ACÏINOrODOUS HOLOlIIUmOIDEA. löö 



(tcxtfig. 2ob) are mostly in the rudimentary condition, i.e., there 

 stand on a disk, which varies very much in the degree of 

 its development, three or four short teeth representing the rudi- 

 ments of a spire. In some the spire may be developed as far up 

 as the cross-beams, which may however not unite all the four 

 pillars. In the more developed ones, the four pillars grow beyond 

 the cross-beams to varying lengths and each end in a simple top 

 without being united into a crown. I have seen tables which 

 have two stories of cross-beams, and Ludwig (1874) also figures 

 such. These seem to be more common on pedicels. I have tried 

 to see if perfect tables are not commoner in the young, or if 

 there do not exist tables with several cross-beams in each, f^o far 

 as the smaller specimens in my possession go, even the youngest 

 of only 4.5 cm. length, seem to have the tables all rudimentary. 



The rods are much more numerous than the tables and form 

 a layer inside these. The most numerous form among them is 

 that shown in textfig. 23 e. It has enlarged ends and from both 

 its sides there arise several laterally directed branches, which 

 mostly terminate in a small, longitudinally directed i)iece. The 

 result is that the rods in certain cases assume the form of small 

 buttons [c, d), there existing sorts of intermediate forms. In the 

 Tomari-mura specimen (Xo. 1240) the buttons in shape approacli- 

 nearest to Selenka's figures, i.e., they are rods with irregular jn-o- 

 tuberances on each side. 



Ventral pedicels are very numerous ; they sliow a black band 

 near the tip and have well- developed end-plates. In their wall there 

 arc present, besides the same calcareous bodies found in other 

 parts, large supporting rods 0.10 — 0.15 mm. in length and with 

 a latticework developed on each side. Dorsal papilUc are much 

 more sparsely scattered and more slender. So far as I can make 



