'2')4 On tlie GeograjiJiical Distribution of Fishes. 



Anioiica. Furtlier, it is to be remarked that several of those 

 lorms whose living representatives are most characteristic and 

 peculiar to North America, <'.<7. AmiidiV and Lcpidostcida^, 

 were in ])revious geological e[)Ochs represented by species in 

 Kuroj)e ; while the Cobitidai, now unrepresented by living 

 American species, had, as already remarked, examples in more 

 ancient times in that continent. 



Tropical Asia nourishes a great number of peculiar forms ; 

 but the relations of those are intimate either with cistroj)ieal 

 ones or with marine types. 



Africa h.\s Cyprinoids and Anabantids in common with 

 tropical Asia, and Cyprinoids in common with the cistro- 

 ])ical hemisphere; but it also supports several veiy ])eculiar 

 families for whose relations we have to seek in other con- 

 tinents. 



In tropical America are to be found the nearest relations of 

 some of these African types, and several almost or quite 

 limited to those two continents. On the other hand, in South 

 America are found several families having no analogues in 

 tlie parts of the world yet mentioned, but for which we have 

 to turn our eyes to Australia ; and there we have representa- 

 tives of not only the same families, but even, it has been con- 

 tended, one of the same species. Under these circumstances 

 we are almost compelled to believe that the iish-fauna of 

 South America Avas derived, at a distant epoch, to some ex- 

 tent from a common source with that of Africa and that of 

 Australasia. We have, however, at first sight, contrary in- 

 dications; but they are not irreconcilable : the most conspicuous 

 and, as it were, obtrusive types of South- American fishes are 

 analogues of African forms, members of the families Cichlidaj 

 and Characinida? ; but the species belong to widely different 

 genera. On the other hand, although the types common to 

 South America and Australia are not conspicuous in numbers 

 or economical importance, they are much more nearly related 

 to corresponding Australian species than the former, and, in 

 common with other facts, tend to verify Huxley's views 

 respecting an "Austro-Columbiau " fauna. 



In fine, dividing the earth into regions distinguished by 

 general iclithyological ])eculiarities, several primary combina- 

 tions may be recognized, viz. : — 1, an Arctogcean^ embracing 

 Europe, Northern Asia, and Northern America ; 2, an Asiatic^ 

 embracing the tropical portions of the continent ; 3, an Afri- 

 can, limited to the region south and east of the Desert; 4, an 

 American (embracing the America ^;r»- excellence dedicated to 

 Amerigo Vespucci), including the tropical and transtropical 

 portions; and, 5, an Aiistrahisian. Further, of these (a) the 



J 



