340 Bulletin American Museum of Natural History. [Vol. XXVII, 



assign to the Edentate series the rank of a subclass; Thomas (1887), im- 

 pressed by the anomalous modifications of the dentition, suggested that the 

 Edentate orders might be collectively designated as "Paratheria," in contrast 

 with the Metatheria and Eutheria. Elliot Smith (1898), from his compre- 

 hensive studies on the brain of Edentates concludes that they are remotely 

 related to the Carnivores, perhaps by descent from early Creodonts. 



The general drift of these opinions is that the Edentates are the highly 

 aberrant descendants of some very lowly and ancient mammalian stock. 

 And certainly the modern Edentates retain many characters of low aspect. 

 In addition to the columella-like stapes of the Sloths and Toli/peutes and the 

 relatively large coracoid of the Ground Sloths may be mentioned the follow- 

 ing, gleaned from Weber (1904) and other authorities. A septomaxillary 

 presumably homologous with the element of the same name in certain 

 reptiles, has been recorded (Broom, 1906.1) in Dasypus. The brain is ma- 

 crosmatic and frequently of low type (Elliot Smith); the cochlea auris makes 

 onlv two to two and a half turns; the testes usuallv remain abdominal; 

 the penis is often on a low stage and in Sloths is even of an embryonic type; 

 the vagina sometimes shows traces of former longitudinal division, although 

 the opposite uteri are completely united; the yolk sack is large; two venae 

 cavjie persist; indications of the former presence of scales are not lacking 

 (Weber, 1904, pp. 432-433); the tail is sometimes very thick at the base, as 

 is often the case in archaic mammals. Under ophthalmoscopic examina- 

 tion, the eye, according to Lindsay Johnson (1901), reveals certain features 

 in common with Alonotremes, Marsupials and Rodents: e. g., the fundus is 

 of uniform color with hardly perceptible stippling, the lens appears as if 

 composed of concentric circles. 



Comparison of the Xenarthra tcith other Orders. 



The Xenarthra agree with ty})ical Placentals in the general features of 

 their development (Parker, 18S5-'86). The structural resemblances of the 

 Xenarthra to the Tubulidentata and the Pholidota have already been dis- 

 cussed. The question of their relationshij) with the Ganodonta (= Ta^nio- 

 donta) is still open. According to Wortman (1897), these Basal to Upper 

 Eocene forms tend to connect the Xenarthra with the Creodonta, but Pro- 

 fessor Scott (1903-05) is inclined to regard the resemblances between the 

 Ganodonta and Xenarthra as due to convergence; because the Patagonian 

 INIiocene forms, which might be expected to approach the Ganodonta, in 

 reality show less resemblance to them than do the modern Edentates. Dr. 

 Ameghino also, whose knowledge of the fossil Xenarthra is very extensive, 



