550 THE MAMMALIA OF THE UINTA FORMATIOIn. 



ing depression in the calcaneum. Tlie navicular is shallower than in Triplopus, 

 resembling closely that of Hyracltyus, and showing facets for three cuneifoi-ras. 

 The cuboid is elongate and compressed in the centre as in the above forms, and 

 articulates distally with mt. iv only. The e c t o cunei f or m has a lateral facet 

 infei-iorly for mt. ii (unlike Triploims), thus reverting to Hurachyus. The m e s o- 

 c u n e i f o r m is correspondingly shortened, while the e n t o c u n e i f o r m is flattened 

 and i-otated backwards as in Hyrachyus (IsTo. 10,005, Princeton Coll.). 



In the pes of Pdlophtlierium we have a striking resemblance to that of Triplo- 

 2ms, and of the Eocene horses except for the greater reduction of the lateral toes. 

 This is simply an example of parallelism, for the teeth* show no affinity to eithei- of 

 these lines. The lateral digits are as far reduced as those of HijJjiarion.f 



4. The Rhinoceros Series. 



A. — Acer cither mm. B. — Aplieloj^s. 



The comparison of the carpus and tarsus in the hornless rhinoceroses of the 

 American Miocene shows a wide divergence between the Aceratlierium (lowermost 

 Miocene) and Aphelops (uppermost Miocene) series. The Acerntherium feet are ex- 

 tremely elongate, with the toes closely compressed, reminding one strongly of the 

 feet of the pseudo-i'hinoceros, Hyracodon, from the same beds. The more recent 

 Aplielops is generally considered a successor, but shows a widely different type of 



*Schloss(;i- (Beitriige z. Stamraesg. d. Huft.) points out the anomalous mixture of primitive and secondarj- char- 

 acters in tliis genus. Tlie premolars are reduced iu number, f, but simpler than the molars, while the molar crowns 

 are more advanced than those of Palicotherium, showing cement. 



f Lydekker (Cat. Foss. Mamm., Part IV, p. K!) follows Flower in uniting PuloploiJieritim minus and PaliBotherium 

 (see Cat. Roy. Coll. Surg., Part II, Mammalia, p. 390). The ground assigned is the presence of a " complete transi- 

 tion between the two types." On precisely similar principles we should unite Eyracotherium venikolum with Mesohip- 

 pus ; nay more, because the American genera undoubtedly belong to one line of descent, whereas the two French 

 genera, judging by the extremities and teeth, belong to different lines. Further, is nomenclature actually simplified 

 when a genus is made so comprehensive as to extend over long geological and wide geographical areas, embracing 

 such a variety of forms that it becomes necessary (vide Rhinoceros and Palaoiherium, opera cilata) to arrange the spe- 

 cies into groups? We think not. Genera can no longer be defined in the Linnajan sense. Lamarck, a century ago, 

 anticipated the difticulties which would arise when the gaps between Cuvier'sand Button's genera were filled up. The 

 range of variation which should be included in a single genus is largely a matter of arbitrary judgment. Never has 

 the problem been more perplexing than among the unbroken series found in the western Tertiaries. The writers, for 

 tlie present, have adopted the rule long since employed by Marsh and Cope, viz , diflorences of degree, size, propor- 

 tion, may be talven as specific ; dill'erencos of kind (in the number of teeth or digits) or of actual form («. g., tlie as- 

 sumption of the molar pattern by the premolars), may I)e taken as generic. This rule, when applied, for example, to 

 the horse series, is found to work admirably — each of the subdivisions of the Eocene is characterized by a genus — and 

 by a number of species varying in the direction of the lower and higher forms. Judging by the thickness of the 

 strata, these subdivisions represent long periods of linu^ 



