Development of Tetradimn Cellulosntn Hall. — Rncdenmnn. 2\ 
cult to trace them in the hard and opaque matrix, the writer 
has not succeeded in obtaining other equally \'Oung or young- 
er stages and, therefore, has failed in establishing the fact of 
the symmetrical form of thisstagebeyond doubt. But the sup- 
position of the symmetrical character of the young corallites 
is supported by later stages, in which again two adjacent 
plications appear. (Cf. figs. 8, 9.) 
The next two plications grow so fast that they soon at- 
tain equal length with the first two and, thus, the character- 
istic form with four plications is produced. (Fig. 2.) The 
diameter of this stage is i mm. 
The four plications graduall}' extend deeper and deeper 
into the visceral chamber, until the}' unite in the centre, di- 
viding the whole into four chambers. Figure 10 represents 
a corallite shortly before the completion of the fission into 
four young individuals. As this and all other sections de- 
monstrate (compare figs. 8, 9), two new plications appear 
between the primary ones prior to the coalescence of the 
latter. There are, therefore, now twelve plications present, 
four longer and eight shorter ones, This stage which is 
especially frequently observable in the older compound co- 
ralla, apparent!}' has given rise to the remark, made in some 
papers on Tetradiian, that while generally there are four 
"septa" present, also a greater or smaller number may be 
observed. As mentioned before, the formation of two pli- 
cations in adjacent quadrants of these secondar}- corallites 
renders the latter also symmetrical. In the majority of coral- 
lites there appear prior to the coalescence of the primary pli- 
cations but subsequent to the formation of the two secondar}- 
ones in each chamber two more on the primary ones, thus com- 
pleting the set of four plications in each secondary corallite. 
(Cf. figs. 6a, 10.) The formation of the secondar}' projec- 
tions by folding is not so distinct as that of the primar}' 
ones; it is, however, indicated by the fact that secondar}- 
plications of adjacent corallites are rarely continuous, but 
typically alternating on the primary plications (see figs. 9, 
10); it becomes also apparent from the structure of the wall 
in the original to fig. i6a. 
By the coalescence of the first four jjlications the very 
prett}' and characteristic stage, represented in figs. 3 and 4, 
