NO. 9 MAMMALS FROM CAVES IN HAITI MILLER 2'J 



as those which exist in greater degree between the femora of Cholospus 

 and Bradypus. The less anterior directing of the neck in the Haitian 

 femur is perhaps most readily made apparent by applying the proxi- 

 mal extremity of the bone to a flat surface in such a way that it is 

 supported by the tripod formed by the posterior surfaces of the head 

 and the two trochanters. The shaft of the bone in Acratocnus ( ?) 

 comes then takes a position essentially parallel with the flat surface. 

 When the femur of A. odontrigonns is similarly placed the shaft rises 

 above the flat surface at an angle ranging from about i8° to about 

 23°. The same difference may be observed by tracing the direction 

 of the low but usually evident ridge which crosses the neck from 

 the head to the lesser trochanter. In Acratocnus odontrigonns this 

 ridge extends so obliquely to the inner surface of the femur that its 

 line, when continued downward, passes beyond the contour of the 

 bone at a point situated near the mid portion of the head of the tro- 

 chanter; in the Haitian specimen it passes out nearly 10 mm. farther 

 down the shaft. The lesser inward bend of the neck is best appreciated 

 by " sighting " down the anterior or posterior surface of the shaft of 

 the bone ; it then becomes obvious that the head lies nearer to the 

 main axis in the Haitian specimen than in any of those from Porto 

 Rico. 



Remarks. — The femur on which this species is based resembles 

 in all its general characters the corresponding bone of the Porto 

 Rican ground sloths and of the Miocene South American Hapalops. 

 The peculiarities which I have described as distinguishing it from the 

 femur of Acratocnus odontrigonns separate it equally from the cor- 

 responding bone of Hapalops, at least so far as can be determined 

 from Scott's figures of three species {longiceps, pi. 32. elongatus. 

 pi. 41, and ruetimeyeri, pi. 42). 



Other remains which I refer without much hesitation to Acratoc- 

 nus ( ?) comes are as follows : (a) the proximal two-thirds of a right 

 tibia (pi. 8, fig. i ) not certainly distinguishable from the corresponding 

 part of the tibia of a Porto Rican ground sloth (No. 17711, Amer. 

 Mus. Nat. Hist.) ; (b) an almost perfect atlas (pi. 10, fig. i) of the 

 proper size to fit a skull of Acratocnus odontrigonus ; several canini- 

 form teeth, both upper and lower, agreeing in a general way with 

 those of the same animal; (c) foot bones and ungual phalanges 

 resembling those of the Porto Rican species. 



On the basis of the femur and of the parts which appear to be 

 almost certainly associated with it I do not now feel justified in separat- 

 ing the small Haitian ground sloth more than specifically from Acratoc- 



