136 CEYLON PEARL OYSTER REPORT. 



The most primitive sub-family of this group is undoubtedly the Gelliinge, from 

 which the Renierinse and Chalininae are clearly derived. The other sub-families may 

 be in part Desmacidonidaa which have lost their chelaa ; this I am able in the present 

 Report to demonstrate pretty clearly in the case of the Phloaodictyinse, which I have 

 accordingly removed. TOPSENT has already removed the Tedaniinse, Desmacellinae, 

 and Hamacanthinae as being Desmacidonidae without chelae ; this may be quite right, 

 but until clear evidence that they are descended from chela-bearing forms is 

 forthcoming, it seems to me equally justifiable to leave them in the parent group. 

 A Desmacella, for example, may very well be an Esperella which has lost its chela), 

 but it seems at least equally probable that it has never reached the stage of 

 having any. 



SUB-FAMILY: GELLIIN^. 



Haploscleridte with diactinal megascleres, and sigmata or toxa or microxea for 

 microscleres. 



Gellius, GRAY. 



Gelliinae with little or no spongin, the main skeleton being formed by a reticulation 

 of oxea. 



Gellius fibulatus (SCHMIDT). 



1862, Reniera flbulata, SCHMIDT (47); 1880, Reniera fibulifera, CARTER (4); 1892, Gellius 

 fibulatus, TOPSENT (48). 



There is in the collection a considerable quantity of this sponge growing amongst 

 the branches of a Floridean Alga. I identify it with the European species by direct 

 comparison with a preparation from a specimen from Budleigh Salterton in 

 Mr. CARTER'S cabinet. The differences in spiculation are very slight. In our specimen 

 the slightly curved and gradually sharp-pointed oxea measure about 0'184 millim. 

 by 0'007 millim., and the sigmata about 0'02 millim. from bend to bend. 



E.N. 51 (Gulf of Manaar) ; 299 ; 348 ; 350 (all growing in association with 

 apparently the same kind of alga ; the last three probably fragments of one and the 

 same specimen). 



Gellius angulatus (BOWERBANK), var. canaliculata, nov. Plate IX., fig. 7. 



[For synonymy and references vide RIDLEY and DENDY (1)]. 



The single specimen is massive, rounded and slightly elongated, about 18 millims. 

 long by 12 millims. in transverse diameter. It was probably attached by one end, and 

 bears a group of vents at the other. The surface is even and smooth, conspicuously 

 veined by ramifying exhalant canals, which run towards the (upper ?) end of the 

 specimen and open there by means of the moderate-sized vents. Many of these 

 exhalant canals lie just beneath the surface and are covered over only by a thin, 

 translucent membrane, which easily gets rubbed off', leaving the canals as open 



