G. W. Lee— Trepostornata. 
143 
from its base of attachment, thus forming an invagination or tube lining the zooecia. 
This proeess may stop at any point or persist throughout the ephebic stage, that is, in full 
sized branches. The result is a hollow branch which may attain the full size of the 
species, or a branch which is at first hollow and afterwards solid, that is, made up of 
fasciculated zooecia shaped so as to constitute the axial and peripheral regions alluded 
to. This is, however, by no means an invariable succession. Several examples of 
Tabulipora howsei (Nich.) preserved in the Nicholson Collection show a tubular, epizoarial 
stage following, in the same branch, the solid stage. How the organism had the faculty 
of suddenly secreting an epizoarium is not easily explained. In the second case the 
epizoarium does not raise itself, and the branches are solid, the branching mode being 
effected solely by the zooecia taking up the fasciculated arrangement. In both cases the 
branches are almost always of uniform diameter throughout, tapering only near the end, 
and incipient secondary branches can be as thick as, or even thicker than, the mother- 
branch. 
It has already been mentioned that the zooecia of a nepiastic colony do not differ from 
those of an ephebic colony except in point of size, i.e ., the relative dimensions of the axial 
and peripheral regions are the same. Furthermore, large suites of ephebic colonies 
belonging to the same species show conclusively that in each species the dimensions of the 
axial and peripheral regions not only are invariably the same in branches of the same 
diameter, but also that in the smallest as well as in the largest these two regions have 
invariably the same relative proportions. In other words, the ratio between axial region 
and diameter in solid cylindrical branches is in each species constant, whatever the size of 
the branch. Thus, in Stenopora redesdalensis the ratio of axial region to diameter is 
0'7 : 1 (rounding the second decimal), as in plate xiv., fig. 5 ; in Tabulipora sc otic a it is 
0*6 :1, as in plate xiv., fig. 4 ; and so on. To put it still more graphically, were one to 
superpose a series of sections of different sizes, the peripheral region of the smaller would 
partly coincide with the axial region of the larger. 
Were these organisms not differentiated into two regions characterized by the unequal 
thickness of the walls, this constant relative proportion could perhaps be explained by 
assuming—as probably done by Sardeson 1 —that the growth-activity was distributed 
throughout the whole length of the zooecial tubes, and regulated in such a way as to 
respect the relative proportions. The finer histological features shown by thin sections are 
however against this assumption ; the outer thickened region exhibits a clearly laminated 
structure, while the proximal end of the walls is apparently structureless and so thin 
22239 
1 35 &. 
