44 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
As to the exact method of attachment, it should be noted that the 
ends of the primary lamellae when entirely preserved, are always 
abruptly bent inwards, very much as the edge of a knife blade may be 
turned by contact with a harder substance, forming a "wire edge." 
This bending or truncating, for it is not a true bend, is always slight 
and might be easily overlooked (pi. 6, fig. 4). It does not resemble a 
break. A close examination of the crura shows the inner edges of the 
blades slightly roughened at the tips but so slightly that, if the surface 
were ever broken, it has been almost completely healed by subsequent 
mineralization, so small in amount as not to destroy the delicate 
grooves on the surface. It appears most likely that the junction was 
very much as shown in figure 3 (pi. 6). The outer edge of the lamella 
rested in the groove on the surface of the crus and the tip was sharply 
bent and attached to the inner edge of the tip of the crus. Not one 
of the lamellae seen, however, shows the end bent as much as this 
explanation requires and the crura do not show any trace of the 
attachment beyond the very slight roughening. When one recalls the 
relative position of the crus and the lamella during the life of the in- 
dividual, it is seen that such an explanation accounts, prol:)ably entirely, 
for the fragile nature of the connection. The weight of the brachi- 
dium is borne by the median septum and the crura and not by the 
attachment. The attachment was either sufficient only to prevent 
the lamella from slipping off the crus or was so imperfectly spiculized 
that little trace of it now remains. 
From the crura the primary lamellae converge to near the center of 
the brachial valve beyond which they separate toward the antei'ior 
margin and converge again near the center of the pedicle valve, as in 
Merista and Meristella. Each spiral ribbon makes from eight to 
ten, sometimes possibly twelve, volutions in the adult. The cones are 
somewhat flattened in accordance with the shape of the shell and their 
bases are approximately parallel to the longitudinal axis of the shell. 
The anterior curves of the ribbon are fimbriate on the outer margin. 
Not rarely these fimbriae are found over the entire anterior half of 
the whorl and sometimes entirely around it. So far as observed they 
are always located on the side of the ribbon toward the base of the cone 
(see figures). 
The jugum or loop is modeled after the same plan as that of Merista 
and jNIeristella but is more massive and compact. It is situated on the 
primary lamellae about one third the length of their longer axis from 
