B. Porifera incalcaria. 2. Gornacuspongiae. 5 



II. Demoterellida. 

 1. Spiculispongiae. 



A. Lithistina. 



See, supra p 4, Topsent ( 2 ), p 3, Hinde & Holmes. 



Rauff ( 3 ) found that the puzzling spicules of Anomoeladina can really 

 be reduced to tetraxons, which grow together with their actines. A full de- 

 scription and illustrations will be given later. 



B. Tetraxonina. 



Oscula etc. in Synops, Cydonium, etc. see, supra p 3, Vosmaer. 

 Fossil genera see, supra p 3, Hinde & Holmes. 



Recent genera and species see, supra p 4, Topsent ( 2 ). 



> 



C. Oligosilicina. 

 See, supra p 4, Topsent ( 2 ), p 3, Hinde & Holmes. 



D. Pseudotetraxonina. 

 See, supra p 4, Topsent( 2 ), p 4, Topsent ( 5 ), p 3, Hinde & Holmes. 



E. Clavulina. 

 See Gourret, supra p 4, Topsent ( 2 ), p 3, Hinde & Holmes. 



2. Cornacuspongiae. 

 A. Halichondrina. 



See, supra p 4, Topsent( 2 ), p 4, Topsent ( 5 ), p 3, Hinde & Holmes. 



Maas (*) observed the development of some species of Esperella [which he 

 calls Esperia], especially E. lorenzi and lingua. The larvae of these two species 

 differ in many points. Those of E. lorenzi are much smaller and possess other 

 spicules, which are besides arranged in a different way. In E. lorenzi, each 

 embryo lies in a comparatively narrow mesh of bundles of spicules ; in E. lingua, 

 the whole group of ova lie in a comparatively wide mesh together. It is probable, 

 though not certain, that the larvae of E. lorenzi leave the mother through an 

 osculum. They, then mount to the surface of the water and try to avoid the light 

 side of the tank. The larvae are not naked at the hinder end ; there is epithelium 

 all over; that of the greater part of the sponge is high and ciliated, that of the 

 hind part flat and destitute of cilia. The flagellated cells are extremely thin 

 and long, their nuclei a good deal thicker than the average diameter of the cell 

 itself. As the peripheral portions of the cells are close together, their nuclei are 

 not placed at the same level. Hence it seems that there are several layers of 

 nuclei, which is however not the case ; this is proved by the isolated cells after 

 maceration. The cells of the inner mass of the larvae are of two types: some are 

 spherical, irregularly granulated, and possess a clear round nucleus ; others are 

 generally fusiform or amoeboid, regularly granulated, and possess an elongated 

 nucleus. A (proximal) portion of the larva shows a large hole, where cells are 

 absent or very scanty. In the axis of the larva occur only large rods; in the 

 proximal area only toxa, in the distal area only chelae. The latter lie in spheres 

 together. It is possible that the whole sphere originates from one mother-cell ; 



