DISTRIBUTION OF ANIMALS, 227 
and Riitimeyer, in his treatise already cited. In what 
follows we may essentially adhere to the latter. 
Our knowledge of the regions of distribution of the 
animal world is still extraordinarily deficient. What do 
we know, for instance, of the occurrence of marine ani- 
mals? Few years only have elapsed since the depths ot 
the sea were rendered accessible to research, and the 
result has almost entirely upset our earlier notions of 
the geological significance of the sea-bottom and its 
habitability. After the strong impulse given by Maury 
to the investigation of the physical condition of the sea, 
we are now occupied in ascertaining the submarine tem- 
peratures and currents, the constitution of the sea-bottom, 
the occurrence of deep-sea organisms, and the conditions 
of their existence. We are therefore just beginning to 
collect the material for a future geography of marine 
organisms. Among terrestrial animals, certain groups 
of which the actual distribution can be defined, are use- 
less for our general purpose. 
Butterflies, for instance, which are an easy prey to 
currents of air, defy geological barriers, and, above all, 
that important partition which from the tertiary era 
has been erected, or rather excavated in the bottom of 
the sea, between Australia and India." It is the same 
with bats, and also with migratory, predatory, and 
aquatic birds; while, as Wallace shows, the other orders 
of this class are in tropical regions very reliable and 
stable inhabitants of their often limited districts, seem- 
ingly suggestive of migration. Exclusive of these, 
there remains therefore little more than the Mammalia, 
whose extraction may be inferred with certainty from 
a comparison of their present cantonments (Cantonirung), 
