FEATHER-STARS, RECENT AND FOSSIL. 
195 
Actinometra (PI. VI. fig. 11), the anal tube is nearly or quite 
central. 
In our English Comatulce the disc is almost or entirely bare ; 
hut in many tropical forms it is covered by a very complete 
mosaic pavement of closely fitting plates (PL VI. fig. 12), which 
occupy all the spaces between the food- grooves, and even extend 
out for some distance on to the upper surface of the arms and 
pinnules at the sides of their median grooves.* Most of 
the plates nearest the grooves are pierced by minute holes 
( w.p .), the ‘ water-pores.’ These are the upper openings of tiny 
funnel-shaped tubes, which open below into the body cavity, 
and are lined by cilia all working inwards, so that the body 
cavity is in free communication with the external water. These 
pores are likewise present in Comatulce with naked discs (PI. V. 
fig. 1. w.p.) ; and they have also been found on the lowest 
parts of the arms, and even piercing the plating of the pinnules, 
where they lead into a tubular extension of the body cavity into 
the arm, containing the generative gland. 
On the naked discs (PI. VI. fig. 11) each side of the food- 
groove is formed of an elevated fold of skin scolloped at its edge 
so as to form a row of minute triangular leaflets. At the base 
of each of these is a group of three delicate tubular tentacles, 
one of which is much longer than the other two. The leaflets 
alternate on opposite sides of the groove, and are ordinarily 
erected, with the tentacles projecting considerably beyond them. 
But the tentacles may be withdrawn and the two folds closed 
down, the leaflets meeting in a sinuous line, so as completely to 
cover the furrow. The same is the case with the grooves of 
the arms and of the pinnules borne by them, as is seen in 
PI. VI. fig. 1, while some arms are occasionally altogether un- 
grooved. (PI. VI. fig. 2.) 
In the plated tropical Comatulce matters are much more com- 
plicated (PI. VI. fig. 10.) The edges of the groove are supported 
by a series of more or less oblong side-plates ( adambulacral ), 
and hinged to the upper side of each of these is an oval covering- 
platef (super ambulacra!). These may be erected and the ten- 
tacles extended, or they may be closed down so as completely 
* The ventral plating of the disc is so very complete in some recent 
Comatulce , that no mouth is visible at all in the dry state. The summit- 
plates of many Palaeozoic Crinoids, such as Cyathocrinus, and also of the 
Blastoids appear to me to be strictly comparable to those of these recent forms. 
Hambach, however, speaks of the summit-plates of the Blastoids as non- 
existent or rather as proving on close examination, to be Bryozoa, or ovulum- 
like bodies ! — Trans. St. Louis Acad. Science, vol. iv. No. 1, p. 150. 
t Some recent Crinoids have covering-plates only, and no side-plates. 
The same variability occurs in the Palaeozoic Crinoids, and perhaps also in 
Blastoids. Some species have both series represented, though, of course, in 
a greatly reduced form, while others have the covering-plates only. Ham- 
bach’s description of the zigzag-plated integument of the Blastoids, as 
‘ probably of an elastic texture during the lifetime of the animal,’ professedly 
