Foerste.] 
326 
[May 1, 
diana, in any of our collections. As far as we can learn, the species 
does not show any great degree of variability, and in all our forms 
presents characteristics much the same as in the European types. 
The following description is drawn up from the forms found at 
Cumberland Gap, Tennessee, where it is very abundant. 
Dorsal valve flattened, with a straight hinge line for five-sixths 
the width of the shell, the postero-lateral margins however rounded, 
and the anterior margin of the shell semi-circular. There is a faint 
median depression increasing in depth towards the anterior margin, 
and which may gain still further in distinctness by a slight increase 
in width and depth of that one of the grooves separating the ra- 
diating plications which occupies a median position. Some of the 
specimens show two small projections along the hinge margin near 
the beak. The beak does not extend beyond the hinge line. 
Ventral valve convex, with often a moderate tendency towards 
the formation of a median ridge along the back of the shell, from 
the beak to not quite the middle of the shell. 
In many specimens this incipient ridge is entirely wanting, and 
in none is it well characterized. The beak always projects beyond 
the hinge margin, and is evenly incurved. The postero-lateral an- 
gles are rounded and the anterior margin is semi-circular. 
Measurements of a considerable number of valves show that in 
both valves the length does not equal the width of the shell. In 
about 50% of both valves the ratio of the length to width of the 
shell varies from 8 to 10 to 9 to 10. While 30% of the dorsal valves 
show a relation of the length to the width, varying from 7 to 10 to 
8 to 10, this is not true of a single specimen of the ventral valve. 
Moreover while 50% of the ventral valves show a relation of length 
to width varying from 9 to 10 to 9.5 to 10, only 10% of the dorsal 
valves show a similar ratio. These measurements show that the 
comparative length and breadth of the valve are very variable, but 
that the extremes of variation in the dorsal valve tend towards an 
increase in the width of the shell while that of the ventral valve 
expresses itself in the increased rotundity of the valve. No rela- 
tionship between these variations and the size of the valves can be 
traced. 
The surface of both valves is covered with simple radiating pli- 
cations, varying from eleven to seventeen. The far greater number 
of valves, more than 80%, show from thirteen to fifteen plications. 
Shells having seventeen plications are almost invariably those in 
