226 Mr. F. A. Bather on British Fossil Crinoids : 
are postpalm ars of at least the sixth series; in other words 
the dichotomy is seen to take place at least 8 times, so that 
the final branches of the arms can have numbered no less than 
1280, and were probably nearer 2000. The ossicles are of 
peculiar shape ; even in the more proximal series their sides 
are seen to curve round in a curious manner towards the 
ventral surface, as is w r ell shown by specimen 57478 h ; and, 
as dichotomy progresses and the transverse axis of the ossicles 
shortens, the dorso-ventral axis becomes much longer, so that 
in the first postpalmars the ratio of depth to width is as 5 to 
3, and in the third postpalmars as 7 to 3 ; this is clearly seen 
in specimen 144 Mason College (Pl. X. fig. 1). In most 
specimens the backs or outer portions of the arms present a 
continuously smooth appearance, but in others the edges of 
the ossicles are more rounded, inducing a slightly moniliform 
aspect ; the smoothness may therefore be due to attrition. Be 
this as it may, there can be little doubt but that the arm- 
ossicles are more ridged at the sides of the arms than on their 
backs, and this in such a manner that the greatest transverse 
diameter of each ossicle is towards its upper or distal end, and 
towards its ventral surface. The condition of the specimens 
does not permit the direct demonstration of a dorsal canal ; 
but that such existed seems certain not only from the shape of 
the ossicles, but also from the fact that in much weathered 
specimens a groove is formed in the median line on the dorsal 
surface ; this is best shown by specimens 138 Mason College 
and 431 Holcroft (Diagram 8). The ventral surface of the 
arms is partially exposed in specimen 57478 b : the covering- 
plates are no longer in situ , and the food-groove, which is 
rather shallow in proportion to the depth of the ossicles, is 
clearly seen ; on either side of it the ventral edges of the 
ossicles rise up like little rounded teeth (Diagram 9). 
Costals, 3 to each ray ; in two instances out of the twenty- 
seven counted there appear to be 4 ; all of the same width, 
which is always more than two thirds that of the radial ; height 
from 1 to 2 millim. according to size of specimen. The upper 
and low^er edges of the second costal are straight and parallel ; 
the lower edge of the first costal is curved conformably with 
the articular facet of the radial ; the lines containing the 
axillary angle of the third costal are concave. The upper 
and lower edges of the costals are slightly bevelled on the 
outside, indicating that these ossicles were united by loose 
suture. 
Distichals vary in number in the different specimens, and, 
to a less extent, in different branches of the same specimen ; 
19 branches have been observed with 4 distichals apiece, 11 
