338 Bulletin American Museian of Natural History. [Vol. XXVII, 



lachrymal foramen, the pterygoid does not share in the backward prolonga- 

 tion of the palate; while the vertebral column differs from that of most 

 Xenarthra (except the Middle Eocene Meiacheiromys, Osborn, 1904) in the 

 lack of accessory zygapophyses in the dorso-lumbar region. The entotym- 

 panic when present is greatly reduced, whereas in Xenarthra it is well de- 

 veloped (van Kampen, 1905, p. 470). 



On the other hand the Pholitlota resemhlc the Xenarthra in the lack of 

 an interparietal, in the high number (7) of ethmoidal scrolls, in the columnar 

 stapes (Doran, 1879) (r/. the barely separated stapedial crura in Bradypodi- 

 dee and Tolypeuies) in the lack of a sinus hypotympanicus and in the large 

 size of the sinus epitympanicus (van Kampen, 1905, p. 473), while the mus- 

 culature of Manis, according to Windle and Parsons (1899), shows many 

 peculiarities not seen elsewhere except in the Xenarthra. The scales seem to 

 involve the same histological elements as do the vestigial scales on the tail 

 of the MyrmecophagidcT (c/. Weber, 1904, p. 433, fig. 330, and p. 420, fig. 316). 

 The third digit of the manus, es])ecially the claw, is much enlarged, as it is 

 in the Myrmecophagid?e and likewise forces the animal in walking to rest 

 on the outer border of the foot and on the dorsal surface of the flexed outer 

 fingers. 



Points of resemblance with several other groups are shown by the embry- 

 onic manus and pes figured by Weber (1904, p. 424). The marked func- 

 tional tridactyly in the manus, digit III being the largest, suggests Insecti- 

 vores or Hyrax; the strictly serial carpus with its rectangular elements is 

 also a point of resemblance to Hyrax. The union of the scaphoid and 

 lunar suggests the Monotremes, certain Rodents and Insectivores; the prom- 

 inence of the "prsepollex" suggests Monotremes, embryonic Didelphis, 

 Insectivores and Rodents. The centrale is not present, at least as a separate 

 element. The proximal end of the fifth digit is spreading and nearly in 

 contact with the ulna. In the embryonic pes, the lower end of the astraga- 

 lus has an extensive contact with the cuboid (cf. Amblypoda); the three 

 cuneiforms are subequal instead of the middle one being the smallest as in 

 normal mammals. The "prpehallux" is present; functional tridactyly is 

 evident but not so pronounced as in the manus. In both manus and pes 

 the digits are spreading, the pollex and hallux rather divergent; the ungual 

 phalanges of digits II-V in the adult are deeply cleft, as they are in Perameles, 

 TaJpa, and Chrysochloris (Weber); in the embryo they are split half way 

 down to the base and the opposite halves spread outward at right angles to 

 each other. 



To these various and ambiguous points of resemblance to other groups 

 Manis adds many peculiarities of its own : first, in regard to the loss of parts 

 (e. g., total loss of teeth, jugal, interparietal, clavicle, centrale carpi, third 



