ANATOMY AND PHYSIOLOGY OF ATRINA RIGIDA. 415 
often being produced in twenty-four hours, the amount varying in different specimens 
between one-eighth and one-twelfth. In one instance a hole which measured approxi- 
mately one-half by three-fourths of an inch was cut in the shell of a vigorous young 
Atrina with the result that it was repaired in three days. Experiments performed to 
determine what parts of the mantle are capable of producing shell go to show that this 
power belongs only to the very edge and is probably confined to a small portion of 
modified epithelium located in a groove in the edge of the mantle.® (Fig. 2, a.) When 
a notch was formed in the shell by breaking out a 
piece, the edge of the mantle was quickly applied to 
the bottom of it with the result that it was soon built 
up even with the general level of the edge of the shell. 
When holes were cut in the shell at a great distance 
from the edge to see if other portions of the mantle 
could produce shell, it was found that the mantle edge 
was drawn back to these places and remained there 
until they were repaired. Although conclusive proof 
that only the edge of the mantle can produce shell is 
lacking, there is abundant evidence that Atrina gener- 
ally repairs all injuries to the shell with this part of the 
mantle, and it seems safe to assume that shell forma- 
tion is confined to this portion.® 
As stated in a previous section, the outer surface 
of the shell bristles with spines, which are distributed 
in rows. They have the same prismatic structure as the 
shell and like it they are secreted by the edge of the 
mantle. When fullyformed, theyare between one-half 
and three-fourths of an inch in length, and, except that 
they are slightly broader at the base than at the top, 
have the shape of a half tube, the hollow side of which 
faces the edge of the shell. During the growth period 
of one of these spines a little fold or tongue of the 
mantle edge extends beyond the shell and fits into the 
hollow surface of the spine. In time the shell, by its 
growth at the edge, extends beyond the spine so that the mantle no longer comes into 
contact with it. This mode of formation accounts for the fact that the spines are 
hollow and open toward the growing edge of the shell. There is no visible differentia- 
tion of the mantle edge in the form of permanent folds to which the formation of the 
spines is due. The edge of the mantie opposite a row of spines does not seem to 
differ from that located between two rows. The tongues which creep out into the 
spines are not permanent structures, but are formed by a local expansion of the mantle. 
o This refers only to the prismatic layer. The second layer, which is laid upon the inner surface of the shell at the anterior 
end, is secreted by the whole of the epithelium of the mantle of this region. 
a 
Fig. 2. — Diagrammatic cross section of the 
body anterior to the adductor muscle, a, 
Modified epithelium which secretes shell; 
b t suspensory membrane; c, descending 
lamella of outer gill; d, reflected lamella; 
e y longitudinal ridge on the mantle to which 
the gill is attached by means of interlock- 
ing cilia; f, vascular interlamellar septum; 
g , longitudinal groove in edge of gill, i, 2, 
and 5, suprabranchial chambers. 
