182 I. IJIMA : HEXACTINELLIDA. I. 



of Hexactinellids in which I have fouud unusually large cells, 

 more or less ovum-like in appearance. 



The one case occurred in a small Kossellid species which I 

 have described under the name of Leucopsacus orlhodocus (Ijima 

 'g8, p. 42). Of that species I havo had only two specimens in 

 all. In one of these, but not in the other, were discovered in 

 some numbers spherical or ovoid, comparatively large cells of 

 20-40 /i diameter, attached to the trabecule^ or lying apparently 

 free in the lacunar spaces. They had smooth, sharply contoured, 

 external surface. The finely, densely and uniformly granular 

 protoplasm is tolerably well stained by carmine or by bsematox- 

 ylin. The centrally situated and exceedingly conspicuous nucleus 

 is spherical, large {iVji-l^ ji dia.), intensely stained, and nearly 

 homogeneous in appearance ; but at times it shows an obscurely 

 heterogeneous structure. A very wide gap in size and in the 

 appearance of the nucleus separates the cells from the largest 

 archteocyte found in the species, the latter measuring not more 

 than 'S^jii'. in diameter and having a nucleus which usually 

 has a single, refringent, chromatic granule in its interior. Being 

 thus unable to trace the cells back to their origin or to follow 

 their subsequent history, I do not feel that I am in a position 

 to form a judgment as to their real nature. Ovum-like as (hey 

 appear, the possibility of their being extrinsic, even perhaps a 

 Protozoan parasite, can not be excluded. In the same specimen 

 I have found, as in Viirollula fertile (p. 162), a number of 

 larvte in different stages of development. This seemed at first to 

 be suggestive of the import of the above-mentioned large cells. 

 However, I have failed to discover segmentation stages, or in 

 fact, any sign of genetic relation between the cells and the 

 larvie. On the contrary, the origin of the latter seemed to be 



