IN THE ALLEGHENY REGION OF SOUTH-WESTERN VIRGINIA. 239 



qommon species of the Lakes and Ohio. Neither this nor any other bony gar is 

 known to the fishermen on the Kanawha above the falls, near the mouth of the 

 Greenbriar, although the species is common in the Allegheny and its tributaries. 



>• PETROMYZONTID^. 



PETROMYZON Linn. 



A species of this genus inhabits the upper waters of the Holston. The landlord at 

 Glade Springs, Mr. Thompson, informed me that he had observed one attached to the 

 side of a black bass {Mlcropterus), from which it had torn the scales, and nearly 

 penetrated the abdominal cavity. He also mentioned seeing a black sucker (C. 

 nigricmis) violently excited, and making great exertions in the water, and unable to 

 pursue a direct course. It became at last so exhausted as to be taken from the water 

 by the hand, when a Petromy?:on was found attached to one side. Oliver N. Bryan, 

 who resides opposite Mount Vernon, on the Potomac, informs me that it is not 

 uncommon to find them attached to the shad, and that they destroy them by their 

 continued suction. 



Sect. in. — Conclusions, 



Before drawing any general conclusions, I will quote from Agassiz his statement of 

 some of the propositions in distribution ; we will see later what approach to their 

 solution can be made : 



" I would remark that there is still another very interesting problem respecting 

 the geograpliical distribution of our fresh-water animals, which may be solved by the 

 further investigation of the fishes of the Tennessee River. This water-course, taking 

 the Powells, Clinch and Holston Rivers as its head-waters, arises from the mountains 

 of Virginia in latitude 37°; it then flows south-west to latitude 34° 25', when it turns 

 west and north-west, and finally empties into the Ohio under the same latitude as its 

 sources in 37°. 



" The question now is this : Are the fishes of this water system the same through- 

 out its extent? in which case we should infer that water communication is the chief 

 condition of the geographical distribution of our fresh-water fishes; or do they difler 

 in different stations along its course ? and if so, are the differences mainly controlled 

 by the elevation of the river above the level of tlie sea, or determined by climatic 

 influences corresponding to differences of latitude ? We should assume that the first 

 alternative were true if the fishes of the upper course of the river differed from those 

 of the middle and lower course, in the same manner as in the Danube, from its 

 source to Pestli, where this stream flows nearly for its whole length under the same 

 parallel. We would, on the contrary, suppose the second alternative to be well 



61 



