200 CEYLON PEARL OYSTER REPORT. 



processes we find a very dense, stout axial core of spicules more or less cemented 

 together by spongin, from which numerous loose bundles of spicules radiate obliquely 

 outwards and upwards into the surface conuli, beyond which their apices may project. 

 Here, again, there is no dermal skeleton. 



Spicules. Oxea (Plate XIII., fig. 1), of various shapes and sizes according to 

 situation; (.) in the columns of the basal skeleton, stout, sub-fusiform, slightly 

 curved, irregularly ended, variable in size, say about 0'5G millim. by 0'03 millim. ; 

 (b.) in the ectosome of the base, comparatively long and slender, slightly curved, 

 irregularly ended, measuring, say, about 0'88 millim. by 0'02 millim. ; (c.) in the 

 digitiform processes the spicules are mostly of the long and slender type just 

 described, but occasionally stout and comparatively short forms occur. 



One of the most characteristic features of this sponge is the thick gelatinous 

 ectosome (Plate XIII., fig. 2, ect.), composed of collenchyma, with a hyaline, or some- 

 times finely granular, matrix containing an immense number of large, stellate, 

 connective-tissue cells. In the basal crust this ectosome is about 0'G5 millim. thick. 

 In the digitiform processes it forms an even thicker layer around the dense central 

 axis of spicular fibre (Plate XIII., fig. 2), but it is penetrated by large longitudinal 

 canals (I.e.), about half a dozen in number, and varying in diameter up to about 

 2 '2 5 millims. These canals are separated from one another by radially arranged 

 longitudinal septa (sept.), in which the bundles of spicules run out from the central 

 axis to the surface conuli. The ectosome on the outer sides of the longitudinal canals 

 is reduced to a membrane of varying thickness, containing no spicules and penetrated 

 by short, narrow canals which place the great longitudinal canals in communication 

 with the exterior. These short canals are arranged in groups at wide intervals, and 

 those of each group unite together into a single larger canal before opening into the 

 longitudinal canal. The dermal pores are doubtless arranged in corresponding 

 groups, probably with one pore at the end of each of the smaller canals, but they are 

 now all closed (their position is shown in Plate XIII., fig. 2, d.p.). The fact that the 

 smaller canals unite together as they pass inwards to the great longitudinal canals 

 seems to indicate that this is an inhalant system. The arrangement of the exhalant 

 system I have not succeeded in making out, and the choanosome is not sufficiently 

 well preserved to enable me to give any details with regard to the flagellate 

 chambers. 



The digitiform processes may contain a good deal of sand outside the spicular axis. 



R.N. 74, 74A (outside pearl banks, Gulf of Manaar). 



OKDER 4 : EUCERATOSA. 



Non-calcareous sponges without siliceous spicules, but with a skeleton consisting of 

 horny fibres developed independently, i.e., not in relation to any pre-existing 

 spicular skeleton. (The skeleton is sometimes replaced or supplemented to a 

 greater or less extent by foreign bodies.) 



