448 
DR- G. J. HINDE ON BEDS OF SPONGE-REMAINS IN THE 
oil the summit of the Blackdown and Haldon Hills in Devonshire, and at Axmouth 
in Dorsetshire. The chert here is only present in beds of subordinate importance. 
Sponge-beds of similar characters to those of the greensand have been described 
from the Hils-sandstein, in Westphalia, which is of neocomian age, and, judging from 
specimens which I have examined, the “gaize de fArgonne,” which is largelv 
developed in the Ardennes, and the “ meule de Bracquegnies,” in Belgium, are 
sponge-beds, filled with spicules and spicular casts like those of the greensand. 
The sponge-remains in the various beds are exclusively those of siliceous sponges. 
In some the silica of the spicules yet retains its original colloidal condition, in which 
it is negative to polarised light, and readily soluble in caustic potash. The matrix of 
the sponge-beds of the malm and firestone is also to a large extent of colloidal or 
amorphous silica, and this material has been deposited in the form of minute globules 
or discs, and it seems to have been derived from the sponge-spicules, with the emptv 
casts of which the beds are throughout filled. 
More generally the origina] amorphous silica of the sponge-remains has been altered 
to chalcedony, and the chert and porous siliceous rock accompanying it, which is filled 
with traces of the spicules, are likewise of chalcedony; occasionally the chalcedony 
gives place to crystalline silica. 
Glauconite very commonly fills the canals of the spicules, and remains after the 
spicular walls have been removed ; it also replaces the spicular walls. 
In some sponge-beds the spicules have been nearly entirely replaced by crystalline 
calcite ; they are imbedded in a matrix of granular limestone. 
As a general rule the sponge-spicules are inclosed in a compact matrix, in which 
their forms can only be partially studied, but under certain conditions they are loosely 
distributed in sand or in fine powder in cavities in chert, from whence they can be 
obtained quite free from matrix. The sponge-beds appear to be wholly composed of 
detached, free spicules; entire sponges are absent. These spicules belong to numerous 
species. All four orders of siliceous sponges are represented, but those of monacti- 
nellid and hexactinellid sponges form but a small proportion, whilst those of tetracti- 
nellid and lithistid sponges, more particularly of the megamorina family, are extremely 
abundant. 
Explanation of Plates. 
PLATE 40. 
Fig. 1. Microscopic section of a sponge-bed in the upper greensand at Warminster, 
showing entire and fragmentary spicules of colloid silica, inclosed hi a 
matrix of transparent chalcedonic silica. Colloid silica in the globular 
form is also present. X 40 diameters. 
