818 W. J. SOLLAS ON THE STRUCTURE AND 



which all appear to have been derived from the chalk-silt in which 

 the Choanite was originally imbedded. 



The central network, or the true Choanite, is of a mixed nature, 

 consisting partly of the skeletal retc of the sponge and partly of a 

 mineral incrustation. Thus one observes in it true, well-defined 

 Lithistid spicules, composed of very dark and clear transparent 

 silica, with a thin axial thread of whiter silica occupying the posi- 

 tion of the axial canal, and a coating of white fluffy-looking silica 

 surrounding- them exteriorly like a growth of some kind of mould. 

 Sometimes this fluffy material is so dense and abundant as to ex- 

 clude the darker flint from the intermeshes of the network ; and the 

 central area then becomes very white and opaque ; sometimes it 

 loses its opacity, becomes less dense and abundant, and fades away 

 into a whitish blue haze, as though permeated by more transparent 

 material: and this produces the more transparent parts of the central 

 area. From the nature of these changes we might conclude that 

 the whiteness of the " fluff" is partly owing to the presence of small 

 empty spaces within it, and that it is rendered more transparent by 

 becoming filled with clear flint, by which its internal cavities are 

 obliterated and the internal reflection of light prevented. In addi- 

 tion to this spicular network we meet also with some isolated hollow 

 casts of quadriradiate spicules, white and opaque in appearance, and 

 excavated with hemispherical pits so extensively as to have become 

 almost entirely eaten away. These casts have in some places been 

 filled in with transparent silica ; and then they lose their whiteness 

 and opacity and are converted into nearly invisible granular films. 

 In some specimens, again, the silica of the transparent spicules of 

 the network has been replaced by granular iron-pyrites. 



In rare instances the whole specimen of the Choanite, excepting 

 the cloaca and canals, which are filled up with opaque white mate- 

 rial, is composed of colourless and transparent silica, and the white 

 network is wanting. The spicules then exist as mere traces only, 

 consisting of scarcely any thing more than the axial canal, which 

 has undergone a slight enlargement, and become filled in with silica, 

 which is only distinguished from that exterior to it by the presence 

 of a few dark-coloured granules. In such a case we notice in the 

 surrounding silica hemispherical bosses, with their rounded surfaces 

 turned away from the spicule, and serving as centres from which a 

 fibrous crystallization of silica radiates towards the centre of the 

 intermesh in which they occur. This arrangement resembles that 

 in Stauronema, where, likewise, the spicule has in places altogether 

 disappeared, leaving only its axial canal, where also bosses have 

 proceeded outwards from the site of the vanished spicule, and a 

 fibrous siliceous crystallization has filled up the interstices surround- 

 ing it. This also appears to be the final stage of the process which 

 led to the production of tubercles on the spicules of S. costata from 

 Wiltshire. 



Finally, in some Choanites the skeletal network, having its fibres 

 simply incrustcd with silica, forms the nucleus of an otherwise 

 hollow shell consisting of the outer and so much as is present of the 



