XXX IjNTTEODUCTIOX. 



and by some peculiar Cyprinodonts, sucli as Belonesocc and Xiphophorus. This may 

 be termed the Guatemalan Province. Costa Rica and Panama comprise a fourth 

 province, in which many of the South-American types attain their northern limit 

 (e. g., Jcara and Gcopliagus of the Cichlidae, Curimatus, Gastroj^elecus and some other 

 Characinidae, the Loricariidoe). In this, the Isthmian Province, the only nearctic form 

 is Lepidosteus tropicus. 



5. The Shoke-Fishes of the Atlantic and Pacific Coasts of Mexico and 



Central America. 



There is a great similarity between the fishes found on the Atlantic and Pacific 

 coasts of Central America. It was formerly stated that a considerable proportion of 

 the species were the same on both coasts, but in recent years the number of supposed 

 identical species has been greatly reduced by more detailed comparison. 



In the present work only those groups of shore-fishes which enter fresh water are 

 considered ; as a result of my studies on these I am inclined to think that if we 

 eliminate pelagic or semi-pelagic fishes of Avide distribution, very few species will be 

 found to be identical on both coasts, and that these will probably be fishes which 

 enter rivers for considerable distances. Such fishes may have become involved in 

 those geological changes which have led to the occurrence of the same fresh-wa,ter 

 species in rivers of the Atlantic and Pacific Slopes. 



Many of the species which were formerly considered to be the same on both coasts 

 are now found to be distinguished by slight but constant differences. In such cases, 

 where an Atlantic species is more closely related to a Pacific one than to any on its 

 own side, it is reasonable to suppose that both are derived from a parent species which 

 inhabited the neighbouring parts of the two oceans at the time when there was a 

 connection between them. Such a marine connection appears to have existed in the 

 Eocene over what is now the Isthmus of Panama, and there are good reasons for 

 believing that it ceased at the beginning of the Miocene. 



A comparison of the shore-fishes of the two coasts should therefore give data 

 as to the nature of the specific characters which may arise during isolation and as to 

 the time required for specific differentiation in certain groups. 



The following species are included in the systematic part of the present work and 

 may be paired together, the members of each pair satisfying the definition that they are 

 more closely related than either is to any other species, and that they represent each 

 other on the Atlantic and Pacific coasts : — 



