476 ASTROSCLERA WILLEYANA, THE TYPE OF A XEW FAMILY OF SPONGES. 



On the Spherulitic Structure seen in Pharetrones from St Cassian and elseiuhere. 



In the triassic deposits of St Cassian and others occurring near Schhiderbaeh and 

 referred to the same horizon, representatives of the Pharetrones occur in which the 

 carbonate of lime trabeculae of the skeleton show no trace of spicules, but, instead, 

 the\' are made up of more or less polyhedral systems of radiating lines which at their 

 periphery are in contact with adjacent systems either by a sharp or a confused 

 boundary'. They thus remarkably resemble the trabeculae of the skeleton of Astrosdera. 



The view of this structure held by Zittel, and by most palaeontologists who 

 have considered the matter, is that it is secondary, being due to a recrystallization 

 of the lime. I have carefully examined sections of specimens which I obtained from 

 these localities, in several of which the sphaerulitic structure is well marked, and I 

 am prepared to accept this conclusion as correct. 



The evidence bearing on the matter may be arranged under separate heads. 



a. If the spherulitic structure of the fossils were primary we might expect that 

 those in which it is present would form a group differing in other respects from the 

 Pharetrones with a spicular skeleton. The opposite is the case. 



Among the genera included in the group there are simple forms, of which Stelli- 

 spongia and Corynella may stand as examples, and remarkable segmented forms con- 

 sisting of hollow segments superposed one upon another, so that the roof of one forms 

 the floor of its successor. These have been divided by Steinmann as a separate group, 

 the Sphinctozoa, and Zittel regards them as so distinct that he removes them from the 

 Pharetrones, and classes them in a separate order. The precise relationship of the two 

 groups does not materially affect our present enquir}-. 



The genus Verticillites is one of these segmented forms. In a representative of 

 it occurring in the Greensand beds of Warminster (F. D'Orbignyi, Hinde) the com- 

 position of its walls of tri- and quadri-radiate spicules has been demonstrated as clearly 

 as in any recent sponge. (Cp. Hinde, Ann. and Mag. ^. H. 5, x. PI x. Fig. 3 &c.) 



On the other hand, the genera Thaumastocoelia and Enoplocoelia of Steinmann are 

 examples of the segmented forms from the triassic strata of St Cassian, and the 

 specimens which have been examined present, wholly or in part, a spherulitic structure 

 of the walls. (The structure was regarded by Steinmann as primary in the fcjrmer 

 genus, but as secondaiy in the latter.) 



Again, the genus Corynella is one of the unsegmented forms and possesses distinct 

 characters'-. 



' The structure is figured by Zittel, " Studien iiber fossile Spongien, in.," Abh. d. ilath.-Phys. Classe d. kon. 

 bayerischen Akad. d. Wisseiischaften, Bd. xiii. Abth. 2, PI. xii. Fig. 5 {Corynella gracilis), also in his GrundzUge 

 der Palaeontologie, p. 59, Fig. 88. 



- According to Zittel's description (" Studien iiber fossile Spongien," p. 3-5) they are usually solitary 

 sponges, cylindrical or pear-shaped, thick-walled, and with a truncated or rounded upper surface. The gastral 

 cavity is funnel-shaped, rarely reaching to the base. Eadial grooves are often present round its mouth and 

 wide tributary tubes pass outwards and downwards from the gastral cavity. As they penetrate the walls 

 they become narrower and ultimately open by fine inhalent ostia at the surface. A thick dermal layer is 

 sometimes present at the base. The skeletal fibres are coarse, chiefly composed of simple spicules, but with 

 some large tri-radiate ones. 



With regard to the 'simple' spicules Hinde points out that those of C. socialis are tri-radiates so 

 modified that they might very readily be mistaken for simple ones. 



