SPONGES. 205 



skeleton. It is very doubtful, however, whether the genus, as constituted by 

 Lendenfeld, is separable from the older Aptysilla. The type species of Dendrilla 

 appears to be a form with a dendritic skeleton, viz., D. rosea, and if we are to 

 separate the species with reticulate skeleton, we can, for etymological reasons, hardly 

 employ the name Dendrilla for them. 



In my report on Mr. Thurston's Second Collection of Sponges from the Gulf of 

 Manaar (3) I attempted to avoid the necessity for erecting a new genus by employing 

 Bowerbank's old name Spongionella for what must now be regarded as the type of 

 the genus Megalopastas, viz., Megalopastas nigra. I must admit, however, that 

 Bowerbank's type of the genus Spongionella, viz., S. pulehella, is probably not an 

 Aplysillid at all (according to Lendenfeld it is a "Leiosella"), while, on the other 

 hand, Bowerbank (8) also applied the name Spongionella to another totally different 

 sponge, viz., Phgllospongia {Spongionella) holdsworthi, and Spongionella is retained 

 by Lendenfeld (66) as a sub-genus of Phyllospongia. I therefore now revert to 

 what was my original intention in 1889, and propose the new genus Megalopastas, 

 the name being chosen in allusion to the large size of the flagellate chambers. 



The anatomical characters of the genus are shown in Plate XV., fig. 4, which 

 represents, somewhat diagrammatically, a vertical section of an Australian species, 

 M. elegans (Lendenfeld's Dendrilla. elegans), of which I happen to have much better 

 preserved material than of the Ceylon species. Excepting that in M. elegans the 

 ectosome is thicker and the outermost secondary fibres of the skeleton do not lie so 

 near the surface, so that there is no " dermal skeleton," the figure would serve almost 

 equally well for either of the Ceylon species. My preparations of M. elegans contain 

 numerous embryos, enclosed in spherical endothelial cajisules, and mostly in the stage 

 represented in the figure, consisting of a solid inner mass of cells surrounded by an 

 outer layer modified near one end to form a pigment ring. This embryo agrees 

 pretty closely with those described by Schulze in Spongelia (71) and Euspo}igia (72), 

 and its occurrence perhaps tends to show the correctness of my views as to the close 

 relationship of the so-called " Hexaceratina " to the other Euceratosa. 



Megalopastas nigra (Dendy). Plate XIV., fig. 7 ; Plate XV, figs. 5-8. 

 1889, Spongionella nigra, Dendy (3). 

 This very remarkable species was originally discovered by Mr. Thurston and is 

 represented in Professor Herdman's collection by two specimens. The sponge 

 (Plate XIV., fig. 7) is sessile, and consists of a number of vertical lamellae, branching 

 and anastomosing with one another, often in a very complex manner. The largest 

 specimen I have seen was about 250 millims. high and the same in breadth, with 

 lamellae about 5 millims. thick. The colour of the living sponge is black (Thurston), 

 when dry, dull black, and in spirit rather lighter, blackish-grey. Texture (in spirit) 

 very compressible and resilient ; moderately tough. Surface granulated, the granules 

 being really minute, close-set conuli. Vents abundantly scattered, but almost or 



