182 P, HERBERT CARPENTER, 
varies considerably in adult Comatule, reaching thirty or 
more in each interradius of Antedon rosacea ; but in Rhizo- 
erinus and in the young Antedon there is but one in each 
interradius (20), while in the early stages of the Pentacrinoid 
larva of Antedon there is only one, which is always situated 
in the same interradius as the foregut (26). Corresponding 
to these water-tubes (“ Steinkaniale,” Ludwig), the ventral 
perisome, whether soft or plated, is pierced by a variable 
number of water-pores, which lead from the exterior directly 
into the body-cavity (figs. 14, 15, w. p.). They are the 
external openings of small canals lined by columnar epithe- 
lium, and expanding almost immediately into enlargements 
in which the epithelium is ciliated. The inner end of the 
canal beyond the enlargement is lined by pavement epithe- 
lium, and opens into the body-cavity. The number of these 
water-pores is very large, but very variable, and they are 
not limited to the disc, for I have found them on the lower 
parts of the arms (fig. 15), and even on the proximal pin- 
nules. In both these cases they open into that section of 
the body-cavity which surrounds the generative apparatus, 
and is known as the genital canal. Ludwig has estimated 
their number in the adult Antedon rosacea at about 1500, 
while in the young Antedon and in Rhizocrinus there is but 
one in each interradius (20), and in the Pentacrinoid larva 
there is only one which pierces the lateral margin of the 
oral plate that is in the same interradius as the single 
water-tube (26). 
By means of the water-pores, body-cavity, and water- 
tubes, therefore, the ambulacral system of a Crinoid is 
placed in communication with the external water; and 
Ludwig regards the water-pore and water-tube together as 
conjointly representing the madreporic apparatus and stone- 
canal of the Starfishes and Ophiurids. A multiplication of 
these organs may occur in both groups, which is merely 
carried somewhat further in the Crinoids, and the nature of 
the epithelial lining of these afferent channels is the same in 
all three groups. It remains, however, yet to be proved 
that the disconnected water-pore and water-tube of the 
Antedon larva are morphologically equivalent to the stone- 
canal and madreporic apparatus of the other Echinoderms, 
though it is certainly exceedingly probable that they are so. 
At present we do not know enough of the earlier stages of 
their development to be quite sure of their homology with 
their undoubted analogues in the other Echinoderms. 
Between the dorsal skeleton of the arms and pinnules and 
the water-vessels on their ventral side, are three tubular 
