DISTRIBUTION AND MIGRATIONS 269 



lie at the southern end of the Mexican plateau) from the Pacific 

 Ocean in remote times, before they had been cut off from the 

 sea by inaccessible falls. The only other fishes in the lakes are 

 a few Cyprinids which have found their way down from the 

 rivers of North America. 



The true fresh-water fishes, that is to say, the fishes which 

 have evolved in the rivers and lakes, may form distinct families, 

 such as the Sun-fishes [Centrarchidae] of North America and the 

 Perches (Percidae) of the Old and New World, or even whole 

 orders such as that which include the Pikes, Mud-fishes, and 

 Black-fishes (Haplomi). To such fishes the sea may be looked 

 upon as constituting a definite and generally impassable barrier, 

 and their distribution is limited by factors which are rather 

 diflferent to those governing the geographical range of marine 

 fishes. The distribution of the members of the order Ostariophysi, 

 including the majority of the fresh-water fishes of the world, is 

 full of interest, providing valuable evidence as to the past 

 history of the continents, their former connections with one 

 another, and the lines at which they were originally severed. 

 There is good reason to believe that their evolution has taken 

 place in fresh water, and that their dispersal, necessarily slow 

 as compared with that of land animals, has been very gradually 

 eflfected by hydrographical changes, among which the capture 

 by one big river of the tributaries of another, the union of two 

 or more rivers due to the elevation of the land, or the joining of 

 two river-systems, the head-waters of which may be separated 

 by only a few miles of swampy land, during abnormal floods, 

 are probably the more important. It has been suggested that 

 a species may have become established in a river system from 

 which it was previously absent through the spawn being carried 

 considerable distances by aquatic birds, through the agency of 

 water-spouts, and by other accidental methods, but there is no 

 direct evidence of such transferences having occurred. It seems 

 more than probable that the present distribution of these fishes 

 was accomplished mainly at the beginning of the geological 

 epoch known as the Tertiary, and that the subsequent land 

 connections and interchanges which had such important eflfects 

 on the movements of the mammals and reptiles did little to 

 influence the distribution of the fresh-water fishes. 



In considering the distribution of the true fresh-water fishes, 

 and particularly those of the order Ostariophysi, the land masses 

 of the globe may be conveniently divided into^^a number of 

 zoogeographical regions. These are: (i) an Australian region 



