HEXACTINELLIDS WITH DISCOCTASTEKS. 



47 



considered as specially developed parenchymalia. In many species of 

 Bhabdocalyphts and Staurocalyptus I have found that in very young 

 individuals the needle-like diactin-prostalia project from all parts of the 

 body, but with growth of body, become restricted to the oscular margin, 

 where they may form an ill -defined, often interrupted fringe. 



The autogastralia are as a rule rough hexactins with some excep- 

 tions. They may be represented by pentactins or by both pentactins 

 and stauractins. The autogastral pentactins has its unpaired ray al- 

 ways distally turned. In one noteworthy case (Staurocalyptus pleo- 

 rhapMâèâ), the autogastrals are to be considered either as being not at 

 all developed or as being represented by short diactins that are not 

 sharply differentiated from parenchymal elements. When autogastralia 

 with cruciate paratangential rays are present in large numbers, they 

 form a continuous lattice-work with small quadrate meshes, covering 

 over the apertures of efferent canals. But when sparingly developed, 

 there are left in the autogastral layer wide gaps, by means of which the 

 efferent canals stand in direct communication with the gastral chamber. 



Hypogastral strands are usually more or less distinctly present. 

 They are nothing else than certain strands of parenchymal diactins, that 

 have dissociated themselves in variable degrees from the parenchymal 

 mass and have entered into the support of the autogastral layer. 



Finally with respect to dictyobasalia— by which name I designate 

 the thin reticular plate that invariably cover the surface of basal attach- 

 ment, — I believe that it is formed mainly by direct as well as synapticu- 

 lar fusion of special spicules developed by the stimulus of foreign bodies 

 in contact with the sponge. AVhere the plate is thin, the axial canal 

 contained in the nodes of beams has the shape of a simple cross, showing 

 that stauractins here lie as the structural basis. On the other hand, 

 where the plate is of certain thickness, the axial canals contained are 

 six-armed and even stout-rayed hexactins themselves are often discerni- 

 ble in the process of fusing with the dictyobasal beams. 



The nine discoctasterophorous species, known to me at present, un- 

 doubtedly constitute a coherent group separated by a gap of not incon- 



