Introduction. xlvii 



It is often argued that fishes are taken accidentally by water birds 

 from one body of water to the other. We have no positive evidence 

 that this has ever been done, but we have enough negative evidence 

 to warrant us in doubting its possibility. Without going into the 

 discussion of this subject I will mention an example bearing upon it 

 which is at least suggestive. When Shoshone and Lewis lakes in the 

 Yellowstone Park were discovered, there were no fishes in them, 

 while only a few miles distant were lakes and streams quite abun- 

 dantly supplied. These streams and lakes were stocked through 

 Two-Ocean Pass* from the head waters of the Snake River, but the 

 falls in the Lewis River prevented the fishes from entering Lewis and 

 Shoshone lakes. If water birds were at all active agents in the dis- 

 tribution of fishes, these two lakes would have undoubtedly been well 

 stocked. 



As shown by the foregoing lists, there are in Mexico four quite 

 distinct fish faunas. Two of these have migrated or resulted through 

 migrations from the north, one from the south, while the other had its 

 origin within the country. The fish fauna of northern Mexico is 

 essentially that of the Rocky Mountains and eastern United States. 

 The two large rivers which have furnished highways through which this 

 portion of Mexico became stocked with fishes are the Colorado and the 

 Rio Grande. The former flows into the Gulf of California, the latter 

 into the Gulf of Mexico. In their upper courses these two rivers are 

 near each other, but their fishes are not the same. The only fish 

 common to both river basins is a small dace, Rhinichthys dulcis (Gir- 

 ard), which is also found in the head waters of the Arkansas, the 

 Missouri, and the Columbia rivers. From the Colorado River thirty- 

 two species of fishes are known, twenty -two of which are thus far 

 peculiar to this basin. Only four or five species of Colorado River 

 fishes are at present known from the Rio Sonora and the Rio Yaqui; 

 however the lower courses of these two rivers, where we would expect 

 Colorado river fishes, have been but little explored. 



There are in all about eighty-seven species known from the Rio 

 Grande Basin; and according to Dr. Evermann and Dr. Kendall, f 

 twenty-three of these are found in the Wabash River in Indiana. 

 In the Rio Grande Basin in Mexico there are seventy species. It is 

 interesting to note that eight of these have been found in the head 

 waters of the Rio Yaqui, and eight in the head waters of the Rio 

 Mezquital. In fact, the fish fauna of the head waters of these two 

 streams is essentially that of the Rio Grande. 



*Evefmann, Bull. U. S. Fish Comm., 1891, 24. 

 fBull. U. S. Fish Comm., 1892, 57-126. 



