1899.] 



PLEXODOXT MOLARS OF MAMMALS. 



561 



are limited to tlie internal margin. A somewhat similar dis position 

 is seen in the molars of some Cretaceous Diprotodonts of the 

 family Garzonidse, e. g. the molar of Halmariphus guaraniticus, 

 represented in fig. 5. This agreement in the disposition of the 

 primitive molar elements seems to imply that the Eodents, the 

 origin of which is still a mystery, may represent a side branch of 

 the Diprotodonts, which originated towards the middle of the 

 Cretaceous period. 



We may next consider the Ungulates, which by their molars, at 

 least those of the present epoch, do not appear to bear an j"^ relation 

 to the Didelphyidse and their predecessors. This, however, is not 

 the case. In a recent publication, I have declared that in the 

 Cretaceous of Argentina all the groups of Ungulates exhibit in 

 the form of their molars a great resemblance to each other : all 

 show the sexcuspidate form ; if not visible in the adult, it is seen 

 in young stages. 



The Argentine Proterotheridse, resembling the Horses in their 

 tridactyle and even monodactyle hoofs, and the Palaeotheridae in 

 their molars, are amongst the most characteristic and most specia- 

 lized of Ungulata. Their oldest known representative is the 

 Deuterotheriam disticluim of the Upper Cretaceous ; its fifth right 

 lower molar, just in the beginning of wear, is represented from the 

 upper aspect in fig 7«. This tooth shows the six conical and 



Fig. 7. 



onnfiy 'h 



'^yi./i' d 



Deuterotheriiim distichum: fifth right lower molar, twice nat. size. — Upper 

 Cretaceous; Patagonia, a, Superior aspect of tooth which has jast 

 pierced the gum ; h, superior aspect of worn tuoth of adult ; c, external 

 aspect of slightly worn tooth. 



perfectly separated cusps, with a disposition closely similar to 

 that of Proteodidelphys, and also with the cingulum (c) on the 

 external side, visible in figure 7 c, which repz-esents a slightly 

 worn specimen of the same tooth. However, in the present 

 genus this conformation had become transitory, as shown by the 

 figure, 76, which exhibits the crown view of the same tooth in a 

 worn condition ; the positions formerly occupied by the primitive 

 elements are marked by the corresponding letters, but the cusps 

 are uo longer recognizable, and without being acquainted with the 

 unworn tooth it could not be guessed that its starting point is 

 almost absolutely identical with the form presented by the same 

 tooth of the Didelph3'idae and of Froteodidelphys. The last-named 



