274 ORTMANN—CORRELATION OF SHAPE AND 
Ohio have a diameter almost uniformly above 50 per cent.; in the 
_region around Pittsburgh (in Ohio and Allegheny), it varies around 
50 per cent.; and in French Creek it falls below 45 per cent. (always 
barring minor irregularities). 
Already this first instance substantiates the general law pro- 
pounded in the introduction: im the larger rivers, these shells are 
more convex and swollen; in the headwaters, they are flat and com- 
pressed; and in the intermediate parts, the intergrades between the 
extremes are found. , 
Under these circumstances it is, indeed, hard to draw a line 
between the two forms called subrotunda and kirtlandiana. Ac- 
cording to measurements taken from Lea (Dia. 69 per cent.) there 
are specimens which are even more obese than any of those ex- 
amined by myself (max. 67 per cent.), and these swollen forms 
differ very strikingly from those which have only 44 per cent. (as 
in Lea’s type of kirtlandiana), or even less (falling to 33 per cent. 
according to Simpson and my material). In order to include politus 
Say in the synonymy of subrotunda, as Simpson has done, I have 
decided to draw the line at the diameter of 50 per cent., so that indi- 
viduals with this diameter and over shall be called Fusconaia sub- 
rotunda, and individuals with the diameter less than 50 per cent. 
shall be called Fusconaia subrotunda kirtlandiana. Thus it is evi- 
dent that two of the shells of subrotunda, for which Simpson gives 
the measurements, having the diameter 44 and 49 per cent., fall 
under kirtlandiana, 
This will bring almost all specimens in the Ohio below Penn- 
sylvania under the main species, while all specimens in French Creek 
are the var. kirtlandiana. In the Pennsylvanian part of the Ohio, 
and in the Allegheny, we have the two forms associated, sometimes 
the one, sometimes the other prevailing, and the various sets from 
these localities should be separated accordingly. This is only what 
we should expect as the natural condition; and also, for practical 
purposes, this clears matters up, and brings sets together, which are 
rather uniform. 
One additional circumstance should be pointed out. The obesity 
of the shell changes a httle with age, so that young shells, in the 
