54 EOCENE AND LOWER OLKIOOENE CORAL FAUXAS. 



If the dissepiments at the end of a growth period are formed in a plane 

 perpeiidicuUir to the vertical axis of the corallite, they are called tabulae. 



The microscopic structure of dissepiments is quite different from any- 

 thing described in the preceding, excepting epitheca. No calciffcation 

 centers could be detected in any dissepiments. They are composed of 

 fibro-crystals standing at right angles to the secreting surface, and show a 

 more or less distinct Inmination parallel to the secreting surface. An 

 enlarged section of a dissepiment is shown in PI. II, fig. 6, Tltysanus excen- 

 triciis, and in PI. II, fig. 2, EusniiUa knorri. These specimens are not quite 

 so satisfactory as the material I have of Diphr'ta cerehrifoniiis, on which the 

 above notes are based. The method of growth of the dissepiments is 

 especially well illustrated l)y the specimens of Haimesiastrma conferta 

 (PI. XV, figs. () and 8). The formation begins against tlie septal faces and 

 extends across the interseptal loculus until the two parts meet and fuse. A 

 distinct suture indicating the ])lace of fusion can nearly always be seen in 

 good fresh specimens.^ 



KEIHARKS ON THE CL,ASSIFICATIOK OF CORALS. 



The classification of corals is in a most unsatisfactory condition, and 

 no classification that will stand the test nf tliorDUgh criticism has as yet 

 been proposed. It would re([uire too niiicli space to review all the systems 

 that have been sugge.sted. In this jniper the following family names are 

 used: Turbinolida' (in which family Parasmilia is included), Oculinidte, 

 Stvl<ii)htiridi\i, Astrangidffi, Astroccenida, Fungida?, Eupsamraidas, and 

 Madreporida*. The name Astreid;e has properly no i)lace in the classifi- 

 cation of corals, as there is no genus bearing the name Astrea (see p. 154). 

 Therefore I have not used it, and several genera that hitherto would have 

 been referred to it, as formerly understot)d, I have not placed in any family. 

 The whole classification of corals from l)eginning to end must be worked 

 over, and in mv opinion manyof tlie old families must be divided into sev- 

 eral if n( it man's' families. A new classificat i( m must lie leased on a knowledge 

 of three things; P^irst, the coniparative anatomy or morphology of the corals, 

 i. e., an intimate knowledge of the skeletal structures and their relations to 

 the soft parts; second, the post-embryonic development of the corals; third, 



' The methods of reiiroductiou .and the combination of zooids into colonies will not be discussed' 

 as iiny late text hook on paleontology will give a general idea of the subject. 



