, however, we know that the restoration is 

 incorrect, since the discovery of a complete skeleton of this species 

 has shown that it was a more slender and longer-necked animal, 

 resembling in its general figure a Llama. 



In that group, which is considered by some writers as generically dis- 

 tinct under the name of Paloplotherium, the last upper premolar has its 



