THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



759 



While we are unable to consider Euproiogonia the direct 

 ancestor of the Perissodactyla, yet it had many characters 

 approximating it to them. It is probable that the common 

 ancestor of the typical Ungulata was thoroughly an unguiculate 

 and that the first separation into the phyla of phenacodonts, 

 perissodaotyls, and artiodact3'ls accompanied or preceded the 

 development of hoofs from claws [p. 300]. * * * This 

 evidence, though not conclusive, points unmistakably toward 

 an alternating carpus as the primitive one. In Phenacodus 

 the carpus is more nearly serial, although it varies in different 

 individuals, and there is always a lunar-unciform contact. 

 But if Phenacodus is a direct descendant of Euprotogonia 

 this serialism must be secondary [pp. 308, 309]. We must go 

 somewhat lower down than the Torrejon [that is, Thanetian] to 

 find the junction of the equine and phenacodont phyla. 



Osborn (1898.148, p. 174) accepted Matthew's 

 conclusion and asreed that the 



•^ puraconc 



primitive ungulate foot was 

 interlocking, or alternating 

 (lunar on unciform, scapho- 

 centrale on magnum), deriv- 

 able from that of such a 

 creodont type as Dissacus, 

 from which the carpus of the 

 Amblypoda and Condylarthra 

 and possibly of all Ungulata 

 may be derived. 



Gregory (1910.1, p. 385) 

 observes : 



The question of the derivation of 

 the perissodactyls as an order and 

 of their relationship to the Con- 

 dylarthra is still open. Cope saw 

 in Phenacodus the atavus of practi- 

 cally all the hoofed orders. Osborn 

 holds the contrary opinion that 

 Phenacodus is a hoofed offshoot of 

 the Creodonta and a member of the 

 Meseutheria, or small-brained Cre- 

 taceous-basal Eocene orders, and 

 that the perissodactyls have sprung 

 from some entirely unknown 

 "Caeneutheria" (p. 457). * * * 



StUl another view may be adduced — that although neither 

 Phenacodus nor Euprotogonia were the ancestors of the Peris- 

 sodactj'la yet they resemble those forms more nearly than do 

 any other known mammals; and that the basal Eocene ancestors 

 of the Perissodactyla would, if discovered, fall under the 

 superorder Protungulata as defined above (p. 383). 



The Protungulata are then defined as typified by 

 the condylarth Euprotogonia. Gregory reaches the 

 conclusion (op. cit., p. 396) that 



The ancestral perissodactyl as thus conceived resembled the 

 more primitive eondylarths such as Euproiogonia in many 

 features (p. 397). * * * To conclude, the derivation of 

 the perissodactyl order from the general insectivore-creodont- 

 condylarth group of placentals seems fairly well estab- 

 lished. * * * On the other hand, genetic derivation from 



any well-known eondylarths {Phenacodus, Euprotogonia, 

 Meniscotherium) is almost equally improbable. But with 

 regard to many important dental and osteological features it 

 is obvious also that Euproiogonia and Phenacodus bridge 

 over the structural gap between the perissodaotyls and the 

 lower unguiculate orders and, in brief, that the stem of the 

 perissodactyls would very likely fall under the Condylarthra, as 

 redefined by Matthew (1897). 



After summing up all the characters of the stem 

 perissodactyls and comparing them in detail with the 

 characters of Euprotogonia Gregory concludes (op. 

 cit., p. 389): 



It is not denied that the ancestral perissodactyl had a larger 

 brain case and better brain than Euprotogonia. It is merely 

 inferred that there was a considerable variation in brain charac- 



■Evolution of the upper molar pattern in eondylarths and titanotheres 



In the earliest known titanotheres, Lambdotherium (D) and Eotitatiops (E), the upper molars had already attained the 

 bunoselenodont pattern, having two low, subcorneal internal cusps and two very large external V-shaped cusps. The 

 probable mode of origin of this pattern is, however, furnished by certain of the Eocene eondylarths (A, B, C). 



A, Protogonodon, a basal Eocene condylarth. In this stage the molar is nearly triangular and the primitive tritubercular 

 pattern is still evident. All the cusps are low; the protoconule and metaconule are well developed. The hypocone 

 (hy) is just growing up from the cingulum. 



B, EtipTotogonia, another basal Eocene condylarth. In this animal the hypocone is more progressive, so that the tooth is 

 becoming subquadrate. 



C, Edocion, a lower Eocene condylarth. This stage shows the assumption of the V-like form of the crest of the 

 paracone and metacone, the reduction of protocone and metaconule, the origin of the mesostyle, and the ineipiently 

 V-shaped form of the hypocone. 



D, Lambdotherium, a lower Eocene aberrant titanothere. The mesostyle is now well developed, and the crests of the 

 paracolic and metacone are completely V-shaped. The parastyle is strong, the metaconule is much reduced, and 

 the hypocone is now subequal to the protocone. The paracone and metacone have also shifted inward toward the 

 center of the crown. All the cusps have retained their low or bunoid form. 



E, Eotitanops, a lower Eocene true titanothere. The molar pattern is now definitely titanotheroid. The protoconule and 

 metaconule are vestigial, but the main outer cusps (pa, me) are still low. 



F, Manieoceras, a middle Eocene titanothere. In this stage the anteroposterior diameter of the molar is considerably 

 increased, the tips of the paracone and metacone are shifted well in toward the middle of the crown, the whole outer 

 wall of the tooth has become considerably deeper, the protoconule and metaconule have practically disappeared, 

 and the protocone and hypocone are low, rounded cusps. 



G, Brontotheriam, SilowQr Oligocene titanothere. Marked intensification of the tendencies noted under F. Paracone 

 and metacone tips at the center of the crown, outer wall of tooth much elongated, anteroposterior diameter for the 

 molar increased. This completes the transformation of a tritubercular molar, with simple conical cusps, into a 

 quadrangular and bunoselenodont quadiitubercular molar, with W-shaped ectoloph and low, conic internal cusps. 



ters in the Condylarthra and that the perissodactyls sprang 

 from some unknown, possibly Asiatic, larger-brained form, 

 which in all its dental and skeletal characters would faU under 

 the Condylarthra as here defined. 



EOCENE CONDYLARTHEA NONPERISSODACTYL 



Thus in the opinion of Matthew (1897) and Gregory 

 (1910) the Perissodactyla sprang from unknown 

 primitive Condylarthra of Cretaceous age. 



The opinion of Osborn (1894-1918) is that since 

 the known Condylarthra have specialized in a dif- 

 ferent direction from the Perissodactyla, the latter 

 probably sprang independently from a different 

 ungulate stock. The evidence for Osborn's opinion 

 is in part as follows: 



