INTRODUCTION. . xxvii 
Eastern Asia, whilst to the last-named family has been referred a marine Indo-Pacific 
genus, with fresh-water representatives in Australia. 
The Nearctic Region may be divided into three sub-regions, the first corresponding 
to the United States west of the Rocky Mountains and Northern Mexico to the west 
of the Sierra Madre. ‘This, the Californian sub-region, is especially remarkable for the 
paucity of true fresh-water fishes. Esocide and Percidz, in addition to the endemic 
nearctic families, are entirely absent, whilst the Siluride and Centrarchide are each 
represented by only a single species. The Cyprinide are fairly abundant, several 
peculiar generic types inhabiting this sub-region. 
The fish-fauna of the rivers of Sonora and Sinaloa is very little known; the few 
fishes which have been recorded from the Rio Sonora show that its fauna.is similar to 
that of the Rio Colorado, and the same may be said of that part of the Rio Yaqui 
which drains the western slope of the Sierra Madre. Only two strictly fresh-water 
fishes are known from streams of Sinaloa and Jalisco to the west of the Sierra Madre ; 
one of these (Cichlosoma beani) is a neotropical type, the other (Moxostoma mascote) 
a nearctic one. 
The second sub-region of the Nearctic Region includes Alaska, Canada, the United 
States east of the Rocky Mountains, and Mexico east of the Sierra Madre and north 
of the Lerma System. This sub-region may be characterized in nearly the same terms 
as those used to define the Nearctic Region as a whole: there is a northern zone 
in which the Salmonide are dominant; further south the Cyprinidae become well 
established, and in the Mississippi System attain their maximum, together with the 
Cyprinodontine, Centrarchidee, and Etheostomatine ; in the Rio Grande a number of 
characteristic types disappear. These changes in the character of the fish-fauna render 
it necessary to divide this sub-region into several provinces, the southernmost of which _ 
(Rio Grande Province) comprises the Rio Grande System, with the rivers of the 
Mexican plateau north of the Lerma System and the streams of the Atlantic slope in 
Tamaulipas and Northern Vera Cruz. 
A number of streams in Chihuahua and Durango flow in the direction of the Rio 
Grande, but never reach it, terminating in lakes which vary in magnitude according to 
the season. As has been pointed out by Dr. Meek, the nature of the fish-fauna justifies 
the supposition that these rivers were formerly tributaries of the Rio Grande. Fora 
similar reason he considers that the portions of the Yaqui and Mezquital Rivers to the 
east of the Sierra Madre also at one time belonged to the Rio Grande System, and that 
d 2 
