94 THE VOYAGE OF H.M.S. CHALLENGER. 



!)}• the blood- vascular plexus (PL LYII. figs. 1, 4, Ir), which has a water-vascular trunk 

 ou each side of it ; and the water-tubes extend outwards in a radial direction as long as 

 the Avatcr-vessels remain double. This is very evident in some horizontal sections 

 through the tapper part of the disk of Pentacrinus naresianus, which were made for 

 Sir Wyvdle Thomson by Dr. Stirling. The double row of water-tubes may be seen 

 extending along the sides of the ambulacra beneath the line of tentacles, to a distance of 

 3 "5 mm. from the peristome, and then it is not complete. 



The position of the water-tubes beneath the primary ambulacra is well shown in 

 PI. LVII. figs. 1, 3, 4. They are seen, as it were, coiling round the subambulacral 

 plates to open below into the upper part of the circumvisceral coelom. When, however, 

 the radial extensions of the labial plexus gradually thin out and the two lateral water- 

 vessels unite into a single median trunk, the water-tubes become less numerous, and are 

 only found in the first two or three sections beyond the point of union. They are thus 

 really in the same position as in the Comatulse, if we consider the double water- vessels as 

 expressing extensions of the angles of the water-vascular ring in the direction of the 

 rays. 



The water-pores which pierce the ventral perisome, whether it be plated or not, are in 

 a close functional relation with the water-tubes. They are the openings of minute canals 

 which are lined by columnar epithelium, and expand almost immediately into enlarge- 

 ments where the epithelium is ciliated (PI. LVII. figs. 1, 3, 4 ; PI. LIX. figs. 2, 4, 6 — zf^j). 

 The inner end of the canal beyond the enlargement is lined by pavement epithelium, and 

 opens into the body-cavity. According to Perrier ^ the primary water-tubes of the early 

 larval Comatute are directly continuous with the inner ends of the water-pores, without 

 the intervention of the body-cavity. 



He has not yet figured this connection, however ; and after the careful observations of 

 Ludwig upon the subject, with which my own are in complete accordance, I have 

 considerable hesitation in accepting Perrier's statements, especially as he admits that the 

 water-pores of mature Comatulte do establish communication between the body-ca\dty and 

 the exterior, just as described by Ludwig in the Cystid phase. This subject is discussed 

 more fully elsewhere.^ [See Appendix, Note D.] 



In the smaller and simpler types of Crinoids the water- pores correspond in number to 

 the water-tubes. The young Antedon has one in each interradius ; and the same is the 

 case in Rhizocrhms, the single water-pore piercing the oral plate. In Hyocriniis, 

 however, the number of pores is larger. In both the sjiecimens which I have examined 

 there are two pores in the oral jjlate of the anal interradius, and there are no others in 

 any of the anambulacral plates which lie between it and the edges of the radials. In the 

 other interradii the disposition of the water-j)ores is as follows : — 



1 Sur le cUveloppement des Coraatules, Comptes reiidns, t. xcviii., 1884, pp. 444-446. 



2 On some points in the Anatomy of Larval ComatiiI;p, Quart. Journ. Micr. Sci., N. S., vol. xxiv., 1884, p. 320. 



