1910.] The American Edentates. 339 



trochanter); second, in regard to the unicjue modification of parts {e. g., 

 xiphisternum, stomach, extreme development of scales, extreme length of 

 tail). Finally there are primitive Placental characters: e. g., large olfactory 

 organs, large cerebellum, moderately small cerebrum, very large yolk 

 sack, i)rimitive conditions of the testes and accessory glands and of the 

 female reproductive organs and independent optic foramen (c/. Weber, 

 1904, pp. 425-429). The dorso-lumbars either retain the primitive number 

 of 19 (D 14, L 5) or rise to as many as 22. 



The possibility of relationship with the Xenarthra is indicated by the 

 characters cited above, especially by the presence of a reduced entotympanic ; 

 but Manis may equally well, so far as other evidence goes, be a highly modi- 

 fied survivor of vmknown Mesozoic Placentals. A South American origin 

 for the group has been suggested by Ameghino (190(5, p. 377), who figures 

 an astragalus described under the name " Argijromani.s- jxtfagonica," which 

 shows a resemblance to the astragalus of Manis; but such evidence is often 

 insufficient to establish relationship. 



The Xenarthra. 



In 1696 John Ray, finding that certain mammals would not fit into his 

 dichotomous scheme, collectively characterized the Armadillo, Tamandua 

 Sloth and others as "Quadrupeda vivipara unguiculata anomala"; and it 

 must be confessed that after more than two centuries of research the Eden- 

 tates are still "anomala" both to the systematist and phylogenist. It has 

 indeed been shown that the Aard-Varks and Pangolins are ordinally separate 

 from each other and from the American Edentates, and as regards the latter 

 it has further been shown that the Sloths, Ground Sloths and Antbears are 

 relatively nearly related and form a natural suborder, the Pilosa of Flower or 

 Anicanodonta of Ameghino, coordinate with the suborder Loricata of Flower 

 or Hicanodonta of Ameghino, including the Armadillos and Glyptodonts; 

 but no structurally intermediate forms between these two suborders have 

 been discovered even in the lowest Tertiary deposits of Patagonia, and in 

 spite of an extensive knowledge of the numerous recent and fossil forms, the 

 relationship of the whole order, both to the other "Edentates" and to the 

 typical Placentals, remains very largely a matter of speculation. 



Cuvier and many later writers, perhaps on account of the fossorial adapta- 

 tions common to Monotremes and Edentates, treated the two groups as if 

 they were more or less closely related. W. K. Parker (1885-86) compared 

 the columnar, imperforate stapes of the Sloths with that of the Monotremes, 

 and both Parker and Ameghino, noting especially the relatively large size 

 of. the coracoid in the Ground Sloths, attempted to derive the shoulder- 

 girdle of Edentates directly from the Monotreme type. Gervais proposed to 



