DISTRIBUTION AND MIGRATIONS 275 



Labyrinthic fishes. Of the Cat-fishes there are nine famihes in 

 the Neotropical region, all of them peculiar to this part of the 

 world. The Cichlids [Cichlidae) are for the most part confined 

 to Africa and Central and South America (Fig. loi), and 

 certain genera of the Old and New World appear to be very 

 nearly related. Gyprinodonts {Cyprinodontidae) are likewise 

 found in both continents, but, like the Cichlids, they are fond 

 of living in brackish-water lagoons, and cannot be regarded as 

 strictly fresh-water fishes. Finally, the Lung-fishes are repre- 

 sented by one peculiar genus in each of the two continents. 



How is this obvious similarity between the fish-faunas of 

 Africa and South America to be explained? Some geological 

 authorities hold that the great ocean beds are permanent 

 features, and that the similarities are due to migrations from 

 the north. Others believe that in past times the distribution 

 of sea and dry land was very diflferent from that seen to-day, 

 and that the resemblance between the animals of the two 

 countries may be explained by assuming the existence of a 

 former land connection between them. It is impossible to 

 discuss the mass of evidence for and against both of these 

 conflicting theories, but it may be pointed out that that drawn 

 from a study of the distribution of fresh-water fishes lends 

 strong support to the latter view. It seems more than likely 

 that in Cretaceous times the two land masses formed a single 

 continent (with India), in which the Ostariophysi originated 

 and underwent their evolution, and that at the end of this period 

 Africa became permanently separated from South America. 

 At this time South America was probably separate from North 

 America, and the subsequent connection of the two by the 

 formation of the Isthmus of Panama at the end of the Eocene 

 period, which led to a great southward migration of mammals 

 and reptiles, seems to have had very little effect on the fresh- 

 water fishes. Certain typically Neotropical forms, such as 

 some of the Characins and Cat-fishes, have pushed as far north 

 as the southern part of Mexico, but none have succeeded in 

 reaching the great Mexican plateau (Fig. loi). In the other 

 direction a species of Gar Pike [Lepidosteus tropicus) has extended 

 as far as Panama, but although a few other Nearctic fishes 

 have reached Central America, none have penetrated to South 

 America. It will be observed, therefore, that the very distinct 

 Nearctic and Neotropical fish-faunas, separated by the volcanic 

 chain of mountains which stretches across Mexico from Colima 

 nearly to Vera Cruz, scarcely meet, much less overlap, a state 



