rOTERIOCRINin.E EXTRACRINUS. 99 



Miller, who does not appear to have been aware of the jointed structure of 

 these pieces, imagined they served as buttresses to strengthen the column at its summit 

 where it was most muscular and yielding, and therefore required some such support 

 to enable it to sustain the great weight of the body and rays with their thousands of 

 lentacnla. This was in all probability one of the ends designed by this peculiar 

 contrivance, but had these pieces been rigid as Miller and other naturalists supposed 

 they were, they would have imparted strength and stability to the column near its 

 attachment to the body, while at the same time they would have destroyed in some 

 degree its flexibility. This defect was completely obviated by the beautiful and unique 

 contrivance of making these pieces jointed and flexible, so that they could 3'ield to the 

 lateral movement of the column on either side and at the same time act as moveable 

 abutments, adjusting themselves to any required position, and effecting their object 

 far more perfectly than if it were possible for them to have been of an unyielding nature, 

 without interfering with the flexibility of the column. In PI. 12, Figures c, and d, 

 will be found a correct representation of the manner in which these joints articulate to 

 each other, their pro[ier adhesion being secured by marginal striae arranged around an 

 oval surface of a'ttachment, with intervening muscular fibres. 'I"he specimens represent- 

 ed in PL 12, fig. 1, a, were selected for the purpose of illustrating the' structure of 

 this portion of the animal, and as this part is frequently obscured by the overlj'ing 

 auxiliary side arms, we consider it a fortunate circumstance that our figures will serve to 

 elucidate a fact of some importance to the zoologist as well as to the paljeontologist. 



It is probable that in the earlier stages of growth and before the different parts 

 had arrived at maturity these pieces were not jointed as in the adult animal. 



The Rays. — A single ray of two joints emanates from each of the five radial or larger 

 perisomic pieces, the ravs then branch off into a pair. At about the seventh joint, they 

 again subdivide and form two leading branches, from the inner lateral edges of which 

 many minor rays proceed, but the distance between each of these is not regular, som- 

 times as many as fifteen joints intervene, vvliiie in others only ten occur. The cuneiform 

 joints from whence the minor rays proreed are not equally divided into two sloping 

 surfaces of articulation as in the lower bifurcations, but the inner slope which supports 

 the lesser ray is smaller than the outer one, which scarcely differs from the ordinary 

 articulating surface of the leading branches. All the ray joints are more or less wedge 

 shaped, the thickest part invariably occupying a lateral position, but so placed that a 

 thick and thin part alternate with each other. On the widest side of each joint, but 

 inclining inwards, there is a small rounded projection perforated in its centre for the 

 attachment of a teritaculum, as represented in PI. 12, fig. g. The cuneiform joints at the 



