220 Proceedings of the Royal Society 
whole parallel in their course, though they diverge to a greater or 
less extent as we pass from the fixed base to the surface of the 
colony. The entire corallum is of a bright red colour, the inner 
surfaces of the tubes being, however, usually lighter in colour than 
the outer surface. The cylindrical theca? are bound together by hori 
zontal calcareous floors (the so-called “ external tabulae”) which are 
placed at varying distances apart in different specimens, but which 
are often disposed with considerable regularity, so as to form a series 
of concentric laminae parallel with the upper surface of the colony. 
These horizontal floors unite the outer surfaces of the corallites ; 
and, as pointed out by Professor Perceval Wright (“ Ann. Nat. 
Hist.,” ser. 4, vol. iii. p. 381), they are really produced by the 
coalescence of periodically-produced horizontal offshoots or exten- 
sions from the upper extremity of the tubes. Each offshoot consists 
of “a fold of the ectoderm, into which some of the endodermic 
layer is tucked,” and by the calcification of the ectodermal layer the 
connecting-floors are produced. Erom the upper surfaces of the con- 
necting-floors new corallites are budded forth, the upward expansion 
of the colony being due to the interpolation in this way of new 
tubes, as can be readily observed in the dead corallum. Horizontal 
sections through the connecting-floors show them to be traversed 
by a system of comparatively large-sized, branching, horizontal 
canals, which open at their extremities directly into the cavities 
of the polypes (fig. 1, b). Hence, if we open one of the 
thecae at what we may call one of its “nodes” — that is to say, at a 
point where it is embraced by one of the connecting-floors, we find 
a ring of rounded apertures corresponding with the openings of the 
canals just spoken of (fig. 1 , a). If, also, we examine the broken 
or cut edge of one of the connecting-floors, we observe the round and 
comparatively large orifices of the same canals as they traverse the 
interior of the floor. By means of this system of canals, the 
visceral chambers of the polypes composing a colony are placed in 
direct communication. 
Though the connecting-floors of the colony are often spoken of as 
" external tabula?,” they have in reality nothing in common with 
the transverse plates or true “ tabulse ” which intersect the visceral 
chambers of the corallites in such forms as Favosites. The only 
structures in Tubipora which can be even remotely compared with 
