ME. H. M. BERKAED ON RECENT POEITIDJE. 131 



upon which the system of Milne-Edwards and Haime was based. 

 It must be remembered that this system, though 50 years old, 

 has never yet been superseded so far as the two chief divisions ot" 

 the Stony Corals are concerned, the perforate and imperforate. 

 Though worker after worker fiuds it obsolete, no comprehensive 

 criticism of it has yet been attempted in connection with syste- 

 matic work *. I make no excuse, therefore, for reviewing the 

 situation which recent researches have brought about. If I 

 needed any excuse I should find it in the great trouble and 

 perplexity which the word "trabecular" has caused me during 

 the past few years. The reason of the confusion can best be 

 explained by showing with what total absence of precision the 

 term was originally used. I hope to make this quite clear by 

 means of concrete examples. 



According to Milne-Edwards and Haime the Madreporarian 

 skeleton was built up by the fusion of vast numbers of spicules 

 like those found isolated in the Alcyonaria. Fusion by terminal 

 growth of isolated spicules would naturally result in a reticulate 

 corallum. Direct evidence of this theory was found in the 

 fact that Madreporaria still existed in which the skeleton 

 was reticular, the septa being a lattice-work. Porites showed 

 these primitive skeletal conditions best, but others, e. g. Gonio- 

 pora, Alveopora, and Montipora, were all sufficiently reticular, 

 or, as it was called, " trabecular " t, to be united with Porites in 

 one family. Corals showing a further degree of fusion were the 

 Madreporidse, in which the septa are for the most part laraellatcj 

 and only the walls are reticular. Thus the Poritidae and the 

 Madreporidse were classed as Perforata in contrast to those 

 corals, such as the Astraeids, which showed a still higher degree 

 of specialization, both septa and walls being solid throughout. 



It is needless to criticize the details of this scheme, since it has 

 been recently proved to rest upon an entirely erroneous con- 

 ception of the origin of the Madreporarian skeleton. We now 

 know, primarily through the reseairches of Dr. von Koch, that 

 the Madreporarian skeleton is a purely ectodermal secretion, 

 and that the septa which appear to be internal are always clothed 



* The bearing of Miss Ogilvie's work will be .alluded to in the course of what 

 follows. 



t For criticisms of this term see pp. 137 & 145. 



