270 ORTMANN—CORRELATION OF SHAPE AND 
_and obliquum* and in the case of the group of Fusconaia barnesiana.® 
I have given a fuller account, including the synonymy of the forms 
concerned. Additional references may be found in my paper on the 
upper Tennessee shells® and on the Pennsylvanian Naiades.’ 
However, all this information is rather vague, and not supported 
by detailed measurements, and, before we finally assume that this 
law exists at all, we should substantiate it by more careful investi- 
gations. The present paper is written with the purpose to supply 
the details and to reduce them to figures. 
When we speak of the obesity of mussel shells, and call them 
flat and compressed, or convex and swollen, we refer to the pro- 
portional diameter of the shell. This diameter can be expressed in 
various ways, but I found that the most easy is in percentage of 
the length. The length is always measured parallel to the liga- 
ment,* which can be done, without difficulty, by placing the shell 
between the arms of a vernier caliper measure in the proper posi- 
tion. The diameter is measured at right angles to it, and always the 
maximum diameter of the two valves is taken, no matter where it is 
located. The length is then taken as=100, and the diameter ex- 
pressed in hundredths of the length. 
The material studied has been collected, for the largest part, 
by myself, and thus I am able to vouch for the correctness of the 
localities. It forms part of the collections of the Carnegie Museum 
in Pittsburgh. Yet there is additional material from other sources. 
The Carnegie Museum possesses a fine collection from the Tennessee 
drainage in northern Alabama and southern Tennessee made by H. 
4 Nautil., 23, 1910, p. 117, footnote 2, and Ann. Carnegie Mus., 8, 1912, 
p. 264. 
5 Nautil., 31, 1917, p. 58. 
9 Proc. Amer. Philos. Soc., 57, 1918, p. 521 ff. 
“Mem. Carnegie Mus., 8, 1919. 
8 The greatest length of the shell is sometimes in a diagonal direction, 
when the shell is “ oblique”; but since the obliquity varies much individually, 
the length parallel to the ligament is selected. Previous authors (Lea for 
instance) have not strictly adhered to this rule, and thus we may explain 
certain discrepancies of the figures given in his text, and those taken from 
his illustrations. In my measurements, I have always taken Lea’s illustra- 
tions as the standard. 
