310 ORTMANN— CORRELATION OF SHAPE AND 



be clue, by any means, to accident. Thus we must accept it as 

 demonstrated, that in most shells which shozv a decrease in diameter 

 in the upstream direction, this decrease is compensated, to a degree, 

 by the greater sise of the shell as expressed by its length. 



Further, we have seen in certain forms (Quadrnla metanevra 

 and Qu. cylindrica, Dromus dramas), that a pecuHar kind of sur- 

 face sculpture, namely large knobs and tubercles, tend to disappear 

 in the headwaters. This is also connected, more or less distinctly, 

 with a flattening of the shell. It would be interesting to discover 

 additional cases of this kind, and may be found in Truncilla toridosa 

 and gubernacnhim of the upper Tennessee. But it surely is not a 

 general law, since there are other shells, which show no change in 

 sculpture along the course of a river, and since there are other in- 

 stances, where the opposite seems to be the case. The group of 

 Amblema plicata seems to belong here, w4iere a strongly sculptured 

 and flat form (A. costata) belongs to the headwaters, while a 

 smoother and more swollen form (peruviana) is in the largest 

 rivers.-" 



In this connection it should be pointed out, that a similar case is 

 known in the freshwater Gastropod lo. The facts have been posi- 

 tively established by C. C. Adams-'^ ; a tuberculated or spinose form 

 (lo spinosa Lea) is found in the larger streams; a smooth form (lo 

 fluvialis Say) in the headwaters. The present writer is able to con- 

 firm this by his own observations. The smoothness of the speci- 

 mens of lo in the upper Powell, Clinch, and Holston in Virginia is 

 very striking ; tuberculate individuals begin to appear farther down, 

 and real spiny ones not till the state line of Tennessee is reached. 

 The transition is quite gradual. 



Conclusions. 



Certain Naiades change their shape along the course of one and 

 the same river in such a way, that 



/. the more obese (szvollen) form is found farther dozvn in the 



-- See Wilson and Clark, '14, p. 63; Utterback, '16, p. 41. 

 ^^Mem. Nat. Acad. Sci., 12, 1915. 



