102 MOSTLY MAMMALS 
are concerned, as closely allied to the sloths. It may be 
added that the great divergence of the two series of 
teeth in the mylodon skull indicates the presence during 
life of a tongue of great width and size. Mylodons had 
a number of ossicles, like large beans, embedded in the 
outer surface of the skin; but in the nearly allied 
glossothere, of which portions of skin covered with long 
sloth-like hair have been discovered in a cave in Pata- 
gonia, nearly similar ossicles were embedded in the inner 
side of the skin. Strange to say, these ground-sloths 
appear to have been kept in caves as domesticated 
animals by the ancient inhabitants of Patagonia. 
Thus far I have shown how the ground-sloths are 
related to the sloths in the characters of their skulls; 
but other members of the group, known as the scelido- 
theres (Scelidotherium), although still retaining the same 
number of teeth, present a certain approximation in these 
respects to the ant-eaters. Thus their skulls, instead of 
being short and broad like those of the mylodons, are very 
long and narrow, and have the muzzle much produced in 
advance of the anterior teeth. Indeed, it would require 
only a still greater elongation and narrowing of the skull 
of a_ scelidothere, coupled with the total loss of the 
teeth, to produce one very similar to that of an ant-eater. 
So far as I am aware, palaeontologists have not yet 
been able to trace a complete transition from the gigantic 
ground-sloths of the Pleistocene deposits of Buenos Aires 
to their diminutive representatives from the older Tertiary 
deposits of Patagonia, although it is known that some of 
the species from the intermediate formations were inferior 
in point of size to their more recent allies. It is, how- 
ever, very interesting to find that the pigmy ground-sloths 
of these Patagonian deposits had transversely ridged 
