THE TANGANYIKA PROBLEM. 315 



The inhalant canals which pass from the subdermal cavity into the chambers are 

 narrow and difficult to make out. In some cases these canals are short, owing to 

 the flagellated chambers being situated close to the floor of the subdermal cavity. 

 Those canals which pass into the chambers which are situated more deeply in the 

 sponge are long and narrow, following a winding course, and keeping nearly always 

 between the chambers and the fibres of the skeleton. On their way down into the 

 deeper parts of the sponge they give off branches which open into the chambers 1 >y 

 way of prosopyles, which are so small that it is almost impossible to make them out. 

 The apopyle was easily distinguishable as a wide opening, communicating directly 

 with the wide exhalant canals, and occupying nearly a fourth of the surface of the 

 otherwise almost spherical flagellated chamber, which is lined by collar-cells with 

 nuclei situated at their bases. The canals of the exhalant system are much wider 

 than those of the inhalant, and, as a rule, occupy a central position between the 

 fibres. As they pass down into the deeper parts of the sponge they converge and 

 unite together, forming wider canals, which are few in number, and which open 

 into the somewhat spacious gastric cavity, which communicates with the exterior by 

 way of an osculum situated at the summit of each of the mound-like elevations of 

 the surface. 



(4) The Gemmule. — The gemmules, which are few in number and scattered 

 about singly, are spherical in shape and small in size, measuring only -35 mm. in 

 diameter. They have a thin coat, which is not surrounded by spicules specially 

 characteristic of the gemmule, but by the ordinary skeleton spicules. Their cellular 

 contents present the same characters as do those of the common species of Spongilla, 

 and each individual cell is full of the two kinds of granules which are quite charac- 

 teristic of the cells of Spongillid granules. It is just possible that, had the 

 material been preserved later in the year, the gemmules would have been more 

 numerous, though there would appear to be no absolute necessity for the production 

 of gemmules, since the sponges live at a depth of three hundred fathoms and cannot 

 possibly be either dried or frozen. 



THE AFFINITIES OF ' SPONGILLA MOOREI. 



The presence of the gemmule is the most important 

 character tending to fix the position of Spongilla moorei 

 among the Spongillicke. Gemmules have been described in 

 marine sponges, and this fact diminishes the importance of 

 the existence of gemmules in a newly-discovered sponge as 

 a character supposed to be distinctive of the Spongillidse 

 (Topsent). It appears that there is no special feature in the 

 structure of the skeleton of Spongilla moorei that would 

 cause it to be separated from the Chalinidse had it been a 

 marine sponge. It most decidedly possesses more spongin 



