MUSSELS OF CUMBEIU.AND RIVER AND TRIBUTARIES. 21 



approximate rather than absohite. But, even so, thoy will be of 

 service to tli« mussel fishermen, for whom they are primarily intended. 

 Only a very small percent Jif3:e of the shells seen and handled could 

 be kept for the final collection. 



An endeavor was made to retain typical specimens of each species 

 encountered, and also all puzzling and aberrant forms, since the latter 

 add much to the actual knowledge of a species, though they may 

 render positive identification more diflicult. 



SUMMARY OF MUSSEL DISTRIBUTION. 



The practice of the Bureau of Fisheries in examining a river and 

 its tributaries from source to mouth, in regular order, throws unex- 

 pected light on the distribution of species which could be obtained 

 in no other way. The fauna of a river has a coherence never found 

 and not to be expected in an artificial division of the country, such as 

 a towmship, county, or State, whose boundaries are purely arbitrary. 

 The larger the river and the more thoroughly the main stream and its 

 tributaries are examined the more illuminating become the results. 

 The study of the entire fauna of the Cumberland River and its tribu- 

 taries leads to the following general conclusions, which are r.mply 

 confirmed in all the river faunas that have been examuicd: 



1. Wlien two closely related forms difi'er essentially in their degree 

 of inflation, the flatter and less inflated one wUl be found in the upper 

 portions of the river and in the tributaries, while the rounder and more 

 inflated one is confined to the lower portions of the main river, where 

 there is a weaker current and more mud. To this there are, however, 

 some noteworthy exceptions, such as Symphynota complanata. 



2. The swiftness of the current, the size of the stream, and the 

 kind of bottom affect other shell characters besides that of inflation. 

 Consequently, where there is a mixture of conditions there is also a 

 mixture of characters, and two species which in other localities may 

 be well defined and easily separated will be found to merge imper- 

 ceptibly into each other. In a miscellaneous collection of shells it is 

 easy to find the blue-point (Quodmla undulata) from one stream and 

 the three-ridge {Q. plicata) from another, the southern mucket 

 {Lampsili.9 ligamentina gihha) from one locality in a State and the 

 pocketbook (L.ventricosa) from another. But when specimens of the 

 entire fnuna of a river are spread out on a table in order from the 

 source to the mouth there is found such a mingling of characters that 

 it is often a mere matter of individual judgment to determine some 

 of the species. This is essentially true of Q. undulata and Q. perpH- 

 cata in the upper portions of the Cuniborl.-md. 



3. There is sometimes a pecuhar similarity in the faunjis of widely 

 separated tributaries, where the conditions at first would seem to be 



