in the interior. The surface is now rough but appears to have 
been originally smooth. An average specimen of the ellip- 
soids, which is the prevailing form, measures 0,292 mm. and 
0,247 mm. in the longer and shorter diameters respectively. 
I have not noticed any trace of depression or hilum in any 
of these bodies, which are abundant. 
There can scarcely be any doubt that these small ellip- 
soids belong to the dermal crust of a sponge similar to Geodia. 
The beautiful radiate arrangement of the <siliceous balls» of 
the recent sponges has however, through the alteration of 
the silica, quite disappeared in these examples, but I have 
had the opportunity of comparing them with similarly shaped 
spicules from the Upper Chalk of Coesfeld in which the ra- 
diate structure is still preserved equally as perfect as in those 
of existing sponges. In the Coesfeld examples I have noticed 
a small cavity in the central portion of the spicule, into 
which the radiate structure does not extend, and to the en- 
largement of this originally small cavity may be attributed 
the irregularly shaped interior hollows of the Horstead spe- 
cimens. This is further confirmed by finding in one of the 
Coesfeld examples, an abnormally enlarged central portion, 
while the exterior portion of the spicule yet retains the ra- 
diate structure. 
Excellent illustrations of the radiate structure of these 
bodies in the existing sponge Geodia Mc Andrewit, Bower- 
bank are given by Dr. Bowerbank (Mon. Brit. Spong. Vol. I, 
Pl. XXIII, XXIV, figs. 326—335. Similar siliceous globules 
are described by Mr. Carter from the Haldon Green Sand 
(An. Mag. Nat. Hist: Vol. VII, Pl. IX, figs. 55. 56); by Prof. 
Zittel from Coesfeld (Ueber Coelop. Taf. V, figs. 18—20); 
and also by Rutot from the Eocene Sand of Brussels (op. 
eit Plvd» figs: 36,37); 
The second form of these siliceous bodies (fig. 26), which 
Mr. Carter has termed globo-stellate consists of a minute 
spherical body whose surface is thickly studded with very 
