1876.] and Development o/Antedon rosaceus. 215 



of the visceral mass, and the formation of membranous adhesions (as at m, 

 fig. 12) between the apposed surfaces of its peritoneal covering and of the 

 peritoneal lining of the calyx, separate into an upper and a lower portion, 

 whose communication with each other is maintained by the axial canal. 



The oral perisome is perforated in the spaces between the radial fur- 

 rows (especially near the angles formed by their junction in the mouth) 

 by numerous minute pores surrounded by ciliated epithelium, whose cilia 

 I have seen in extremely active movement, apparently inwards. These 

 are the orifices of funnel-shaped canals, which first pass vertically through 

 the perisome, and then contract and take an oblique direction, opening 

 (I believe) into a system of sinuses, which are left by the incomplete 

 adhesion of the peritoneum covering the upper surface of the visceral 

 mass to that which lines the under surface of the oral perisome. If 

 this be the case, they obviously establish a direct connexion between this 

 limited portion of the perivisceral cavity and the medium inhabited by 

 the animal ; and the water thus introduced probably serves a respiratory 

 purpose*. 



Perivisceral Cavity. — The adhesion of the under surface of the oral 

 perisome to the upper surface of the visceral mass is so close that they 

 can only be detached by tearing, after both have been hardened in spirit. 

 The peritoneal covering of the lower or dorsal surface of the visceral mass, 

 on the other hand, is but slightly adherent to the opposite peritoneal sur- 

 face lining the deeper part of the concavity of the calyx, although usually 

 in close contact with it ; and thus, when the periphery of the oral disk 

 has been cut through, the visceral mass adherent to its lower surface may 

 be lifted entire out of the calyx, without rupture of any thing but a 

 pedicle (h, figs. 2, 3) which passes up as a continuation of the original 

 Crinoidal axis, through the " rosette " (Part I. § 35) formed by the meta- 

 morphosis of the original Basals. Hence it often happens that, through 

 the tearing of their oral perisome, which is very soft in the living state, 

 specimens of Antedon come up in the dredge completely eviscerated ; and 

 I have strong evidence that even this loss may be repaired (as the like 

 evisceration is said to be in Holothurida) by the production of a new 

 visceral mass from the " axial prolongation." 



Although the visceral mass so nearly fills the cavity of the calyx that 

 scarcely any space exists between their apposed peritoneal surfaces, yet 

 the existence of a true peritoneal cavity becomes obvious at the bases of 

 the arms ; for it is there prolonged, in each arm, into an extension (cc, 

 fig. 7) which forms a canal that underlies the subtentacular canal (stc) 

 proceeding into it from the oral disk ; and, like that canal, it not only- 

 passes along the whole length of each arm, but also gives off: lateral exten- 

 sions (dd) that proceed to the extremities of the pinnules. This canal, 



* These canals have been noticed by Grimm (Bulletins de l'Acad. Imper. de St. 

 Petersb. 1872, pp. 3-8) ; but he has not been able to trace their internal connexions. 

 By Perrier (loc. cit.) they are described as caeca. 



