I9I3-] ORTMANN— THE ALLEGHENIAN DIVIDE. 353 



verse valley. If there was any stream piracy in the past it would 

 have been the Tennessee, which had the advantage over the New 

 River, so that the latter could not receive anything from the former. 

 This seems to be supported by the general character of the fauna. 

 The two cases mentioned above (Anculosa and Cambarus longulus) 

 will be taken up again further below. 



Fact L, 2, (b). 



The main fauna of the Ohio reaches, as we have seen, in the 

 Kanawha and the mountain tributaries of the Monongahela only 

 up to the lower end of the falls-line, marked by a canyon. It is clear 

 that here the upward migration of the Ohio fauna is checked by the 

 physiographical character of these streams. The upper Allegheny 

 and its tributaries are Plateau streams, originating upon the Alle- 

 gheny Plateau at elevations of about 2,000 feet (see pi. XIII, fig. 1), 

 and the West Fork River of the Monongahela falls into the same class 

 (see pi. XIII, fig. 2), and in these streams the fauna goes way up» 

 But in the case of the tributaries of the Monongahela, Youghiogheny, 

 Cheat, Tygart, and also in New River (including Greenbrier) of the 

 Kanawha system, the sources are in mountains of 3,000 to over 

 4,000 feet elevation. These rivers have a very steep grade, and in 

 a certain region they all run through a more or less well developed 

 canyon. The lower end of this canyon forms the upper boundary 

 of the Ohio River fauna in the Youghiogheny at Connelsville, Pa., 

 in the Cheat at Mont Chateau, W. Ya., in the Tygart at Grafton, W. 

 Ya., in the Xew River at Kanawha Falls, W. Va. 12 (Compare our 

 profiles, PI. XIII., fig. 2, and PI. XIV., fig. 1.) 



We have to regard it as an ecological fact among the Xajades 

 (and some other freshwater Mollusks, for instance, the genus Pleu- 

 rocera), as well as in the river-crayfishes (Ortmann, 1906, p. 412), 

 that they do not like rough water and unstable, shifting bottom. 

 The canyons of the falls-line of these rivers are, next to their upper- 



12 Of course, exceptional cases, where single species have found a way 

 up and through the canyon, may be disregarded. Such are the cases of Quad- 

 rula tuberculata and Rotiindaria tuberculata in the New River at Hinton, and 

 probably also of Symphynota costata in the Tygart at Elkins. 



