Grier: Morphological Features of Mussel-shells. 163 



The above measurements were resolved to factors for comparison 

 by division with the length, with the exception of the thickness of 

 the shell, which it seemed desirable to compare with the dextro- 

 sinistral diameter, as giving more of a biological balance to this 

 measurement, making it independent of any extraneous physical 

 condition which might be involved. To facilitate the work of calcu- 

 lation, a Mannheim slide-rule was used, since the writer found by 

 checking with calculations made by long hand that he was able to 

 secure a fair degree of accuracy. Computations were usually made 

 to the fourth place. 



In recording the data thus obtained, it was found convenient to 

 use a standard figuring book of twelve columns, double-page form, 

 obtainable from most stationers. This provided ample space for any 

 comments to be attached to any particular group. In comparison, 

 the shells were first grouped into their respective drainages, after which 

 the calculated data for the different localities were arranged in the 

 order from headwaters to mouth of stream, and in the lake region from 

 east to west. (See accompanying map, PI. I.) Since the problem had 

 to adjust itself to the amount of material on hand, it was not always 

 possible to compare as large a number of shells as desirable from 

 some localities. The smallest number was three, but since the results 

 obtained consist of the average for the most part of a fairly large 

 number of shells they are thought to reasonably justify the results 

 obtained. The procedure was ordinarily to compare equal numbers 

 of shells in each case. The kinds of comparison were as follows: 



1. Comparison of Naiades of Lake Erie with their representative 

 species in the streams entering the lake, principally the Maumee 

 Drainage. 



2. Comparison of these species of Lake Erie with their representa- 

 tive species in the streams of the Upper Ohio Drainage (Upper Alle- 

 gheny and tributaries, Beaver and tributaries, etc.), as close as possible 

 to the divide. Species from the latter drainages were at the same 

 time compared with those of the Maumee. 



The physiographical and geological considerations leading me to 

 make my comparisons under these two heads are found in the sections 

 dealing with "Physical Conditions" and "Characteristics of the 

 Mussel Fauna." It is true that in Pennsylvania few or none of the 

 shells have gone over into the lake, yet excellent authority states that 

 the shells of the Upper Ohio resemble in a striking way those of the 



