E. MAKSHALLT. — DEVELOPMENT OF HEXASTEES. 199 



both disappear. Interposed between the terminals in a sheaf 

 exists a certain amount of protoplasm, the exact disposition of 

 which can not be determined (fig. 35). The protoplasm contains 

 irregularly distributed nuclei, which are at first few, but increase 

 in number with the growth of the sheaf (figs. 33-35). 



Of the oxyhexasters in E. marshalli, I rarely came across 

 such as were still apparently far from being complete in the 

 development of their parts. The small, thin-rayed specimen, 

 shown in PI. IV, fig. 18, represents one such case. There can 

 be no doubt whatever that the general mode of development is 

 here essentially the same as in the other forms of hexasters. 

 Since the few cases I have seen of young oxyhexasters were all 

 found in unstained preparations, I can say nothing in particular 

 about their scleroblasts, which should have been present at the 

 center. However, in certain Eosellids (e. g., Ehabdocalyptiis 

 capillaius Ij.), I have observed numerous cases of evidently im- 

 mature as well as nearly mature oxyhexasters in connection with 

 the surrounding soft parts. Unlike the floricome or the graphio- 

 come, a capsular wall seemed to be wanting here, while the 

 scleroblast nuclei were always strikingly few in number. In 

 optical sections, at most only four of these nuclei at a time came 

 into view, more or less regularly disposed in the angles formed 

 by the principals around the central node. Judged by the size 

 of the nuclei and _of the extent of space occupied by them, there 

 could be present in each case not more than eight nuclei in all, 

 which maximum number, if regularly distributed, would bring 

 each principal to the middle of every four nuclei. In fact, I 

 think I have observed this regular arrangement in some cases ; 

 but I am far from maintaining this with any degree of assurance. 



