KHABDITE-CELLS IN CEPHALODISCFS. 263 



secretion-mass, as shown in iigs. 2 and 6 a, seems to indicate that 

 they perform the same function as the larger rods, although 

 such explanation does not meet all the facts. They are of fairly 

 frequent, but not of universal occurrence, and they are the only 

 bodies in the rhabdite-" cells " which are of a fairly constant 

 structure and position, and do not vary within wide limits. Eor 

 the rest, the irregular outline or sac of the "cell" shown in 

 fig. 2 contains a granular substance which seems to be either 

 pure plasma or a more or less slight modification of that substance. 

 It is to be noted that it is more densely granular at the two 

 poles, the base, however, being more granular than the apex. 

 Particular attention must be directed to the fact that in this 

 " cell," as in all the others examined, there is no structure 

 whatever that can with any justification be called a nucleus. I 

 have hence in describing the bodies as rhabdite-cells placed the 

 word cell in inverted commas. 



Figure 3, drawn from a single field, illustrates three conditions 

 well: — (1) the variation in position of the "cells" — d lying 

 somewhat below the surface, a and h immediately beneath it, and 

 c and f completely outside it ; (2) the fragmentation of the 

 secretion-mass as shown in/", d, ff, andi; and (3) the splitting- 

 up of the secretion-mass to form rhabdite-rods, as shown in all 

 except c. In the latter we have only the secretion-sphere and a 

 small vacuole — the most undifi'erentiated rhabdite body that has 

 been seen. In b, however, the sphere has fragmented into two 

 pieces — each piece lying apparently in a vacuole (an interesting 

 feature also seen in the rhabdite-cells of certain Tui'bellaria), and 

 just commencing to split up. Hence the serrated appearance of 

 the periphery of the two pieces. In g, d, e, and /, successive 

 stages in the splitting up of the secretion-mass are seen, whilst 

 in a the splitting has proceeded so far that a number of rods 

 have been formed connected by a central mass of hitherto un- 

 difi'erentiated secretion. The latter contains a small vacuole (see 

 also fig. 4), whilst outside the clump of rods are two small 

 secretion-masses, which may either be the two aster anlagen or 

 fragments of the original secretion-sphere. 



In figure 4 we have a condition intermediate between/" and a 

 of the preceding figure. The secretion-sphere lies in a large 

 vacuole, and the splitting up, though somewhat far advanced, has 

 yet not proceeded as far as in 3 a. The central vacuole has been 

 already noticed. Above, at the apex of the " cell," is a bent 



