CO PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
tion it seems difficult to harmonize the widely different types of spondy- 
lium in the pedicle yalye. The "shoe-lifter" of M crista typa (pi. 10, 
fig. 57-59), however, suggests a method by which the spondylium of 
Camarophorella might haye been eyolved. But first it is necessary 
to describe a minor feature of the spondylium of Camarophorella 
which has been referred to briefly. In the detailed description of C. 
mutabilis, reference was made to two accessory plates in the beak of 
the pedicle valve, one on each side of the spondylium, connecting the 
outer surface of the spondylium with the inner surface of the valve. 
In the molds they are seen as clefts on either side of the pedicle open- 
ing and anterior to it. In young shells the plate is well developed 
and is located about halfway between the hinge and the ventral sur- 
face of the mold, so placed that they are not visible in a dorsal view 
of the shell. With growth, this plate is gradually pushed around close 
to the hinge line and in some cases is resorbed to a considerable extent, 
so that in the adult it is readily visible from the dorsal surface and in 
some cases invisible from the side, quite the contrary of the younger 
stages. Not infrequently it becoines so reduced that the filling of the 
cavity between it and the dorsal edge of the spondylium is reduced 
to the merest spine (pi. 6, fig. 5-12). Apparently the plates are func- 
tionless. They have no value in the majority of cases as a strengthen- 
ing brace in the adult, and there is no apparent need for such a brace 
in the young. 
The "shoe-lifter" of M crista typa from the Helderbergian of Mary- 
land is extremely variable. In figure 58 (pi. 10), is shown a specimen 
which is believed to represent, in a general way, one of the stages in 
evolution which has resulted in the spondylium of Camarophorella. 
The broad curved plate of the structure is reduced and the dental 
lamellae occupy much the same position as in Camaro])horella. If 
the reduction of the arched plate were carried further and the dental 
lamellae increased and continued to the l)ottom of the valve as a me- 
dian septum, so that they took up the entire muscle strain, a structure 
very similar to that seen in the young of Camarophorella would result. 
The accessory plates shown in figures 42 and 43, h (pi. 8), are better 
developed than usual and show very well what a late stage in such a 
line of develo])ment would be like. This explains the apparently 
useless acce\ssorv plates of Camarophorella as a vestigial remnant of 
the broad arched plate of ^Nlerista and Dicamara, and makes possible 
the suggestion that Camarophorella was developed from a Dicamara- 
