ORDERACTINOIDEA. 65 



interrupting ridges are prominent and large, to others, where the 

 surface is smooth. Some traces of them are seen in the recent Mussse 

 and EuphylUse.* 



The transverse dissepiments secreted across the cells of the Pocil- 

 loporse, Favosites, and many Cyathophyllidaj (§ 46), appear to be 

 connected, as suggested by Ehrenberg, with this process of dying or 

 removal below. The base of the polyp, or, at least, the central part 

 of it, is withdrawn at intervals, and after each withdrawal, a new plate 

 is secreted by the base of the animal. 



It is obvious from the preceding, that the polyp, which is the germ 

 of a compound zoophyte, loses its identity, and cannot be said, in any 

 proper sense, to have the long life which is attributed to the full- 

 grown zoophyte itself; or else, we might have, among the huge 

 Astra?as of the Red Sea, polyps that were cotemporaries with the 

 builders of the pyramids. 



C. Coalescence of Branches. 



63. The forms of zoophytes ai-e farther modified by the frequent 

 coalescence or growing together of branches. A clump is sometimes 

 so united in this way, that only the branchlets at the extremities 

 are entirely free ; and occasionally a branching corallum finally be- 

 comes nearly solid, a few holes intersecting or riddling the mass, 

 being the only indications within that it was a ramose species. 

 When the species ramifies in a plane, the coalescing branchlets some- 

 times produce a complete network, as in the sea-fan (Gorgonia 

 fiabellum) of the West Indies. The vase Madrepores are other 

 examples of the same. This coalescence is so complete in some of 

 the horizontally growing Madrepores (M. palmata and fiabellum), 

 that they form broad solid plates or folia, with perhaps an inch or so 

 of the coalesced branches, free at the margin of the plate. 



In foliaceous zoophytes, the same coalescence may take place. In 

 certain species, the folia curve around until the edges meet and 

 grow together, and produce a chimney or tubular form, as in the 

 Echinopora reflexa. Again, a plate folds upon itself, and the parts 

 unite, back to back, so that a species, which usually has polyps only 

 on one surface, — unifacial, — may change its character and resemble 

 bifacial species, in which polyps open on both sides. 



A broken piece of live coral, placed against another of the same 

 species, will soon grow to it and continue its existence as if unin- 



* See plate 6, figure 'ib. 

 17 



