24 MEMOIRS OF THE CARNEGIE MUSEUM. 



creek-form of the river-form flava trigona. As I have pointed out under flava 

 trigona, the latter has (or had) in the Ohio and Monon^ahela near Pittsburgh a 

 distinct tendency to become more flattened. There are all intergrades in this 

 respect before me, and it is absolutely impossible to draw a natural line between 

 the swollen trigona-iorm, and the flat flava-iorm. Farther up in the Monongahela 

 and Allegheny, and in their tributaries, and also in the smaller tributaries of the 

 Ohio proper, the pure and tjq^ical F. flava is exclusively found. 



Conditions like these force us to unite these forms specifically. But never- 

 theless trigona is to be regarded as a distinct variety, constituting a geographical, 

 or rather ecological, race of F. flava, which has its definite habitat in larger streams. 

 There are indications that similar conditions prevail outside of Pennsylvania. 



Attention should be called to the singular parallelism of these forms with 

 those of the Fusconaia subrotunda-gronp. In the latter, we have seen that there 

 is a greatly swollen form, with highly elevated and incurved beaks (F. ebena) 

 living in the largest, deepest rivers, with muddy bottoms, and therefore in those 

 regions, which are nearest to the center of the interior basin. In the great rivers 

 with sandy and gravelly bottoms and somewhat stronger current, that is to say 

 in the upper Ohio, this is replaced by a form, which, although more or less swollen, 

 has less elevated beaks {F. subrotunda). This in turn in the smaller streams 

 gives way to a flat form (F. hirtlandiana) . The same is true apparently in the 

 F. flava-group. We observe a very swollen form with high beaks having the 

 Mississippi as the center of its range {F. undata) ; a swollen form with less elevated 

 beaks in the upper Ohio {F. trigona) ; and a flat form in the smaller creeks (F. flava) . 

 To these is added, in this case, a dwarfed lake-form in Lake Erie {F. parvula). 



The boundaries of these corresponding forms of the two series do not coincide, 

 inasmuch a F. subrotunda goes farther up in the rivers than does F. flava trigona. 



It is well to keep this peculiar phenomenon in mind, for later on we shall 

 become acquainted with other instances of the same kind. 



Further, attention should be called to the peculiar fact that, while the form 

 flava is entirely missing m northwestern Pennsjdvania, and is represented in Lake 

 Erie by the form parvula, it turns up again in western New York. The Carnegie 

 Museum has specimens from the environs of Rochester, New York, and it had 

 been previously reported from the region of Buffalo and the Erie Canal, from 

 the Genesee River and the Mohawk, crossing over to the Atlantic slope (See Call, 

 1878, and 1885; Dewey, 1856; IMarshall, 1895; Baker, 1898b). According to 

 CaU, it has migrated along this route in recent times. Since this range has no 

 connection with the rest of the range of F. flava, we must assume that it came 



