80 A JOURNEY IN BRAZIL. 
Old World, and in all th<} rivers and lakes of the temperate 
zone, communicating with the Atlantic Ocean. They occur 
in smaller numbers in most tributaries of the Mediterra- 
nean, but are common in the Volga and Danube, as well 
as in the Mississippi, in some of the rivers on our north- 
ern Atlantic and Pacific coasts, and in China. This fam- 
ily has no representatives in Africa, Southern Asia, Austra- 
lia, or South America, but there is a group corresponding 
in a certain way to it in South America, that of the Go- 
niodonts. Though some ichthyologists place them widely 
apart in their classifications, there is, on the whole, a 
striking resemblance between the Sturgeons and Gonio- 
donts. Groups of this kind, reproducing certain features 
common to both, but differing by special structural modifica- 
tions, are called ' representative types.' This name applies 
more especially to such groups when they are distributed 
over different parts of the world. To naturalists the com- 
parison of one of these types with another is very interest- 
ing, as touching upon the question of origin of species. To 
those who believe that animals are derived from one another 
the alternative here presented is very clear : either one of 
these groups grew out of the other, or else they both had 
common ancestors which were neither Sturgeons nor Goni- 
odonts, but combined the features of both and gave birth to 
each. 
" There is a third family of fishes, the Hornpouts or Bull- 
heads, called Siluroids by naturalists, which seem by their 
structural character to occupy an intermediate position be- 
tween the Sturgeons and Goniodonts. There would seem 
to be, then, in these three groups, so similar in certain fea- 
tures, so distinct in others, the elements of a series. But 
while their structural relations suggest a common origin. 
