CHAPTER X 



THEORIES AS TO THE ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION OF THE 

 TITANOTHERES AND OTHER ODD-TOED UNGULATES 



SECTION 1. ORIGIN AND RELATIONSHIPS OF THE 

 PERISSODACTYLA 



NATURE AND HABITAT OF THE ANCESTRAL 

 PERISSODACTYL 



As shown in Chapter I, the branching off of the 

 titanotheres from the other lower Eocene perissodactyl 

 families — the horses, tapirs, and lophiodonts — points 

 to the very great antiquity of the titanothere stock, 

 to basal Eocene or closing Cretaceous time, when there 

 probably existed a stem perissodactyl, doubtless a very 

 small animal. This little quadruped was probably 

 an inhabitant of the plateau regions of Asia, from 

 which its descendants diverged through adaptive 

 radiation into the ancestral nine great perissodactyl 

 families that entered Europe and North America. 

 It is certain that this adaptive radiation took place 

 long before the beginning of Eocene time, for there is 

 evidence that at the beginning of the Eocene epoch 

 at least five of these perissodactyl families had acquired 

 their distinctive family characters, especially those 

 of the teeth. 



The grounds on which are based the theory of the 

 derivation of the perissodactyls from a single perisso- 

 dactyl stem are twofold. First, the members of each 

 of the families preserve certain persistent primitive 

 (paleotelic) ancestral characters, inherited in common 

 from the perissodactyl stem form; second, each of 

 these families exhibits in common certain potentialities 

 (cenotelic) of evolutionary development, expressed 

 in the progressive or advancing characters that each 

 exhibits. 



This twofold affinity displayed in similar ancestral 

 character and in similar progressive characters is a 

 dominant feature in classification, which has been 

 exemplified repeatedly in the preparation of this 

 monograph. It is now seen to apply to orders as 

 well as to families, genera, and species. 



The common characters of members of the perisso- 

 dactyl families are listed below. 



Persistent primitive ancestral characters 



1. A complete series of upper and lower incisor and canine 



teeth, as in all primitive Ungulata, Condylarthra, and 

 Amblypoda. 



2. Upper and lower grinding teeth originally with six rounded, 



conical cusps, as in the Condylarthra and Artiodactyla, 

 but differing from those of the Amblypoda. 



3. Base of the skull with an alisphenoid canal for the trans- 



mission of the external carotid artery, as in Condy- 

 larthra and many Creodonta; reduced or absent in Artio- 

 dactyla. 



20 or 21, as compared with 

 19-20, and Artiodactyla, 



4. Dorsal and lumbar vertebrae : 



Condylarthra, which have 

 which have 20. 



5. Femur with a third trochanter, as in Condylarthra; reduced 



or absent in Artiodactyla. 



6. Manus with only four digits; poUex absent or much reduced; 



five digits in Cond3'larthra; five digits in primitive Artio- 

 dactyla. 



7. Pes with distinctive perissodactyl ankle joint, quite distinct 



from that of Condylarthra or Artiodactyla. 



Potential new progressive characters 



8. Premolars tending to transform into the molar pattern, a 



very rare character in Artiodactyla. 



9. Manus originally mesaxonic, secondarily paraxonic; in the 



Artiodactyla originally and persistently paraxonic. 



10. Tarsus losing the ancestral middle ankle joint by flattening 



the articulation between the astragalus and the navicular; 

 perfecting the tibiotarsal joint by mechanical play 

 between tibia and astragalus. 



11. Fibula gradually losing its calcaneal facet; fibulocalcaneal 



facet persistent and specialized in Artiodactyla. 



12. Progressive specialization and adaptive radiation of the 



brain, skull, teeth, trunk, limbs, and feet for a great 

 variety of habits and habitats in the nine different 

 families. 



The characters indicated in this list debar the 

 Perissodactyla from community of descent with the 

 Artiodactyla (see Gregory, 1910.1, pp. 385-386), as 

 was believed by Cope when he proposed the super- 

 order Diplarthra to include the perissodactyl and 

 artiodactyl stems. In the opinion of Osborn they also 

 debar the Perissodactyla from descent from the Con- 

 dylarthra, typified by Phenacodus of the lower Eocene 

 and Euprotogonia of the basal Eocene, as believed by 

 Cope in regarding the perissodactyl order as an off- 

 shoot from the Condylarthra. Structurally, as Greg- 

 ory has shown (op. cit., pp. 387-397), the ancestral 

 perissodactyl resembled certain more primitive con- 

 dylarths (Euprotogonia) in many features, although it 

 differed from them fundamentally in the ankle joint; 

 but in their potentiality of progressive evolution in 

 the characters shown in Nos. 8 to 12 above, the 

 momentum of the Perissodactyla can not be derived 

 from the inertia of the Condylarthra. We thus can 

 not agree with Gregory (op. cit., p. 397) that "the 

 derivation of the perissodactyl order from the general 

 insectivore-creodont-condylarth group of placentals 

 seems fairly well established." It appears more 

 probable that the perissodactyls sprang from some 

 entirely unknown progressive placental source — large- 

 brained, plastic — to which Osborn has given the name 

 Ceneutheria, as distinguished from the archaic non- 

 757 



