Prototheca of the Madreporaria. 9 



(cf. the minute colonies of Madreporids already mentioned). 

 Such series of gradually expandino- colonies might grow into 

 columnar or massive stocks widening at the top. In all such 

 stocks the tabulae which run through them must he regarded 

 as the floors of successive saucers. This is well exemplified 

 in the genus Goniopora, as I have already explained *. In 

 this genus too we have, as we have in Montlivahia, irregular 

 bands of epitheca running round tlie stocks. These are the 

 rims of the protothecal saucers showing irregularly at the 

 surface. In Alveopora the rims all run together to form 

 continuous epithecal investments^ except, perhaps, in their 

 branching forms, in which the prototheca3 may be lifted up 

 above one another by the growth of the spiny septal skeleton. 



For an understanding of the morphology of the coral 

 skeleton we must bear in mind that essentially the same 

 process, viz. a succession of epithecal cups or saucers^ occurs 

 throughout the whole of the Madreporaria. Tliey may be 

 simple conical cups fitted one into the other {Zaphrentis) or 

 flat plates piled up {Montlivahia, Goniopora) , or their epithecal 

 floors may be thrown into complicated folds and both the 

 cup and its repetitions may be difficult to unravel, but the 

 fundamental principle is the same throughout. Tliere is only 

 one group I can think of in which the epitheca is not nor- 

 mally repeated, namely in the highest Madreporids — Madre- 

 ptora, Turbinaria, Montipora, Astrceopora, and their simpler 

 ancestors the Eupsaramiids. In these the purely septal 

 skeleton rises rapidly above the original flattened prototheca 

 which is then left behind. This is the reason of that well- 

 known characteristic of these forms that the calicle cavities 

 run continuously for long distances through the skeleton. 



We repeat, then, for the sake of emphasis that wherever 

 the epitheca occurs it represents the rim or the coalesced rims 

 of one or more protothecal cups or saucers, tiie floors of which 

 are represented by the tabulae. In any individual case the 

 tabula below the living layer is the n\\\ repetition of the 

 original prototheca of the parent polyp. 



The main problem, then,, of the student of coral morphology, 

 that is, taking the skeleton alone into account, is to trace the 

 various moditications of the prototheca from its earliest simple 

 cup stage to the many different shapes and positions it now 

 assumes and occupies as part of the coral skeleton. 



Roughly speaking, we may say that there are two periods 

 in the evolution of the Madreporaria — that in which the 

 prototheca, though modified, remains in evidence, and that in 



* Vf, vol. iv. Brit. Miis. 3Iadiepor;u-ia, p. 2-1, diagram A, 



