138 The American Geologist. September, is93 
it is best defined there, while the elevation of the ridge be- 
comes in many eases less defined towards either end. At the 
junction of the horns of these crescents (as ends of these 
ridges may be called) with the succeeding ridges, they are 
sometimes indistinct, especially exteriorly. The scar areas are 
too poorly preserved to show the surface marking. With this 
exception, the above description also gives the characteristics 
of the scars of the Huffman specimen. 
The average width of the scars is N.5 to 9.5, sometimes 10 
mm.; their average length is 11.5 to 12.5 mm. from ridge to 
ridge. The length of the scars is therefore relatively greater 
than in the Huffman specimens, where the length and width 
are more nearly equal. The great length of the scar, com- 
bined with the great curvature of the defining ridges, makes 
it probable that the Eaton Glyptodendron, if congeneric, rep- 
resents a fragment of the dorsal side of these shells. The 
scars are arranged in diagonally intersecting series. The 
Huffman specimen, as a rule, does not show the same 
faultless regularity in the arrangement of the scars, except 
indeed on the dorsal surface, as far as the preservation of the 
scars there extends in our specimen; this being another rea- 
son why we suppose the Eaton specimen to represent part of 
the dorsal side. The curvature of the Eaton fragment, as 
shown by the cast, is that of an arc having a radius of 65 mm. 
If the fragment be part of the dorsal side, it must therefore 
belong to a point of the shell having considerably greater di- 
ameters than the Huffman specimen, a fact already indicated 
by the larger scars of the Eaton specimen. 
Finally, the condition of the cast itself seems to suggest its 
connection with that of a shell rather than with that of a 
piece of wood, since the upper boundary of the fragment is 
that of a regular and even though curved line, such as is not 
very infrequent in the case of the fracture of a shell, but one 
that would hardly occur in a branch or trunk of a tree of 
equal diameter broken transversely. It would be more likely 
to splinter, and to fracture irregularly. Of course there will 
always be an element of uncertainty about Glyptodendron, 
owing to tin- fact that it shows neither the interior structure, 
the general form of the fossil, nor the detailed surface charac- 
ters of the scars, the only clear features being the boundaries 
