ORTMANN: monograph of the naiades of PENNSYLVANIA. 105 



the greatest width is situated in the region of the posterior ridge, but not in the 

 anterior part of the shell, as in E. dilatatus. This anterior swelling, and the tapering 

 posterior end of the shell, distinguish this latter species from E. violaceus. 



In other respects E. violaceus is extremely variable. First of all in size and 

 shape (longer or shorter) , then in the degree of compression (compare our measure- 

 ments). Furthermore the color of the outside and of the nacre is variable, and 

 finally the epidermis may be more or less smooth, or may be roughened by con- 

 centric lamellae. Since the shell is not very particular as to station (see below), 

 it is not astonishing that the responses to the environment are mmierous, and 

 that it assumes in many places peculiar features, which, however, arc not at all 

 constant, and may turn up anywhere under proper conditions, and are connected 

 by innumerable intergrades. There is no tendency in Pennsylvania to develop 

 local races or varieties, and I am perfectly satisfied that there is only one species 

 in our region. 



"Species-making" within this form has gone beyond all bounds, and in a 

 large number of Lea's "species," chiefly from the southern states, the question 

 may be raised whether Lea was actually in earnest, when proposing them, or 

 whether he only wanted to mystify contemporaneous and subsequent students 

 of naiadology. Great credit is due to Simpson for straightening out the worst of 

 this tangle; but I think that Simpson has not gone far enough. I shall mention 

 here a few additional instances of species created by Lea, which in my opinion 

 simply fall as synonyms under E. violaceus, merely representing individual phases. 

 This chiefly concerns such forms as are actually found in our state. 



Unio complanatus jejunus (Lea): Simpson (1914, p. 658) unites jejunus and 

 percoarctatus Lea, and makes this a variety of the ]:)resent species. Both are 

 southern types (from the Carolinas), and both are said to be more compressed, 

 while jejunus is reported to have an inclination to be biangulatc behind, and 

 percoarctatus is said to have the surface covered with loose concentric striae. All 

 these characters are found in the Pennsylvanian E. violaceus in certain individuals. 

 They may be very much compressed, they have very often a tendency to be bi- 

 angulate behind, and they have often a rough epidermis, with loose stria?. I have 

 specimens, which completely agree with jejunus, as well as with percoarctatus, and 

 sometimes such specimens (See Plate VIII, fig. 5) prevail at certain localities, 

 but they are always connected with the normal form by intergrades. 



Unio roanokensis Lea, and northamptonensis Lea: Simpson (1914, p. 666) 

 regards these as one species, distinct from our species. The form northamptonensis 

 is reported from as far north as Massachusetts, and might be expected in Penn- 



