22 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.63. 



RESTUDY OF EOCENE AND PALEOCENE TARSIIDS. 



In order to arrive at any definite or satisfactory conclusions regard- 

 ing the approximate stages of evolution which may have been reached 

 by any group of animals at a given geologic period, it is first neces- 

 sary to determine as nearly as possible what characters presented are 

 basic or primitive, and what features may be regarded as marking 

 definite lines or trends in development. Where ancestral forms and 

 immediate descendants are known, the problem is comparatively 

 simple and there is little chance of disagreement; but where these 

 are not known, the task is more difficult and the chances for agree- 

 ment are, in the very nature of the case, much diminished since most 

 of the conclusions are arrived at through inferences drawn from com- 

 parisons with similar but often wholly unrelated forms. The dif- 

 ficulty seems to be increased when dealing with an order such as the 

 Primates, where many features in all groups even of its living repre- 

 sentatives are still in a stage relativel}' not far removed from the 

 generalized base-structure of primitive mammals. 



Keeping these facts in mind, and also remembering that various 

 specializations in certain groups have progressed much more rapidly 

 than in others, and that as we approach the common origin of the 

 groups the less conspicuous become the differences in these modifica- 

 tions, let us reexamine the evidence presented by the early Tertiary 

 representatives of the order. 



The Lower and Middle Eocene Tarsiids include nine genera referred 

 by Matthew, as already stated, to four key groups, as indicated by 

 lower jaw characters. ^^ Among them are represented a rather wide 

 variety of forms, yet all show a definite trend in the same general 

 direction. 



The Fort Union Tarsiids, represented by Paromomys, PalaecJithon, 

 and Elphidotarsius , as defined, add one or possibly two more groups 

 of similar rank, making in all at least five major groups of the Tar- 

 siidae represented in the early Tertiary deposits of America. The 

 exact taxonomic rank of these groups is not known. As Matthew 

 has quite justly concluded, our present knowledge is insufficient to 

 properly define them as families or even subfamilies, either of which 

 they may prove to be. The important fact in this connection is that 

 as at present known they all exhibit certain features, some more 

 than others, which suggest Tarsius affinities, and on present evidence 

 seem properly to belong in the same family as the living Tarsiers, 

 although none of the Eocene genera exhibit a combination of charac- 

 ters which would warrant considering them directly ancestral to the 

 latter, and much less to any other modern group of Primates. In 

 fact, there is little ground for assuming, as seems to have been done 



i» A fifth group (numbered " I " in Mattha.v's key) is represented by the living Tarsiers. For com- 

 plete definitions of thasj groups see Matthew, Bull. Amer. Mus. Nat. Hist., vol. 34, 1915, pp. 429-483. 



