SPONGES. 129 



The walls of the excavations have a finely granulated or, under the microscope, 

 frothy appearance, due to the presence of innumerable shallow, conchoidal depressions, 

 which are in contact with one another all over the surfaces of the walls. 



There is only room in the thickness of the shell for a single layer of chambers. 

 When these have attained their full size, their transverse diameter is about 1"5 millims., 

 but they may be considerably elongated. The diameter of the cylindrical canals 

 which terminate on the outer surface of the shell varies up to about - 65 millim. 

 The outer end of each is closed, completely or partially, hy a thin membrane containing 

 numerous micro- and megascleres. Where the membrane in question stretches com- 

 pletely across the end of the canal, I assume it to be an inhalant pore-sieve, though 

 the pores cannot now be seen ; such a membrane may be supported by bundles of 

 tylostyles converging towards the centre. The vents, on the other hand, are more 

 or less widely open and surrounded by a thin, membranous diaphragm, also supported 

 by tylostyles. 



The body of the sponge itself (in spirit) forms for the most part a very thin lining 

 (of a pale yellowish-brown colour) to the chambers. 



Spictdes. (1.) Tylostyli (Plate V., fig. 9, a) ; straight, or nearly so, slender, very 

 gradually and sharply pointed, and with well-developed globular heads ; size about 

 0"25 millim. by - 004 millim., with head about 0'006 millim. in diameter. These 

 spicules occur scattered generally, as well as in the neighbourhood of the vents and 

 pores. 



(2.) Spirasters (Plate V., fig. 9, b, c, d) ; usually with about four angulations, fairly 

 stout, and abundantly but rather minutely spinous, size about 0"024 millim. by 

 0"004 millim., but variable. These spicules pass by transitional forms (Plate V., 



fig- 9 > e > /> u) mto 



(3.) Spined microxea (Plate V., fig. 9, h, 1-) ; usually bent or angulated in the middle, 



and measuring about 0"06 millim. by - 0027 millim., but variable. These forms are 



extremely numerous in the deeper parts of the sponge. The gradual transition 



between spirasters and microxea in this species is extremely interesting. 



The excavations made by Cliona margaritiferce, even down to the spine-shaped 

 outgrowths of the chambers, are closely similar to those of Hancock's Cliona 

 spinosa (38) and Thoosa cactoides (38), but the spiculation is very different. Cliona 

 spinosa occurs in shells of Perna and Placuna, and Thoosa cactoides in shells of 

 Meleagrina margaritifera. Another distinct species, of closely similar form, occurs 

 in shells of Meleagrina albino,? and has been described by the same author (38) 

 under the name Cliona cervina ; the spiculation of this species makes a much closer 

 approach to that of C. margaritiferce, but the two appear to be quite distinct. 



Topsext (37) has already described, under the name Cliona indica, a boring sponge 

 infesting a pearl oyster from Ceylon. This species is evidently nearly related to ours, 

 but the describer gives no information as to the character of the excavations or 

 apertures, and the spiculation differs so much from that of our species that it is 



s 



