STATION IN FRESH-WATER MUSSELS. 271 
i 
H. Smith; and finally I had the privilege of examining upper Ten- 
nessee shells in the collection of Mr. B. Walker in Detroit, col- 
lected by C. C. Adams (particulars about these may be found in 
my paper on the upper Tennessee Naiades). 
Of course, the shells collected at one and the same locality vary 
somewhat, and thus it was necessary to compute the average for 
each set. In some cases, only few specimens were at hand, in 
others, a great many of them. Thus I do not give, in the tables, the 
individual measurements, but only the extremes and the average. 
We shall see that this is entirely sufficient to establish the law. 
In each river, the localities are arranged from downstream upward, 
taking up the tributaries of the main stream generally in ascending 
order. The exact situation of each locality may be found on the 
map (page 258). 
According to obesity previous authors frequently have dis- 
tinguished separate species, and often corresponding forms in dif- 
ferent river-systems have received different specific names. Thus 
it has been a difficult task to bring order in the nomenclature, and 
I am compelled to devote considerable space to this. Since it will 
become apparent that the various groups of forms treated actually 
are connected by intergrades, each group should be regarded, taxo- 
nomically, as one species. But since, on the other hand, the ex- 
tremes are often rather sharply contrasted, and since they very 
generally appear as geographical (or rather ecological) races, it is 
convenient and justified, to use the old specific names in a varietal 
(or subspecific) sense. On account of the existing intergrades it is 
difficult, even impossible, to draw sharp lines between these varie- 
ties, and the dividing lines selected by myself (according to per- 
centage of diameter) may seem, and are, arbitrary: I have en- 
deavored, in this respect, to preserve the older names as far as 
possible. Sometimes it was possible to distinguish only two forms, 
a swollen one and a flat one; in other cases, three have been ad- 
mitted (an intermediate one being added to the two extremes). 
This latter course was made necessary in those instances, where the 
oldest name was given to an intergrading form. 
