336 BuUctiu American Museum of Natural Hisfor;/. [Vol. XXVII. 



covered with cement but are without enamel, and composed of irregular 

 usually hexagonal prisms of dentine; each prism in cross section shows a 

 central pulp cavity and radiating dentinal canals. In correlation with the 

 excessively large rhinencephalon, the number of ethmoid scrolls rises to 

 eleven (the highest among mammals), of which the posterior five extend 

 beneath the olfactory bulbs, much as in Echidna and are thus vertical 

 in position (Weber, 1904, p. 416, fig. 313). This condition is probably 

 secondary, however, in both forms, because the primitive position of the 

 cribriform plate (as shown in Ormthorhijnchus, Marsupials, Rodents, In- 

 sectivores, etc.) is anterior and not partly inferior to the rhinencephalon. 

 The nasoturbinal and first four endoturbinals in Orycteropus are normal 

 and resemble those in Insectivora. The carpus is very abnormal in the wide 

 separation of the lunar from the unciform leaving a lacuna between them. 

 There is no free centrale; in the U. S. National Museum specimen and in the 

 individual figured by Cuvier the scaphoid appears to be fused with the lunar, 

 but Weber states (p. 415) that the scaphoid and lunar are separated; the 

 unciform is very small. The poUex is lacking; there is a tendency toward 

 didactyly in both manus and pes. In the manus the phalanges of digits II 

 and III are symmetrically modified with reference to the median line and 

 superficially suggest those of camels ; the unguals too are more or less came- 

 loid in type, being high and compressed. In the pes the unguals are more 

 depressed; the distal keels of the metacarpals extend on to the front face. 



Comparison xvith other groups. 



Orycteropus contrasts with the Xenarthra and Pholidota in many char- 

 acters, i. e., in the possession of a distinct interparietal, in the characters of 

 the. brain, the uterus, placenta, testes and teats. The astragalus, how^ever, 

 agrees better with the Xenarthrous t}^De than with any other (cf. Ameghino, 

 1908, p. 18), and the skull and skeleton, especially the scapula and limb 

 bones, show some marked adaptive resemblances to those of Priodon. 

 The musculature, according to Windle and Parsons (1899), presents "two 

 features not found elsewhere than in the Edentata." Dr. Matthew has sug- 

 gested to the writer that the Tubulidentata may possibly have been derived 

 from the Ganodonta, because the Stylinodontidse show some interesting 

 analogies to Orycteropus in the general appearance of the cheek teeth, and 

 phalanges. 



No striking resemblances to the Rodents are evident, except the great 

 length of the entocuneiform, which is merely a primitive mammalian char- 

 acter. 



The hypothesis that the Tubulidentata are remotely related to the 



