112 THE VOYAGE OF H.M.S. CHALLENGER. 



manner. The arrangement will he seen in Plate I. fia;. 4. In the membrane lining the 

 calicles, in transverse sections, a peculiar structure (shown PL I. fig. 3) is to be constantly 

 observed. Stout offsets from the median connective tissue layer pierce the outer layer 

 of connective tissue cells, and hang loose externally as flattened tags, which appear as if 

 broken off, and are often somewhat curled up. I have been unable to determine the 

 connection of these tags of tissue with the calicular wall (PL I. fig. 3). Exactly similar 

 structures occur in Tubipora, being specially developed around the lower part of the 

 polyps. Beneath the uppermost tabulae scarcely any organic lining remains to the tubes, 

 if any at all, and the deeper central parts of the corallum are, in the specimen of 

 Heliopora which I have examined, almost entirely filled with the tubes of the boring 

 annelids (Leucodora sp.). Thus when a mass of Heliopora, after being hardened, is 

 decalcified, the whole of the deeper parts are removed, and a thin layer of soft tissue 

 only remains behind, which above presents a similar appearance to that of the surface 

 of the undecalcified coral, but beneath is seen to be composed of a series of villous 

 processes derived from the tubes with the bottoms of the calicular sacs appearing 

 as tubercles amongst them. Since the tubes of the coenenchym and calicles 

 have no lateral connections with one another except close to the surface of the 

 corallum, in decalcified preparations they are, excepting at their very upper extremities, 

 entirely separated from one another ; hence it is extremely difficult to prepare fine 

 transverse sections in the deeper regions, since the structures afford no support to one 

 another. 



Canal Systems. — The summits of the cavities of the sacs of soft tissue lining the 

 ccenenchymal tubes communicate freely with one another and with the cavities of the 

 polyps by means of a system of short transverse canals, which cross over the margins of 

 the walls of the calcareous tubes at the lower parts of their mouths, as already described, 

 p. 104, and shown in Plate II. fig. 7. These canals are mostly very short; they are cir- 

 cidar in section, and have the same three layers in their walls as which compose the sacs 

 within the tubes. In older parts of the coral, where the calcareous tubercles on the 

 surface are much developed and the mouths of the ccenenchymal tubes are consequently 

 contracted, a series of open channels are to be observed in the corallum running at the 

 bases of the tubercles, when the coral is looked at with a hand magnifier. It is in these 

 channels that the system of transverse canals runs. This caual system I have termed 

 the " deep canal system," to distinguish it from the system of smaller canals lying 

 superficially to it. The tube cavities communicate with the polyp cavities by means of 

 the deep canal system, through a system of large apertures shown in Plate II. fig. 2. 

 These apertures open in the intermesenterial spaces all around the summit of the calicle 

 at its periphery, a single aperture being situate in the space formed by each externally 

 projecting fold of the calicular wall. 



The superficial canal system consists of a .series of small canals and sinus, which take 



