232 REVIEWS 
Essentially different is the ventral structure of the Fistulata, which 
have no interradial plates in the dorsal cup, the anal plate excepted, but 
which have these pieces extensively developed in the tegmen. Four of 
the interambulacral spaces are raised but little above the level of the 
arm bases, while the posterior area is extended abruptly upward, and 
is formed into a tube or sac of variable shape and size, rising beyond the 
tips of the arms. ‘This sac, which may be regarded as a greatly 
extended anal area, probably lodged a large portion of the visceral 
mass. ‘The sac is generally composed of longitudinal rows of hexago- 
nal plates, and is often perforated by pores. The structure at the four 
other sides of the disk is rarely observed except among the Cyathocrinidze 
in which it is probably more substantial than in other groups. In Cyath- 
ocrinus there are six plates, interradially disposed, resting against the 
inflected upper edges of the radials, the lateral margins being covered 
_by the ambulacra. Four of them are large and of equal size, the two 
others, lying at the posterior side, are quite narrow and enclose a 
madreporite. The margins of the larger plates are roofed over in per- 
fect specimens by numerous small irregular pieces, while the perfo- 
rated plate is exposed to view. 
Most of the Ichthyocrinidz have interbrachial plates, which in some 
forms are large and massive, in others small ; some are arranged regu- 
larly, others irregularly, but all are movable. The plates of the tegmen 
are very minute and irregularly arranged, the ambulacra are tegminal, 
and the mouth and food grooves are open. ‘Thus there is among pale- 
ozoic crinoids a tegmen having all the characteristics of the disk in 
recent species, demonstrating conclusively that the disk as a vental 
structure is not confined to the neocrinoids as generally supposed. 
Moreover, acareful study of the various tegmens in the different groups 
shows that there are represented among them all intermediate stages 
from the simplest disk to the most rigid and complicated “vault” of 
the Actinocrinide, and that the so-called vault is a highly modified 
form of the disk. 
The anal plates bear a most important part in the phylogeny of 
paleozoic crinoids, and they are among the best criteria for pur- 
poses of classification. When present they occupy, in the Camerata, 
the median line of the posterior area so as to divide the interbrachial 
plates into two equal sets, and being in rows containing an odd num- 
ber they have the effect, as it were, of breaking up the middle plate into 
two, as in cases where no anal plate is inserted between the sections. 
