252 BULLETIN NO. VII. 
the semilunes are interpolated folds or figures, and here certain 
transitional stages seem to show that the outer and inner 
enamel layers are parts of the same layer. The details may be 
gathered from the figures. 
CONCLUDING REMARKS ON THE DISTRIBUTION OF RODENTIA. 
The rodents are perhaps the most widely distributed of the 
orders of Mammalia, yet but one of its families (Muride), in- 
cluding the universally distributed rats and mice, is found 
represented in all the major land areas. In this respect the 
hares and squirrels (Leporide and Sciwride) come next, they 
being found everywhere except in Australia and Madagascar. 
In the former the only members of the placental Mammalia are 
six genera of mice of which four are not found elsewhere, 
while in Madagascar three genera peculiar to that island alone 
occur. South America is richest in rodents and, except for a 
few genera shared with North America, is the exclusive habitat 
of its species. Thirty-seven genera are credited to it of which 
thirty-two are endemic. 
Africa is next in numerical richness and affords about thirty 
genera, several of which are found also in India and the Medi- 
terranean province, only twenty-one being endemic. The North- 
ern continents are less rich in species and exhibit the custom- 
ary uniformity. The palearctic region has some twenty-five 
genera of which only about fifteen are not found either in India, 
Africa or North America. Southeastern Asia is remarkably 
poor in species, and the few genera considered endemic are 
nearly allied to widely distributed groups. North America, 
although possessing several peculiar types, is closely similar 
to the Asiatico-European continent in its rodents. ) 
These facts lead to the assumption that the rodents are de- 
rived from several ancestral lines which have continued inde- 
pendent in the principal discrete land-areas. The principal 
centres of divergence are supposed to be in South America and 
Africa. Most of the important families are represented by 
remains as early as Eocene Tertiary. In Europe, during that 
period, there lived members of the existing genera Sciwrus, 
and Myoxus. There is some reason to suppose that the squir- 
rels appeared in America rather later than in Europe, although 
the synchronousness of epochs of like names in the two conti- 
nents is pure assumption. Remains of Sciwrus have not been 
found earlier than the Miocene here, while in the Eocene there 
were types of such extinct genera as Paramys and Sciwravus. 
