SPONGES. 61 



Hexadella, Topsent. 



Myxospongida with large, sac-shaped flagellate chambers and a distinctly differen- 

 tiated, tough ectosome. 



The description given by Topsent (loc. cit., p. 119) of this genus and of the two 

 species which he includes in it is evidently of a preliminary character and is unfor- 

 tunately without illustrations. His own diagnosis of the genus runs, " Hexaceratina 

 revetantes, molles, sans spicules ni fibres cornes ; un peu plus epais que dans les 

 genres voisins, l'ectosome jouit seul de quelque consistance et sert, dans une certaine 

 mesure, de squelette externe a la masse." 



As the genus is as yet so little known, I shall postpone the discussion of its 

 relationship until I have described in some detail the anatomy of the Ceylon species. 



Hexadella indica, n. sp. Plate I., figs. 1-3. 



Sponge thin, encrusting ; spreading extensively over masses of calcareous Polyzoa 

 and other organisms, but usually only about 1 millim. in thickness. Colour in spirit, 

 grey (in life, red ?). Surface smooth, glabrous, minutely reticulate. Vents few, 

 minute, scattered. Consistence (in spirit) rather tough and membranous (owing 

 chiefly to the ectosome), so that the sponge can be stripped off from the substratum 

 like a skin. 



The ectosome (Plate I., fig. 1, Eel.) forms a tough pellicle about 0"08 millim. thick, 

 which can be stripped oft* from the underlying choanosome. The choanosome evidently 

 consists of a thin lamella (fig. 1, Ch.L., fig. 2), very much folded upon itself. The 

 lamella itself is little, if any, thicker than the ectosome, and contains a single layer of 

 large, sac-shaped flagellate chambers (figs. 1,2, F.C.), whose arrangement, of course, 

 follows the curvature of the lamella. The structure of the choanosome thus reminds 

 one very strongly of what occurs in OscareUa, but there is no extensive internal 

 portion free from flagellate chambers as in that genus. 



Between the folds of the choanosomal lamella lie the primary inhalant (P.I.C) and 

 exhalant (P.E.C.) canals. The former expand at their outer ends into large, irregular 

 crypts (fig. 1, Cr. ), roofed over by the ectosome, which merges insensibly into the 

 choanosome at the points of contact. These crypts, resembling subdermal cavities, 

 may sometimes be seen from the surface, ramifying horizontally beneath the ectosome. 

 The primary exhalant canals open at their inner ends into larger exhalant channels 

 (fig. I, EEC). 



(wing to the development of the ectosome, the water does not enter the primary 

 inhalant canals directly, as in Oscarella, but by means of numerous well-developed 

 "chones" (fig. 1, Ch.), which penetrate the ectosome at right angles to the surface. 

 Each chone expands somewhat at its outer end in a trumpet-like manner, and is 

 covered over by a membrane (fig. 3, Mem.) which doubtless contains the inhalant 

 pores, but the pores are now all closed and cannot be recognised. I have, however, 



