THEORIES AS TO ORIGIN, ANCESTRY, AND ADAPTIVE RADIATION 



795 



teeth as the titanotheres, namely, the bunolophoseleno- 

 dont, gradually evolved hypsodont molars with highly 

 complex enamel foldings and thereby attained an 

 efficient grinding mechanism capable of triturating 

 the driest and hardest kinds of grasses. (See fig. 718.) 

 Consequently the imperfect browsing and grazing 

 mechanism of the teeth may have been one 

 of the principal causes of the extinction of the 

 titanotheres. (See Chap. XI.) As more fully 

 discussed in Chapter XI all bunoselenodont 

 Eocene perissodactyls and artiodactyls became 

 extinct in late Eocene or early Oligocene time 

 except only the chalicotheres, which theoret- 

 ically retreated to the forests and retained the 

 pure browsing habit. 



GEADUAI CHANGE OF DIET 



The evolution of the grinding teeth in the 

 titanotheres may indicate that all the phyla at 

 first fed upon coarse tubers, roots, and the 

 twigs and leaves of low bushes. The cypto- 

 cephalic Dolichorhinus may have fared upon 

 the harder and smaller varieties of plants and 

 thus presented a nearer approach to a true 

 grazer. The Oligocene titanotheres probably 

 fared upon larger and coarser varieties of plants, 

 which were plucked up by the powerful pre- 

 hensile lips. The degeneration or complete loss 

 of the incisor teeth was supplemented by the 

 prehensile lip action. The swollen premolar 

 teeth of the ProtitanotJierium stage, ancestral 

 to Brontops, seem adapted for browsing and 

 crushing food rather than for grinding. The 

 pestle and mortar construction of the Oligocene 

 titanothere grinding teeth implies that the food 

 required to be broken and crushed before being 

 ground. The very size of the molars seems to 

 imply very coarse food. The selection of food was 

 apparently never of the harder siliceous varieties 

 sought by the contemporary rhinoceroses and 

 horses, otherwise hypsodoiit evolution might have 

 been more rapid. The shape of the skull in Oligo- 

 cene forms seems to imply that the head was not 

 rapidly thrown backward, but that the food was 

 sought near the ground, among the low shrubs. 

 The browsing African rhinoceros R. (Opsiceros) 

 hicornis seems to seek shrubs and browsing food in 

 a country affording admirable opportunity for 

 grazing. 



diameters of the teeth, whereas in the mesaticephalic 

 titanotheres there is an excess of the anteroposterior 

 diameters over the transverse. The facts set forth 

 below should be noted: 



1. All the species measured show a considerable 

 degree of fluctuation in the anteroposterior and trans- 



Cer. 



Figure 715. — Brain of Menodus compared with that of Rhinoceros and 

 other quadrupeds 



Proportions of brain in archaic fleft column) and modernized (right column) mammals of 

 similar size. Dots show olfactory lobes; black lines show cerebral hemispheres; dashes show 

 cerebellum and medulla. All one-fourth natural size. Modified after Osborn, 1910.346. 

 A, Araocyon, Eocene flesh eater; Cams (the dog), modern flesh eater. B, Plicnacodus, 

 Eocene primitive ungulate; Sus, the domestic pig. C, Coryplwdon, Eocene ungulate; 

 Ehinoceros, living rhinoceros of same size. D, Uintalherium, massive Eocene ungulate 

 E, Menodus giganteus, Oligocene titanothere. 



HARMONY OF PROPORTIONS OF HEAD AND GRINDING 

 TEETH 



A consideration of the proportions of the head and 

 the grinding teeth affords a demonstration of the 

 principle that the evolution of the 11 principal char- 

 acters of the titanothere skull and dentition is corre- 

 lated with pervading bjachycephaly or dolichocephaly. 

 In the brachycephalic Oligocene titanotheres there is 

 an excess of the transverse over the anteroposterior 



verse diameters of all the teeth, a fact which may be 

 due in part to geologic crushing of the rock matrix 

 and in part to actual variability in proportions. 



2. There seems to be a higher degree of fluctuation 

 in the diameters of individual teeth than in the total 

 anteroposterior measurements of p'-m^ which may 

 indicate compensatory growth. 



3. Apparently each species has a typical mean for- 

 mula and curve for the excess of the transverse over 



