﻿12 MADREPORARIA. 



According, however, to the description above given (§ Morphology), the trabeculaj of the 

 Montiporre are not true trabecula3, but are purely secondary productions correlated with the 

 formation of protective tubercles on the surface. As the corallum thickens the protective 

 tubercles are gradually submerged, and appear in sections as though they were trabeculse. 



It is obvious that the whole question of the relation of Montipara to Porites turns 

 primarily upon tliis point : Are the trabeculte of the former primitive or secondary ? 



They are here unhesitatingly claimed as secondary, for the following reasons : — 



(1) They are always associated with the tubercles which are purely specialisations of the 

 surface. 



(2) They in no case extend through the corallum down to the epitheca, but are 

 developed in the uppermost thickening layer, the stroma of which is arranged at right angles 

 to that of the more primitive streaming layei'. 



(3) All stages of their formation can be traced from the straightening of certain threads 

 of reticulum ending in fine surface echinuLu to the distinct trabecules projecting above the 

 surface as stout tubercles. 



(4) They only develop late in the growth of a colony, and are not found in very young 

 forms, nor near to the growing edge. 



(5) The great majority of the known types show no trace of this formation, being, like ■ 

 the youngest colonies, purely reticular. 



These arguments, then, seem to the writer conclusive against classifying Montipora with 

 Porites, for the trabecule, which appear to be primitive structures in Porites, are secondary 

 specialisations in Montipora. Other arguments pointing the same way, based upon the method 

 of budding, will be adduced in the volume dealing with the genus Porites. 



The alternative relationslup, according to previous writers, is that with Madrepora. 

 Montipora, it was thought by Dana, might be deduced from Madrepores without distinct 

 axial corallites by the degeneration of the protuberant calicles. On the other hand, it must 

 be pointed out that there is certainly no evidence tliat Montipora ever passed through a 

 Madreporan stage with its higUy specialised method of budding. 



The relationship with the Madreporidffi, however, can be established in a far more con- 

 vincing manner by showing that all the genera can be deduced from a common parent form. 

 This common ancestral form is not purely hypothetical, but can be practically demonstrated 

 for three of the genera, viz. Turbinaria, Montipora, and Astrmopora, while our views of the 

 morphology oi Madrepwa requii-e the assumption of a siniilar parent form. 



The following are the chief characters of this common ancestral polyp : — 



(1) The secretion of a thick porous wall buUt up of lamellate septa and costse with 

 their synapticular junctions. 



(2) A saucer-shaped epitheca. 



(3) Early budding from the walls of the polyp, i.e. while it was still comparatively small. 

 From such a form all the Madreporidse can be deduced. 



In Madrepora, the parent polyp shoots up, secreting a tall conical wall, from the sides of 



