166 I. IJIMA : HEXACTINELLIDA. I. 



Hexactinellid species as will be demonstrated in the course of this 

 series of Contributions, are spherical or approximately so ; some- 

 times they are ovoid and, when several lie close together, they 

 may by mutual pressure approach a polygon in shape. In dia- 

 meter, they measure only S-oYo ,«, rarely as much as 5/^. The 

 variation in size is mainly due to the greater or less quantity of 

 the cytoplasm which is on the whole sparingly present, being 

 represented by a thin layer around the nucleus. The outer 

 contour of the cell is even and distinct, without any indication 

 of the presence of an investing membrane. 



The cytoplasm appears nearly homogeneous. Osmic acid 

 slightly browns it. In specimens hardened with corrosive-sub- 

 limate or with absolute alcohol, it takes up coloring matter very 

 well, so that the outline of the nucleus within is somewhat ob- 

 scured. If this fact is considered together with the smallness of 

 the cells, which are often of nearly the same size as the trabe- 

 cular nuclei, there ought not to be much wonder if the cells 

 should be thrown together with the latter under the nuclei, in- 

 stead of being recognized as complete cells. In fact, in my 

 earlier notes and sketches concerning not only E. marshalU, but 

 also a number of other species in which I later recognized their 

 true nature, I find them put down as nuclei ; and I venture to 

 say that F. E. Schulze possibly fared similarly in his study of 

 Schaudinnia arctica (see ante, p. 1Ö9). Examined under an 

 immersion system in successful preparations, the cell-body stands 

 out as distinctly as I have represented it in PI. V, figs. 37 and 43, 

 inclosing an indubitable nucleus which in its turn contains darkly 

 stained chromatic bodies. Whereas, if seen under only a moderate 

 power, the entire cell-body might easily pass for a nucleus and 

 the real nucleus for a chromatic inclosure. 



