348 Louis Agassiz on the Natural 
America, the hippopotamus of Africa, and the walruses of the 
Arcties ; others which have only a small number of representatives 
beyond the fauna which they specially characterize, as, for in- 
stance, the marsupials of New Holland, of which America has a few 
species, such as the opossum; and again, others which have a 
wider range, such as the bears, of which there are distinct species 
in Europe, Asia, or America, or the mice and bats which are to be 
found all over the world, except in the Arctics. That fauna will, 
therefore, be most easily characterized, which possesses the largest 
number of distinct types proper to itself, and of which the other 
animals have little analogy with those of neighbouring regions, as, 
for example, the fauna of New Holland. 
The inhabitants of fresh waters furnish also excellent een 
for the circumscription of faune. The fishes, and other fluviatile 
animals, from the larger hydrographic basins, differ no less from each 
other than the mammalia, the birds, the reptiles, and the insects of 
the countries which these rivers water. Nevertheless, some authors 
have attempted to separate the fresh water animals from those of 
the land and sea, and to establish distinct divisions for them, under 
the name of fluviatile faunee. But the inhabitants of the rivers and 
lakes are too intimately connected with those of their shores to allow 
of a rigorous distinction of this kind. Rivers never establish a sepa- 
ration between terrestrial faunze. For the same reason, the faunz 
of the inland seas cannot be completely isolated from the terrestrial 
ones ; and we shall see hereafter that the animals of southern Europe 
are not bound by the Mediterranean, but are found on the southern 
shore of that sea, as far as the Atlas. We shall, therefore, distin- 
guish our zoological regions according to the combination of species 
which they inclose, rather than according to the element in which 
we find them. Yh 
If the grand divisions of the animal kingdom are primordial and 
independent of climate, this is not the case with regard to the ulti- 
mate local circumscription of species ; these are, on the contrary, 
intimately connected with the conditions of temperature, soil, and 
vegetation. A remarkable instance of this distribution of sitimals 
with reference to climate may be observed in the Arctic fauna, which 
contains a great number of species common to the three continents 
converging towards the North Pole, and which presents a striking 
uniformity, when compared with the diversity of the temperate and 
tropical faunze of those same continents. 
The Arctic fauna extends to the utmost limits of the cold and 
barren regions of the North. But from the moment that forests ap- 
pear, and a more propitious soil permits a larger development of 
animal life and of vegetation, we see the fauna and flora not only 
diversified according to the continents on which they exist, but we 
observe also striking distinctions between different parts of the same 
continent; thus, in the Old World, the animals vary, not only from 
