﻿16 MADEEPORARIA. 



Important as are the surface characters, macroscopic and microscopic, of both upper and 

 under surfaces, vertical sections through the corallum yield points of almost equal value. The 

 middle streaming layer comes to sight. This, it is here suggested, represents the direct 

 outgrowth of the reticulum into whicli the primitive radial skeletal structures of the 

 ancestral parent polyp degenerated. It is, therefore, in a way the most primitive layer of 

 the cceuenchyma. It can be seen at the growing edges and at the tips of branches, but in 

 sections, its real thickness and character (dense or loose) can be gathered. It is sometimes a 

 mere filamentous reticulum, at others it retains a trace of the old septal or costal lamellae 

 out of which it has developed. In this latter case, its elements are band-like, the bands 

 streaming outwards and fusing to form anastomosing canals with perforated walls. Seen in 

 cross section, the edges of the bands form delicate line patterns. In branches, the middle 

 streaming layer becomes the reticulate axial strand, and shows the same variations in 

 coarseness and texture. 



The lower layer of coenenchyma, which arises by the bending downwards of threads of 

 the middle streaming layer, shows a few variations, wliich, however, on account of its lesser 

 development, are not so important as those seen in the upper layer. In this latter layer 

 vertical sections reveal the stages of variation, from pure reticulum to palisades of stout 

 trabecule. 



Variations in the epitheca. — The extent to which this important element in the coral 

 skeleton is developed varies greatly : it may either foUow the edge regularly with its growth 

 from the earliest saucer-shaped young colony to its widest expansion, or else it remains 

 behind and the corallum shoots freely out beyond it. Its variations in thickness and degree 

 of concentric wrinkling have also perhaps a systematic value. In thin growths, the 

 coenenchymatous reticulum in immediate contact with it often changes into a solid layer 

 fused with it, so that the epitheca as such is not distinguishable in section. 



VIII. CONCLUDING REMARKS. 



In grouping these variations under so many specific headings, it will be seen that I have 

 selected those presented by the cceuenchyma as of the greatest importance. It is, of course, 

 somewhat bold to rely upon so variable a tissue ; and yet, one has no choice. If the diagnosis 

 of the genus above given is correct, the Montiporte are specialised by the great development 

 of the coenenchyma, hence its variations alone can form the basis for a natural classification. 



While, however, there can be little doubt that the leading ccenenchymatous variations 

 supply us -nith very good characters, one's confidence fails on coming into the region of the 

 finer surface markings. It is especially in these that the accidental variation is so great, and 

 we can thus have no hope of finally classifying the genus until we have much larger series, 

 enabling us to eliminate the chance and accidental variations from the more far reaching and 

 important structural variations. 



