20 MEMOIRS OF THE CARNEGIE MUSEUM. 



also seems to be somewhat different, F. flava undata (Barnes), belonging to the 

 larger rivers of the central basin, being absent in Pennsylvania. In the middle 

 West (Illinois) these two forms seem, however, to overlap, and in the Southwest 

 also both seem to be present, and, according to what Walker says, intergrades are 

 present. The real /''. flava undata is, in consequence of the higher beaks, more 

 subtrigonal (not subtrapezoidal) in outline, and in addition has the tendency in 

 the epidermis to become greenish rather than brownish. 



From the measurements given above we see that some of our specimens agree 

 rather closely with Lea's figure of U. trigcmiis. Others reveal transitions 

 in the direction of F. flava. Specimens from C'harlcroi have been determined by 

 Simpson as intergrades between trigona and ndnginosa {— flava), and this is 

 entirely correct. We have here (or rather had) in the Ohio below Pittsburgh, and 

 in the Monongahela above, a region, where flava gradually passes into trigona. 

 Further upstream, chiefly in the Monongahela system, only more or less typical 

 F. flava are found. The only way to bring order out of the chaos is to draw an 

 artificial line, as I have done. 



Localities in Pennsylvania represented in the Carnegie Museum: 



Ohio River, Neville Island, Allegheny Co. 



Monongahela River, Westmoreland Co. (G. A. Ehrmann); and Charlcroi, Washington Co. (G. A. Ehr- 

 mann). 



Other localities represented in tJie Carnegie Museum: 



Elk River, Sutton and Gassaway, Braxton Co., and Shelton, Clay Co., West Virginia."^ 

 West Fork White River, Riverside, Greene Co., Indiana (J. D. Haseman). 

 Wabasb River, New Harmony, Posey Co., Indiana (A. A. Hinkley). 



(A form indistinguishable from this has been received from various localities in the Southwest; 

 but they cannot be distinguished from U. chuni Lea.) 



Marais des Cygnes River, Rich Hill, Bates Co., Missouri (W. I. Utterback). 

 Cache River, Nemo, Craighead Co., and Sedgwick, Lawrence Co., Arkansas (H. E. Wheeler)." 

 White River, Cotter, Baxter Co., Arkansas (A. A. Hinkley). 

 Saline River, Benton, Saline Co., Arkansas (H. E. AVheeler).'* 

 Kiamichi River, Tuskahoma, Pushmataha Co., Oklahoma (F. B. Isoly). 

 Sabine River, De Soto Parish, Louisiana (L. S. Frierson). 



Distribution and Ecology in Pennsylvania (See fig. 3) : F. flava trigona is found 

 in Pennsylvania only in the Ohio and lower Monongahela Rivers, and reaches here 



'° Smaller than the Pennsylvanian form, but agreeing in all other respects. A few specimens from 

 Gassaway would fall under typical flava, but the diameter remains above fifty percent. 



1' In part cotypes of F. sckda Wheeler (Nautilus, XXVIII, 1914, p. 76, PI. 4). I cannot separate 

 them from this form. 



" In some of these, the dianaeter fs^lls below fifty-five percent, but remains above fifty percent. 



