68 



leton is so close that it seems highly probable that they be- 

 long to Cystispongia. There is also a remarkable resemblance 

 between these fragments of mesh and the skeleton of Asty- 

 lospongia praemorsa. Goldfuss and under the microscope there 

 was little difference to be noticed in the dimensions , form, 

 and arrangement of the spicular mesh of this Silurian sponge, 

 and the examples from the Upper Chalk. 



Coscinopora sp. 



(Plate V, figs. 22 24). 



Portions of skeletal mesh having circular or elliptical in- 

 terspaces of unequal dimensions; composed of the usual six- 

 rayed spicules with open or < lantern nodes at their centres. 

 Diameter of some of the larger interspaces 0,45 mm. Frag- 

 ments of outer or surface layer consisting of a delicate mem- 

 brane perforated with circular openings of different sizes, the 

 larger, which appear to be disposed at regular distances from 

 each other, have a diameter of 0,38 mm. Fragments of the 

 root-fibre composed of nearly straight spicular fibres with 

 transverse connections. 



These different portions of the structure of a Hexactinel- 

 lid sponge appear to correspond with the Genus Coscinopora 

 Goldfuss. The type of the genus C. infundibulijormis, Goldf. 

 (Petref. Germ. i te Theil, p. 30, Taf. 30, fig, 10) is a cup 

 shaped sponge with thin walls, which are covered by a deli- 

 cate perforate surface membrane similar to that of fig. 23. 

 The inner skeleton has spicules with lantern nodes, form- 

 ing elliptical interspaces, which however are less regular 

 in their arrangement than in the Horstead examples (fig. 22). 

 The sponge appears to have been attached to the sea bottom 

 by a root-like extension of the base, which is iormed of 

 nearly parallel fibres of silica with transverse connections at 

 frequent intervals similar to that shown in fig. 24. Coscino- 

 pora infnndibuliformis appears to be abundant in the Upper 



