THE OSTEOLOa Y OF SIX0PA—3fA TTIIEW. 217 



types, as distinguished from the far more common instances of parallel 

 adaptation. 



As has alread}" been intimated, the geological occurrence of the 

 known species forbids their being considered as in direct genetic 

 secjuence; Init the genera may be properly so regarded (excei)t Delta- 

 thrriuiii)^ and the features of skull and skeleton entirel}- accord with 

 the teeth in indicating a direct genetic sequence of the genera. 



The species of Tlymnodon differ very considerably' in certain adapt- 

 ive features of the base of the skull, dependent upon the pushing- 

 backward of the glenoid articulation to a position almost opposite the 

 occipital condyles. In all of them, and in Ptcrodon as well, the basi- 

 occipital is somewhat shorter than in Sinopa and the petrosal promi- 

 nence of irregularly rounded form, situate at the bottom of a deep pit. 

 The posterior nares are roofed over to a varjdng extent b}^ union of 

 the pterygoid plates of the palatines and alisphenoids. The tympanic 

 bulla is ossified to a varj^ng degree. The fossa between the cond3des 

 and lambdoidal crests is tilled up as in Daphse/xis. The limbs show a 

 more or less cursorial adaptation. These features are developed to 

 the greatest extent in the large American species If. horridus; the 

 European JI. hrachyphynchus is the most primitive (except that the 

 bulla is completely ossitied according to Filhors statement). A skel- 

 eton from Colorado referred to H. cruentus shows a mere ring of ossi- 

 fication of the tympanic, while in other species the bulla was complete 

 but small (according to Scott). //. paucidens is the most primitive of 

 the American species. 



Pterodon is much like ITyeenodon in the features of the base of the 

 skull, but has the united mastoid and paroccipital processes extended 

 into broad wing-like "jugular apophyses" (Filhol), while in Uyseno- 

 don the}' are less developed than in Sinopa. The fossa behind the 

 lambdoids is deep, the post-nareal gutter narrowed anteriorl}' but not 

 ^oofed over, and in other respects the skull is very primitive, but 

 resembles the primitive species of Ilyn'iiodon and differs from Shwpa 

 in the details of form of the bones and processes. 



Tritemnodon is very close to Sinopa in all the details of skull and 

 skeleton structure, as maj' be seen by comparison of the figures and 

 description of T. agilis given by Wortman. 



Cyiwliya&nodon is near to Sinopa and Tritemnodon., but has a shorter 

 basicranial region and larger brain case. It appears from Filhol's 

 figures of 61 cayluxi to show various other distinctions from these 

 genera in the form of the otic region and arrangement of the foramina, 

 as well as in the shorter, higher crowns of the premolars, all placing ifc 

 more directly in the ancestral line of llyainodon and Pterodon. 



