Marsh Collection, Peabody Museum. 337 



but we do know that such a type as Didelphops Marsh, in its 

 dentition and palate, resembles the living carnivorous Marsupials, 

 and it is to some such type in particular that I would refer the 

 origin of the Creodonta. 



The classification herein adopted with respect to the major 

 divisions of the Carnivora is essentially that of Flower and 

 Lyddeker in their work upon the Mammalia, with the excep- 

 tion that I have substituted for their name "Carnivora Vera" 

 the subordinal term Carnassidentia. We shall then have the 

 three suborders, — Creodonta, Carnassidentia, and Pinnipedia, 

 but the difficulties of determining the limits and framing 

 exact and satisfactory definitions for these groups, especially 

 the two former, are just as great when considered as suborders 

 as they are considered as orders. Two main features of their 

 organization have hitherto been used to separate the Creodonts 

 from the higher Carnivora or Carnassidentia, viz: the union of 

 certain elements of the carpus and the modification of particular 

 teeth into a sectorial or carnassial dentition. The relative 

 importance of these two sets of characters in constructing 

 the primary divisions of our classification is of course a matter 

 upon which different opinions are held. Speaking for myself, 

 I am convinced that the tooth characters are of greater 

 moment in making these primary divisions than the union or 

 non-union of certain carpal elements, for the reason that the 

 ununited carpals are undoubtedly expressive of a generalized 

 condition, which applied to all phyletic lines of this series in the 

 early stages of their existence, and cannot, therefore, give the 

 faintest hint of the breaking up into subordinate series. It is, 

 moreover, largely dependent upon time, since, later than a 

 given epoch these carpals are united and earlier than this they 

 are free, in any group belonging to the order. 



If, on the other hand, we base our primary divisions upon 

 the modifications of the teeth, we have from almost the earliest 

 deposits in which we have knowledge of the remains of this 

 group, many of the various phyla more or less distinctly out- 

 lined. Thus, as early as the Torrejon, we find representatives 

 of the Viverravidee, which not only have the teeth constructed 

 upon the same identical pattern as that of certain living Car- 

 nassidentia, but the number is exactly the same. In the suc- 

 ceeding Wasatch, representatives of the modern Canids and 

 Felids appeared, while the various lines of the Creodonta may 

 be said to have been fully established by this time. 



The one distinguishing feature of the dentition of the Car- 

 nassidentia is found in' the fact that the fourth superior pre- 

 molar and the first inferior molar have been exclusively 

 developed into carnassial teeth. Some of the living represen- 

 tatives, such as the Bears and Kaccoons, have largely lost the 



