Marsh Collection, Peabody Museum. 27 



exclusion of the metatarsals is not certain, but it is in all proba- 

 bility in some way associated with the retention of the oppos- 

 able hallux in the development of the elongated pes. It is of 

 interest to note just here that the fourth digit of all those 

 Primates with elongated tarsals is very perceptibly the longest 

 and strongest of the series, more so, in fact, than in those 

 species in which the tarsals are not elongated. It is all but 

 certain that if this modification were to continue in an exclu- 

 sively terrestrial habitat long enough to cause the opposable 

 hallux to disappear, the fourth digit would become enlarged 

 and modified, as in the kangaroos and their allies. The chief 

 differences would, of course, be in the elements elongated. 

 The shifting of the axis to the outside and the specialization of 

 the fourth digit are the strongest possible proof that the foot 

 of the kangaroo has been derived from an ancestral type in 

 which the hallux was fully opposable, and hence indicating an 

 arboreal habitat for its possessor. Tarsius is the most highly 

 modified of all the Primates with respect to the elongation of 

 the tarsals, as is shown by the reduction in size aud the coossifica- 

 tion of the fibula with the tibia, as well as in the grooving and 

 broadening of the astragalus. It is in this species that the 

 preponderance in the length and strength of the fourth digit 

 over its fellows is greatest. 



Of the known representatives of Eocene Primates in North 

 America, there are no less than six or seven genera, including 

 at least twelve species, which are more or less closely related 

 to Tarsius. The skull is known in two of these species only, 

 the remainder being represented by teeth and jaws exclusively. 

 On account of the incompleteness of many of these remains it 

 is quite impossible to determine whether they are members of 

 the Paleopithecini or J^eopithecini. It will require a knowl- 

 edge of the relations of the lachrymal, as well as of the struc- 

 ture of the limbs, before these points can be finally determined. 

 That they do not belong in the Lemuroidea is shown by the 

 characters of the lower incisors and canines, which are known 

 in all the species with the exception of one or two. As a mere 

 matter of convenience in grouping, I arrange a number of these 

 species temporarily in the Paleopithecini. In so doing, how- 

 ever, I wish to state distinctly that there are very good 

 reasons for regarding some of them, at least, as true monkeys, 

 directly ancestral to certain of the living South American 

 forms. This will be further discussed under the descriptions 

 of the species. 



The divisions of these species of Primates into family and 

 subfamily groups is in the present state of our knowledge 

 attended with much difficulty. In one series including the 

 genera Omomys, Hemiacodon, and probably Euryacodon also, 



