W.-B. SCOTT — MAMMALIAN FAUNA 243 
pler than that which prevails among the Pampean genera. The skeleton, 
though typically glyptodont, yet shows many interesting points of 
approximation to the armadillos, which indicates that these two orders 
are more nearly allied than either is to any other edentate group. 
The Gravigrada are extraordinarily numerous and varied, so variable 
indeed that it is not yet possible to make any satisfactory taxonomic 
arrangement of them. More important is the fact that the three families 
which occur in the Pleistocene are already recognizable and that the 
ancestors of nearly all the Pleistocene genera of both North and South 
America may be identified. As in the other groups of edentates, the 
Santa-Cruz Gravigrada are of relatively small size, some of them extre- 
mely small, and in structure they are far more primitive than their 
Pampean successors. Time will permit the mention of only a few of the 
more significant differences between the ground sloths of the two epochs. 
1. The dental formula is invariably */, and in each jaw the first tooth is 
always more or less distinctly eaniniform. 2. The trunk is very long and 
consists of numerous vertebræ, as many as 25 dorso-lumbars; the articu- 
lations of the posterior thoracies and of the lumbars is less complicated 
than in the Pampean ground-sloths or the armadillos and anteaters of 
the present day. 3. The humerus always has an epicondylar foramen and 
the femur has a large third trochanter. 4. All the parts of the manus 
and pes are free and uncoôssified; the feet are always pentadactyl and 
the digits all have the full number of phalanges, including the claw. 
From the form of the astragalus and calcaneum it is evident that the 
foot was plantigrade and that rotation, so as to bring the fibular border 
to the ground, could have been present only in an incipient degree. 
A careful study of the Santa-Cruz Gravigrada reveals important resem- 
blances, not only to later members of the same group, but also to the true 
sloths and the anteaters, and we have the strongest evidence that all the 
American edentates are descended from a common ancestry, to which 
the so-called edentates of the old world have no apparent relationships. 
As is well known, continental South America contains no Insectivora 
at the present day, and it is therefore somewhat surprising to find re- 
presentatives of this order in the Santa-Cruz beds. AS yet only a single 
genus (Necrolestes) has been obtained and this is a somewhat proble- 
matic type, concerning which I am not yet prepared to express any very 
definite opinion. Among recent insectivorous genera, Chrysochloris is 
undoubtedly the one which most resembles Necrolestes and, if this re- 
semblance is really due to relationship, as it seems to be, it gives inte- 
resting suggestions as to the former land connections between South 
America and Africa. 
No Cheiroptera have yet been found, but that is, of course, no sufficient 
reason to doubt their presence in South America at that period. 
