M. K. A. Zittel on Fossil Hexactinellida. 409 



to such an extent with that of many fossil Hexactinellida, 

 that it is sometimes quite impossible to say whether a micro- 

 scopic preparation is derived from the recent Philippine sponge 

 or from a fossil sponge from Frankish Switzerland. The same 

 phenomenon is presented bj several genera of fossil Hexacti- 

 nellida when we compare them with one another. 



Consequently we require other characters for the discrimi- 

 nation of the genera. As the flesh-spicules are usually defi- 

 cient in the fossil skeletons, the relation of isolated siliceous 

 structures, when any such are present, to the neighbouring- 

 skeletons is rarely demonstrable ; these can only be regarded 

 as of subordinate value in a system of the Dictyonina which 

 is intended to extend to the fossil forms. The peculiarities 

 presented by the surface of the solid skeleton, the water-canal 

 system, with the ostia belonging to it, and the external form 

 of the sponge-body are more important for this purpose. 



1. The surface of the skeletons of Dictyonina differs in 

 many genera in no respect from the inner parts. The lattice- 

 meshes are of the same size ; the siliceous trabeculse are the 

 same in arrangement and nature as the rest of the skeleton j 

 in short, the surface is perfectly naked. This comparatively 

 rare case is known in certain species of the living genera 

 Eurete^ Farrea^ and Myliusia ; among the fossil Dictyonina, 

 Pachyteichisma and Verrucocoelia, for example, have naked 

 surfaces. 



Much more frequently both surfaces, or at the least the 

 outer one, are entirely or partially provided with covering 

 layers {JDeckscMchten) of different constitution. Etallon, 

 E,dmer, and Pomel have already called attention to the im- 

 portance of these covering layers {PeriencJiym, Etallon ; Epi- 

 dermis, F. A. Romer; couche pellicidaire^ Pomel). 



A frequent form of covering structure is produced thus : — 

 The arms of the outermost layer of sexradiates situated in 

 the plane of the surface are thickened, or widened in the form 

 of laminse ; or they form rough and perforated, rectangularly 

 crossed beams of very dissimilar form, by the emission of 

 lateral branches, which in their turn become soldered together 

 and at the same time thicken. The outwardly directed ray of 

 the sexradiate is always aborted ; the inwardly directed one, 

 on the contrary, is united with the latticed framework. 



Such covering layers are only slight modifications of tlie 

 latticework itself. They generally still allow a mesh-like ar- 

 rangement to be recognized, although the lumina of the 

 meshes may be considerably and very unequally narrowed. 

 According to the greater or less secretion of silica, the surface 

 acquires a coarsely or finely perforated constitution. Never- 



