98 
lodged in a socket or concave impression on the column ; but when increase 
of size renders a stronger support necessary, two or three succeeding joints of 
the side arms become imbedded in this socket, (for which its extension as 
already noticed, allows room) and these joints instead of being arranged in @ 
series branching off at right angles from the column, become oblique, their 
direction inclining upwards so as to aid in bearing the additional weight. The 
first joint of the side arms where thus obliquely inserted in the columnar 
socket have that portion of their circumference which is presented towards 
the upper part of the column, truncated, in such a curve as may fit 
them to the concavity of the impression where they rest against it. (Pr. 
A. levis, fig. 29. to 33.) The surface of these joints which fit into the co- 
lumnar impression is smooth (Pt. A. levis, fig. 31.34. and 38.) being des- 
tined for adhesion only, but the articulating surface between the contigu- 
ous joints (fig. 32. 33. and 35.) where motion also is to be allowed, exhibits the 
usual mechanism of radiated ridges and furrows. ‘These joints are convex on 
the side nearest the column, and concave on that most remote. The auxiliary 
side arms have sometimes others laterally branching from them, (A. levis, 
fig. 30.) I have, however, never been able to ascertain their termination.* 
The Pexvis (Pt. 111. fig. 1. to 8.) is formed of three plates and saucer- 
shaped, presenting at its upper edge six depressions for the adhesion of as 
many plates of the next series. 
The Six Frast Costans (PL. 1. fig. 3.) adhere to the upper edge of the 
pelvis, are plate-like, and of two distinct forms; five of them are hexagonal, 
and one (the irregular first costal) is pentagonal. 
The Srconp Cosrats. (Pun. fig. 3. to 5.) On the upper edge of each of 
the five hexagonal first costal plates, is placed an hexagonal second costal 
plate. 
* Sinte the mode of insertion of these side arms is such as would facilitate their becoming de- 
tached, and the vessel by which they receive nourishment from the alimentary canal very small, 
the idea once suggested itself to my mind, that these complicated side arms might possibly 
be the rudiments of young animals, which after drawing their nourishment for a time only 
from the parent, drop off and become independent. In our present state of ignorance concern- 
ing the propagation of these animals, it seems improper to suppress the statement of this possi- 
bility, which would at once be confirmed or confuted by the discovery of the termination of these 
side arms; since, if the conjectnre is correct, they must present the rudiments of the pelvis, &c. 
On the whole, however, I am rather inclined to reject than adopt this idea, 
