574 THE VOYAGE OF H.M.S. CHALLENGER. 



and even in some Coccodiscida (principally in the peripheral part of the disk or its 

 chambered arms); but both flat (or convex) surfaces of the disk (at least in the central 

 part) remain here constantly as simple lattice-plates, whilst in all Spongodiscida the whole 

 surface of the disk is spongy. 



When I constituted the family Spongodiscida in my Monograph (1862, pp. 452, 460) 

 I had separated from them the Spongocyclida, exhibiting in the central part of the disk 

 a more or less distinct concentric arrangement of the spongy chambers, whilst in the 

 former the delicate spongy framework is quite irregular, composed of branched siliceous 

 threads, connected and interwoven in all directions. But in all Spongocyclida the whole 

 surface of the spongy disk is quite as irregularly rough and deprived of smooth sieve- 

 plates as in all true Spongodiscida, and the more or less concentric structure of the 

 central part of the disk in the former (very variable and often scarcely able to be 

 recognised) seems not sufficient to separate both groups ; even the single genera cannot 

 be sufficiently separated by this character. I now therefore give up entirely the 

 group of Spongocyclida (as already done in my Prodromus, 1881). Nevertheless the 

 concentric annular structure in the dark central part of some Spongodiscida is very 

 interesting as transition to the Porodiscida ; it indicates already that the former are 

 derived from the latter. Even the single genera in both families are corresponding. 



In the new system of " Polycystina," which Ehrenberg gave, 1875 (Abhandl. d. k. 

 Akad. d. Wiss. Berlin, p. 157), are enumerated under the Calodictya four genera "with 

 spongy disk," viz., Spongodiscus, Rhopalodictyum, Dictyocoryne, Spongaster. Indeed 

 these four genera, which I here retain, are true Spongodiscida, and must be separated 

 from the other Calodictya, the greater part of which are Porodiscida. The number of 

 species of true Spongodiscida now amounts to sixty-seven, which I dispose in 

 thirteen genera. 



The whole family may be divided into three subfamilies. The first of these are the 

 Spongophacida (corresponding to the Trematodiscida among the Porodiscida), in which 

 the circular margin of the spongy disk bears no radial appendages ; either the margin is 

 quite simple, spongy (Spongodiscus), or surrounded by a hyaline, solid, or porous 

 equatorial girdle (Spongophacns). The disk is either more lenticular (biconvex) or 

 more flat discoidal (a shortened cylinder), rarely a little biconcave (thicker at the 

 margin than in the centre). The spongy framework of the solid disk is either quite 

 irregular (Spongodiscidus), or in the central part with concentric circular rings 

 (Spongocyclia), or in the central part spirally convoluted (Spongospira). 



The second subfamily, Spongotrochida, corresponds to the Stylodictyida (among the 

 Porodiscida), and is distinguished by solid radial spines on the margin of the disk, 

 disposed in the equatorial plane either irregularly or regularly (after the same order as 

 in the other families of D i s c o i d e a). 



The third subfamily, Spongobrachida, correspond perfectly to the Euchitonida 



