122 KEPORT— ]871. 



certain extent, that between the Palaeozoic and recent faunas. Thus the find- 

 ing of species of the great Perforate genus Madrepora in the Oberburg 

 carries the genus a step further back than their discovery in the OHgocene 

 of Brockenhurst, and when taken into consideration with the presence of 

 the StephanophylUa, a perforate simple coral, in the Crag, Eocene, and Lower 

 Cretaceous deposits, and with Actinads, a highly developed compound form, 

 in the Lower Cretaceous strata of Gosau, the immense break between the 

 next form of the family and the existing is materially diminished. The next 

 form is not met with until the Carboniferous deposits of Indiana are reached 

 in a downward course ; and we owe to the late Jules Haime the knowledge 

 of the structures of PaJceacis cimeiformis, Haime, MS., from Spurgeon Hill, 

 Indiana. It is indeed remarkable that the vast eoralliferous strata which 

 intervene between the Carboniferous and the Lower Chalk should not present 

 a satisfactory proof of the existence of those members of the existing great 

 reef -building family. There is a curious fact which may be taken for what 

 it is worth in considering the absence of genera which have been represented 

 in some ancient deposits and which have not been found in intermediate 

 strata. Thus the existing West-India reefs contain abundance of the species 

 of the genus Madrepora and Millepora ; indeed they, with the forms of 

 Porites, constitute the bulk of the formations. Now, although Porites is 

 common in the Miocene reefs of the area, the others are very rare, for the 

 coral structures were principally composed of tabulate forms and Heliastra?ans. 

 Yet we know that before the Miocene reefs flourished, Madreporae and MUle- 

 pora) were common enough ; they were living all the while in other coral 

 tracts. But the break between the Palaeozoic and the Lower Cretaceous forms 

 cannot be bridged over without investigating the value of the classification 

 which separates the most closely allied subfamily of the Perforata, although 

 the Perforata are found in the Great Oolite. 



II. The Perforata characterized by a porous ccenenchyma and other tissues 

 present many modifications of their hard parts. Some approach the Aporosa, 

 and others would hardly be considered corals by the uninitiated on account of 

 the sponge-like reticulations of the skeleton. The genus Madrepora is defined 

 as follows by MM. Milne-Edwards and Jules Haime : — 



The coraUum is compound and increases by budding. The cojnenehj'ma 

 is abundant, spongy, reticulate, slightly or not at aU distinguishable from 

 the walls, which are very porous. The visceral chambers are subdivided by two 

 principal septa, which meet by their inner margins, and are more developed 

 than the others. 



The septa, especially the two largest, although perforated, are continuous, 

 and very often lamellar. 



MM. Milne-Edwards and Jules Haime distinguish the Poritidae in the fol- 

 lowing manner : — 



The coraUum is compound, and entirely formed of a reticulate ccenenchyma, 

 which is formed of trabeculae and is porous. The corallites are fused 

 together by their walls, or by an intermediate ccenenchyma, and they multiply 

 by budding, which is usually extracalicular and submarginal. 

 ^ The septal apparatus is always more or less distinct, but never completely 

 lamellar, and is formed by a series of trabeculae, which constitute by their 

 union a sort of lattice-work. The walls present the same structure as the 

 septa. The visceral chambers sometimes have rudimentary dissepiments, 

 but are never divided by tabulae. 



This family is divided into two subfamilies — 



1. The Poritinae, with a rudimentary or absent ccenenchyma. 



2. The Montiporinae, with a well-developed ccenenchyma. 



