1882.] I'ROF. W. H. FLOWER ON THE EDENTATA. 361 



III tracing true affinities we must not look to these obviously 

 adaptive characters, but to others of really deeper significance. 



To commence with the skeleton, one of the most striking characters 

 by which the Myi-mecophafjidce differ from ordinary mammals is 

 the presence on the posterior do;sal and the lumbar vertebrae of 

 accessory articulating surfaces, in addition to the true zygapophyses 

 common to all mammals, and causing a remarkable interlocking of 

 the arches of these vertebrae. Such articulations are found equally 

 well-developed in the Megatheriidce and in the Dasypodidce, and 

 in a comparatively rudimentary state in the Bradtjpodidce, especially 

 in Bradi/pus ^ 



The whole vertebral column of the Sloths, especially the dorsal 

 and lumbar region, is poorly developed, evidently in relation to the 

 subordinate function of the muscles attached to it. In the ordinary 

 position the animal hangs below the branches of the trees in which 

 it dwells, the trunk being merely slung between the two pairs of 

 extremities ; progression is effected chiefly by the body being 

 dragged along by the fore limbs ; the hind legs, which usually take so 

 active a part insupportingand propelling the trunk, aresmall and weak, 

 andthefunctionsof walking, running, andjumpingare alike in abeyance. 

 Hence all the processes of the vertebral column — the spinous, trans- 

 verse, and accessory — are but feebly developed. The existence then, 

 even in a rudimentary form, of these additional articulations is ex- 

 tremely significant. It may almost be said that they prove that 

 the Sloths are descended from animals in which they existed in a 

 fully developed form. On the other hand, like as in some respects 

 the vertebral column of Manis is to that of Mynnecophaga, not a 

 trace of either of these articulations or of the processes on which they 

 are situated are to be found in the former. There is a strong inter- 

 locking of the lumbar vertebrse ; but it is formed by the greatly curved 

 form of the true zygapophyses, and not by the addition of any 

 superadded parts. In Orycteropus also they are entirely absent. 

 On this ground alone we might be justified in assuming that the 

 Old-World Edentates are not closely related to the American forms. 



Nothing can be more different than the characters of the sternum 

 q{ My nnecophaya and of Manis. In the former the numerous meso- 

 sternal segments are small, later.nlly compressed, and articulated with 

 each other and with the strongly ossified sternal ribs by synovial 

 joints. In Manis the sternum is broad and flat, and the sternal 

 ribs only partly ossified, and connected with it in the normal manner. 

 In these characters, the Bradypodidoe, Megatheriidce, and Dasy- 

 podidce agree with the Auteaters, and Oiycteropus is more nearly 

 related to Munis. 



The gigantic post-Tertiary Edentates of the New "World, Mpga- 

 therium and its allies, throw much light upon the close affinity of 

 the Sloths and true Auteaters. By common consent they have been 

 placed among the former when the order is divided into two divisions, 

 and are spoken of as " Ground-Sloths ;" yet in many important 

 characters, perhaps in all those not relating to the functions of 

 ' See ' Osteology of Mammalia,' 2nd edit., 1876, p. 55. 



