- 67 - 



bed a somewhat similai" skeletal mesh in Eubrocluis clatisus 

 from the Cambridge Coprolitc-bcd. Geo. May. 1876. p. 398, 

 PI. XIV). 



Cystispongia sp. 



(Plate V, figs. 20, 21). 



The skeletal mesh is formed of spicules with relatively 

 short arms and large sub-spherical solid nodes. The arrange- 

 ment of the .skeleton is ver)- irregular. Distance between the 

 nodes 0,225 mm.; thickness of spicular arms 0,033 mm.; 

 diameter of central node 0,11 mm. In addition to these frag- 

 ments there are others with the spicular arms much closer 

 arranged which appear to be dermal portions of the same 

 .sponge. The remains of this .sponge are more abundant in 

 the Horstead flint than those of an>' other Hexactinellid. 



The only Cretaceous sponges with which these fragments 

 can be compared are those of the Genus C}stispongia, Roemer, 

 Figures and description of Cystispongia (Ceplialites) bursa, 

 Quenstedt, sp. are given by Prof Quenstedt in his work on 

 Die Schwamme p. 492, Taf 138, fig. 17 and I have had 

 the opportunit)' of comparing examples of this .species with 

 the Horstead fragments. C. bursa is usually of an oval form, 

 the outer surface is composed of a very delicate lamina of 

 silica, immediately beneath which is a layer of closely arranged 

 spicular mesh work with large spherical nodes, much resemb- 

 ling that of my fig. 2 1 . Within the oval-shaped case of 

 silica there are thin folded layers made up of spicular meshes, 

 but the solid spherical nodes are not so large in proportion 

 to the arms as in the Horstead examples. I have discovered 

 at Horstead fragments of lamina not unlike that forming the 

 exterior membrane of Cystispongia, but none of these frag- 

 ments had any portions of the spicular mesh attached, so 

 that it is doubtful whether they really belong to this genus 

 of sponge, but the resemblance of the fragments of the ske- 



