56 THE CRINOIDEA CAMERATA OF NORTH AMERICA. 
basals to a less number, is readily understood among genera in which the 
anal plate is wanting. When the base is quadripartite, it is invariably the 
two anterior plates of the elementary five which are consolidated (Fig. 2). 
In the tripartite base there is a fusion of the posterior with the left postero- 
lateral basal, and another between the right posterior and adjoining antero- 
lateral-plate (Fig. 3). The figure shows that a bisection of the two larger 
plates will reproduce the original five pieces, interradially disposed. 
The case is not so simple in genera with an anal plate, where the form 
of the basal disk is changed from pentagonal to hexagonal (Fig. 4), as a 
bisection of the larger plates would produce six plates instead of five. This 
difficulty, however, is overcome if we consider that the introduction of the 
anal plate into the ring of radials necessitated corresponding modifications 
among the basals, as otherwise these plates would lose their interradial posi- 
tion. It required either the introduction of a basi-anal plate, or an increase 
in the size of the original pieces. That the latter occurred among the Came- 
rata is clearly shown by the diagrams, and the evidence leaves no doubt at 
what part of the base the extra width was inserted. 
Taking first the quadripartite base, and comparing Fig. 2 of the diagrams 
with Fig. 8 — one pentangular and the other hexangular — we find that in 
the latter the posterior basal has doubled in size (Fig. 7), without materi- 
ally changing the orientation of the plates, or disturbing their general 
arrangement. 
In the tripartite base the change was accomplished in a different way. 
There a is added to plate ¢ (Figs. 9 and 10), and the plates a and ed have 
coalesced, and hold relatively the same position as in Fig. 5. 
The bipartite base is probably derived from the tripartite (Fig. 4), which 
preceded it in time, and x, which in the latter constituted a part of ¢, is 
united with ed, and a) with ec (Figs. 11 and 12). 
Now, taking up Fig. 7, and eliminating z, so that the side of plate a rests 
against the plate e, we obtain Fig. 2, and by a similar procedure we are 
enabled to transform Fig. 9 into Fig. 3. The hexagonal base is thus re- 
stored to its primitive pentagonal form without disturbing the orientation 
of any plate, compound or simple. 
A most beautiful confirmation of these observations is furnished by an 
abnormal specimen of Teleiocrinus umbrosus in our collection, in which the 
regular anal plate is wanting.  Zeleioerinus has normally three equal basals, 
but in this specimen the basal plate to the left of the anterior ray is reduced 
