﻿S6 MADREPORARIA. 



which are unsymmetrical, appear in parts of two cycles ; the directives, which are sometimes 

 prominent, may meet a short way down in the calyx ; when an aperture lies below the level 

 of the surface reticulum, one or two septa may appear at the edges of the bordering flakes or 

 threads, showing that the calicles rise slowly in the thickening coenenchyma. 



The ccenenchyma in section consists almost entirely of long, thick, nodulated trabecultB 

 running continuously and very distinctly through the great tliickness of the corallum. Under 

 every tuberosity of the corallum, these trabeculas radiate fan-shaped. They are joined by 

 short stout junctions which do not run in horizontal floors. At the surface, there is no trace 

 of this trabecular structiire ; a very stout reticulum with large open pores spreads evenly over 

 the whole corallum, through the breaks in which the calicles appear. When the trabecule rise 

 at all above the surface they appear to be merely parts of the smooth threads of tlie reticulum. 

 There is no formation of sm'face tubercles. 



This remarkable coral is represented by only one specimen which was originally labelled 

 3f. crosa, Dana, by Quelch. But the peculiar trabecular structure of the ccenenchyma in the 

 specimen does not favour the idea that it could ever produce the corallum figured by Dana as 

 M. erosa. 



The specimen is about 7 cm. high, and shows a base of fracture about 5 by 9 cm. The 

 texture throughout the whole mass is singularly uniform, and not unlike that shown in 

 pulvinate specimens of Astrmopora (see left side of fig. 1, PI. IV.). The surface reticulum is 

 so loose and open that individual meshes are often not much smaller than the calicles and only 

 distinguishable from them to the naked eye by the deep fossa of the latter. 



The specimen is peculiar in the fact that it is one of the very few truly glomerate forms 

 known in the genus (cf. M. verriKosa). Massive Montiporae are generally built up by repeated 

 encrustations (e.g. M. spongodes). There is no other specimen showing such an immense 

 thickness with continuous trabecule running right through the whole depth. Long trabeculse 

 occur in M. cactus and in M. mammillata, but in the latter there is evidence of periodical en- 

 crustations and not a continuous thickening ; while in the massive forms of M. verrucosa the 

 crenenchyma is reticular throughout, without differentiation of its vertical elements into 

 trabecule. 



The upper surfaces of the single specimen have been badly injured, probably in attempting 

 to break the mass away from its attachment, but the sides show the original character (see 

 right centre of figure). 



The surface of fi'acture is interesting because it has largely followed the trend of the 

 trabeculaj which, rising upwards in the axis, bend outwards in all directions towards the 

 surface. 



Certain Balanids seem to have adapted themselves to this coral and its evidently rapid 

 growth in thickness. Settling quite small on the surface they grow longer and longer, and 

 leave long, smooth, calcareous, funnel-shaped shells running deep down into the corallum. 



I originally classed this with the specimens of M. mammillata in which the trabeculte 

 rise into short tubercles above the surface. The complete absence of tubercles on this type 

 lead to its resting place among the glabrous Jilontipores. 



a. Mactan Island. H.M.S. ' Challenger.' (Type.) 



