SOME EXTINCT ARGENTINE MAMMALS _ 105 
habits, we are in a position to realise the appropriate 
nature of the term “ ground-sloths” by which they are 
designated. These creatures may, in fact, be briefly 
described as edentates with a skull, teeth, and shoulder- 
girdle very similar to those of the sloths; while as regards 
their backbone and feet they come very close to the ant- 
eaters, although in the later and more gigantic forms the 
specialisation characterising the fore-feet of the latter has 
been extended to the hinder pair. 
Turning to the question of the mutual relationships and 
phylogeny of the three groups of edentates discussed in 
the course of the foregoing paragraphs, we shall have 
little hesitation in regarding the pigmy ground-sloths, 
which are the earliest known representatives of the group, 
as the direct ancestors of the gigantic megalothere. A 
modification in the structure of the teeth would equally 
well permit of their having likewise been the ancestors 
of the mylodons, which, as we have seen, possess sloth- 
like teeth. This, however, will not permit us to regard 
the mylodons as having been the forerunners of the sloths, 
seeing that the latter have a less specialised type of 
hind-foot ; and we must accordingly regard the sloths as 
a side branch derived from the pigmy ground-sloths or 
some nearly allied forms after the acquisition of cylindrical 
teeth, but before the hind-foot had acquired the specialisation 
characterising the mylodons and megalotheres. Hence 
the curious structural similarity between the front teeth 
of some of the mylodons and the two-toed sloth must be 
another instance of that parallelism in development to 
which reference has so often been made. 
With regard to the ant-eaters, we have already seen 
that the fore-foot of these animals resembles that of the 
pigmy ground-sloths in that the terminal joints of the 
