290 ORTMANN—CORRELATION OF SHAPE AND 
mixture of some specimens of barnesiana. From Sevierville, I have 
ten specimens, ranging from 45 to 38 per cent., the average being 
42 per cent., thus representing barnesiana with an interspersing of a 
few bigbyensis. I have no doubt, however, that, if there was any 
material from farther up stream, pure bigbyensis would appear. 
Finally, the same conditions obtain in the Tennessee drainage in 
northern Alabama. In the Tennessee River at Florence is a form 
(12 specimens measured) in which the diameter ranges from 64 to 
50 per cent., with the average at 56 per cent. This is pure tumes- 
cens. In the tributaries we find mostly the more compressed forms 
of the barnesiana- and bigbyensis-type, so, for instance, in Elk 
River, at Fayetteville (1 specimen), a form with the diameter of 
44 per cent. (barnesiana), and farther up, at Estill Springs (8 
specimens), a form ranging from 40 to 35 per cent., with the average 
of 38 per cent., which corresponds to bigbyensis. 
Thus the contention is substantiated, that twmescens of the large 
rivers passes through barnesiana into bigbyensis of the headwaters. 
Group OF AMBLEMA PERUVIANA (LAMARCK). 
The extremes of this group of forms are marked by Amblema 
peruviana (Lam.) (= Quadrula plicata Simpson, ’14, p. 814) and 
Amblema costata (Raf.) (== Quadrula undulata (Barn.) Simpson, 
p. 819). But there are other described “species”? which fall into 
this group. 
Wilson and Clark (1914) and Utterback (1916) have referred to 
these as showing the same phenomenon of a swollen form inhabit- 
ing the large rivers, and a flat one in the smaller streams (Cumber- 
land and Osage River systems). I have no doubt that this is 
correct, and that it applies also to other rivers. 
According to my experience, there is no question that specimens 
of A. costata in the Ohio River below Pittsburgh are more obese 
than the greatly compressed specimens found, for instance, in the 
Beaver and upper Allegheny drainages. However, I do not possess 
sufficient material from the middle and lower Ohio, to show the 
actual transition into peruviana. Also in the Tennessee-system, I 
did not go far enough down to reach the range of the latter: what I 
