of Edinburgh, Session 1880-81. 
227 
communication by horizontal canals, which run in the interior of the 
connecting-processes or connecting-floors, to which I have alluded 
above. 
(c.) A third point in which it has been alleged that the corallum 
of Tubipora resembles that of Syringopora , is the presence in both 
genera of an axial tube occupying the centre of the visceral cham- 
bers of the corallites. Here, however, we have, in my opinion, to 
deal with a merely apparent similarity and not with a case of true 
homology. It is true that the axial tube of Tubipora, when it is 
present, is not unlike in appearance to the axial tube of Syringopora, 
but there are important differences between these structures to be 
noted. Thus, the axial tube of Tubipora is an apparently inde- 
pendent calcareous structure, which has no connection with the wall 
of the visceral chamber save only at the nodal points of the eorallite 
( i.e ., at the points opposite to the origin of the horizontal connecting- 
floors), where it becomes connected with the inner surface of the 
theca by a funnel-shaped expansion of both its upper and lower 
extremities. In my view, therefore, there exist no u tabulae ” in 
Tubipora. Moreover, the axial tube entirely resembles the wall of 
the theca in being composed of fused spicules, and it is penetrated 
by the same close-set canaliculi. What its precise nature may be it 
is difficult to say, but it is possibly due to a progressive calcification 
of the wall of the gastric sac from below upwards as the polype 
grows upwards in the progressively lengthening theca. At any rate, 
it is interesting to notice that the only points at which it has any 
connection with the wall of the theca are just those points which, 
as shown by Professor Perceval Wright, mark the periodic growth 
from the summit of the polype of the horizontal outgrowths wdiich 
give rise to the connecting-floors of the colony. On the other hand, 
the so-called axial tube of Syringopora is not in any way an inde- 
pendent and definite structure, with an independent and definite 
wall. It is, rather, simply the axial space formed by the fitting 
into one another of a number of irregular funnel-shaped tabulce ; 
and this difference is well shown by a comparison of a transverse 
section of a eorallite of Tubip>ora with a similar section of a eorallite 
of Syringopora. In the former case, we see simply two concentric 
rings, one representing the wall of the theca, while the much smaller 
inner ring represents the axial tube, the two having absolutely no 
