DENDY NON-CALCAREOUS SPONGES 103 



There are five specimens of this sponge in the collection. They are all subcylin- 

 drical, provided with a single vent at the upper extremity and a root-tuft of long, 

 silky spicules at the lower (cf. Figs. lOa-lOc). The general surface is smooth and porous, 

 the consistence soft and compressible, the colour in alcohol light grey. The smallest 

 specimen measures about 19 mm. in height by 5-5 mm. in diameter ; the largest 38 mm. 

 by 13 mm., in both cases excluding the root-tuft. In one specimen the root-tuft extends 

 downwards for 18 mm. before meeting the mass of sand-grains with which it is still 

 in connection. In the other specimens a similar mass of sand -grains is attached by the 

 root-tuft close to the lower extremity of the sponge. 



In one of the larger specimens, which I cut open, the vent forms the terminal 

 aperture of a cylindrical cloacal chamber about 10 mm. in length and 2-5 mm. in diameter, 

 into which numerous larger and smaller exhalant canals open at various levels, the 

 larger ones being continuations of the cloacal cavity deep down into the body of the 

 sponge (Fig. lOa). Just below the point where the larger canals, coming from below, 

 join to form the cloaca! cavity, lies the so-called " nucleus," from which the principal 

 fibres of the skeleton radiate, mostly in a downward direction. Curving gently outwards, 

 these fibres break up, over the general surface, into dense surface-brushes, which, 

 however, do not project sufficiently to render the surface hispid to the naked eye. At 

 the lower extremity of the sponge they are continued downwards, outside the sponge- 

 body, to form the root-tuft. 



Throughout the greater part of the sponge-body the fibres appear to be composed 

 exclusively of very long and very slender oxea, and they are crossed at various angles 

 by irregularly scattered oxea of similar form, but perhaps shorter. 



The dense surface-brushes are composed mainly of long, slender oxea, but mingled 

 with these occur many protrieenes with very slender shaft and almost hair-like cladi 

 of unequal length. 



The anatria3nes appear to be confined to the lower parts of the sponge, where they 

 seem to form the principal constituents of the descending fibres both inside the sponge 

 and in the root-tuft. Their shafts are very long and slender, hair-like, and their cladomes 

 unusually well developed, with sharp, strongly recurved cladi. Protrisenes also occur 

 in the root-tuft, but less abundantly than the anatria3nes. The microscleres are minute, 

 slender, contort sigmata, of the ordinary Tetilla type, not very abundant. 



This interesting and easily recognisable sponge appears to be characteristic of 

 sandy and muddy flats along the shores of the Indian Ocean. Sollas has expressed 

 a doubt whether the specimen from the Mergui Archipelago identified by Carter as 

 belonging to this species is really specifically identical with the type from the S.E. 

 coast of Arabia, but I do not think it at all likely that there is a specific difference. 

 On the other hand, I have myself [1905] described a distinct, but closely-related species 

 (T. limicola) from Ceylon, differing from T. dactyloidea in external form and in the 

 arrangement of the exhalant canal-system. Annandale has recently [1915] described 



