66 BRITISH STROMATOPOROIDS. 



cells, which generate round themselves concentric layers of chitine. In their 

 nature the horn-cells are the primitive " radial pillars," into which they become 

 ultimately converted in old colonies, while the horizontal clathrate fibres represent 

 the first concentric lamina. The interstices of the creeping network are occupied 

 by the coenosarc, from which the polymorphic zooids are given forth. Superiorly, 

 the " horn-cells " project upwards as short tubercles, interspersed at intervals with 

 larger serrated spines. Moreover, the primitive lamina may show shallow branch- 

 ing grooves or gutters, the " astrorhizal grooves," which lodged corresponding 

 stolons of the coenosarc. 



If the colony continues to grow, the " horn-cells," or " radial pillars," grow 

 upwards, and when they attain a certain height, throw out irregular horizontal 

 processes or " arms," by the union of which a second cribriform horizontal 

 " lamina " is produced. By a repetition of this process, the colony may at last 

 assume a considerable thickness ; but, as a rule, it is only in the neighbourhood of 

 the mouth of the invested shell, where the polypites are most abundantly supplied 

 with food, that more than two or three successive lamina? are produced. In the 

 immediate vicinity of the mouth of the invested shell the colony may grow to a 

 thickness of one line or more, partly by the addition of fresh concentric lamina?, 

 and partly by a simultaneously effected absorbtion of the shell on which it grows. 

 This gradual absorbtion of the invested shell goes on over the whole surface, but 

 much more actively near the mouth of the shell than elsewhere ; and hence in old 

 colonies of Hydractinia echinata one often finds the calcareous substance of the 

 shell largely, or in parts wholly, replaced by the horny fibres of the investing crust, 

 the shell being also lined internally by a smooth horny layer. 



It is to be remembered that this kind of transformation of the shell of a Gastero- 

 podous Mollusc, though commonly the result of the growth of a colony of 

 Hydractinia, is also well known to be occasionally produced by investing parasites 

 of quite a different nature. Thus, a similar change is not uncommonly effected by 

 Suberites domuncula, Nardo ; the Sponge in this case further resembling the colony 

 of Hydractinia in the fact that it invariably, so far as I have seen, grows upon a 

 shell which is tenanted by a Hermit- Crab. The same phenomenon is also some- 

 times the result of the growth of certain of the Polyzoa. Thus, colonies of 

 Cellepora edav, Busk, one of the Crag Polyzoa, produce a similar transformation of 

 the Gasteropodous shell upon which they grow. 



I may note, in passing, that, though I have often specially investigated the 

 point, I have never observed any case in which there are indications of a similar 

 transformation of an invested shell or coral as produced by colonies of Labechia 

 or of any other Stromatoporoid. On the contrary, the invested body seems always 

 — as shown by thin sections — to retain its original form and its original surface 

 unchanged, the investing Stromatoporoid simply growing upon its exterior. 



