PEARLS AND PEARL FISHERIES. 
407 
varying extremely; several describe the shells as occurring in small patches of a few 
feet square, but the large majority agree in giving the beds an elongated shape, 
either along the banks or on shoals in midstream. In the smaller rivers they extend 
all across, up to a width of the stream of 100 yards (Tennessee). The length of these 
beds is estimated at from a few yards to several hundred, or in some cases a mile or 
even 4 miles (Arkansas). They are in some cases reported as upon sandy or gravelly 
bottom, in shallow water of moderate swiftness, and a few speak of the shells as 
wedged in among the crevices of rocky or stony bottoms. Very few refer to still water 
or mud. 1 
In two papers (Florida and Illinois) some of the shells are described as in the 
bank, from 1 to 4 feet below the surface of the water. This occurrence is peculiar, 
and it would be of interest to ascertain what species possess this habit. 
In several instances the shells are reported as packed side by side on the bottom 
so closely as to be like a pavement (Tennessee and Wisconsin), and sometimes several 
layers deep in places where there are “holes” in the bed of the stream (Wisconsin). 
The Florida paper states that in lakes the beds extend around the shore, their 
breadth determined by the depth of the water. 
There is a general agreement that the midstream beds are upon shoals or connected 
with islands, bars, or rapids. But the detailed statements vary, some placing them 
above and others below rapids, and likewise as to islands. Evidently they occur for 
the most part in places where there is a moderately rapid flow, but somewhat protected 
from tbe full force of strong currents. Some interesting particulars are given. One 
paper (Tennessee) says that the shells lie in beds from shoals up to deep water, where 
there is rock bottom, and then in crevices in the rocks; and two others (Tennessee) 
are somewhat similar. Another (Tennessee) reports them as usually at the head of 
an island above the “breakers,” usually opposite the bluff side. An Iowa paper speaks 
of the beds as extending along bold banks until the current changes to the opposite 
side, i. e., on the swifter (convex) sides of the curves. The author of a Maryland 
paper states that the beds vary in location with the varying distribution of the sand 
and mud of the bottom, the shells traveling correspondingly if the changes are not 
too sudden. One paper from Texas refers to their seeking and occupying positions 
where they are best protected from the force of the current in freshets. 
It is clear, from all these varying accounts, that the location of the shell-beds is 
determined by conditions which depend on the size and the rapidity of the stream 
and the nature of its banks aud bottom; the main requisite being water of a very 
moderate depth, flowing freely but gently, and so producing almost always a sandy 
or gravelly bbttom. In shallow streams these conditions would extend all across; in 
larger ones they would be found near the shores, or associated variously with islands, 
bars, and rapids. In slow streams, the shells would naturally be found on the convex 
sides of the curves, where the swifter current erodes the banks; in more rapid ones 
they would seek the slower portions of the river, aud avail themselves of the shelter 
of islands, etc., as a defense against the violence of freshets. This last agency is 
spoken of by several, in the answer to another inquiry, as being highly destructive, 
especially to the younger and the smaller shells; hence, those without the protection 
of some island or shoal above them would be most liable to be swept away and 
destroyed in flood time. 
1 But ou this point, see pp. 395-397, above, as to mode of occurrence in Arkansas bayous. 
