116 ART. 7. — I. IJIMA : HEXACTINELLIDA, IV. 



detectable with unmistakable distinctness in the central sphere, 

 as depicted in PL VIII., fig. 28. The periphery of the sphere 

 apj)ears compact, while more centrally in the neighborhood of 

 the axial threads are visible a number of small and irregularly- 

 shaped vacuoles which appear like dark granules. These vacuoles 

 I consider to be the same as those which are so commonly 

 seen, often arranged in rows, in the eight principal arms of 

 Acanthascine discoctasters (c/r., f. i., PI. XII., fig. 27 ; PL XV., 

 fig. 9). They are in all probability spaces which remain unfilled 

 by the siliceous substance during the amalgamation process of 

 spicular parts originally separate, and are in that sense compar- 

 able to the mesh-like spaces inclosed between synapticulse that 

 solder together parenchymalia. I am therefore strongly inclined 

 to believe that the state of things in the macrodiscohexaster of 

 the present species is brought about by the formation of synapti- 

 cular connections between the principals, and not by simple fusion 

 of their surfaces after coming into direct contact nor by their 

 total disappearance in which case the central sphere Avould be 

 identical with the central node of ordinary hexasters, which does 

 not appear to be really the case. 



The appearance of the surface of the central sphere as seen 

 by focussing up and down the microscoj^e, strongly reminds one 

 of that of a Foraminifera shell. The points of origin of the 

 numerous terminals are thickly and uniformly distributed all over 

 the surface (PL VIII., fig. 28). 



The terminals are filament-like and quite obsoletely rough- 

 surfaced. The minute prominences causing tlie roughness have 

 their points directed backwards (PL IX., fig. 9). The terminals 

 thicken just perceptibly tow^ards the outer end, where the breadth 

 measures only about 2 fx. The length is quite various, fluctuât- 



