288 
THE VOYAGE OE H.M.S. CHALLENGER. 
calcite, which, unlike silica, is not known to originate from a colloidal state, recourse 
must be had to a different mode of explanation, and the following may be suggested. 
“ In some sponges — e.g., Thenea wallichii — the walls of the canals in the living state 
are formed by a considerable thickness of gelatinous connective tissue, from which the 
body sjDicules are absent, these occurring immediately about the exterior of the gelatinous 
tissue. If such a sponge should soon after death become covered up by fine calcareous 
mud, it is quite conceivable that the entrance of solid particles of foreign material might 
be prevented for a time by the presence of an investing skin and other soft tissues ; on 
the other hand, this would not prevent the entrance of mineral solutions, which, pene- 
trating the interior, might deposit in the vacant canals calcite or silica, as the case might 
be. Subsequently the organic matter would be dissipated by decomposition, and the 
fine calcareous mud, no longer excluded, would be able to insinuate itself into every 
space thus left vacant, and so fill up the cavity which would intervene between the 
thread of calcite or other mineral which had already been deposited, and the walls of 
the Surrounding skeleton. Thus the problematical threads of our fossil sponge may be 
regarded as representing the original cavities of the canals ; while the consolidated mud, 
or empty space, as the case may be, has taken the place of once-existing gelatinous 
connective tissue. How far this, which is the only explanation I can suggest, is the true 
one, it is hard to say. One difficulty on the face of it is the rapidity which it seems to 
require in the rate of deposition of the minerals now forming the axial threads ; but till 
we know more about the durability of certain organic tissues, and the rapidity with 
which mineral deposition takes place in shallow warm seas, this objection cannot be 
regarded as a fatal one.” 
As the structure assumed in the foregoing explanation has now been observed to 
exist, the explanation itself becomes much strengthened, and that mineral deposition 
proceeds under certain circumstances with a greater celerity than has hitherto been sus- 
pected appears probable. The parallelism between the structure of the canal- wall in this 
species of Theonella and in some species of Thenea, e.g., Thenea muricata, and more 
particularly in the position of the spicules or desmas of the general skeleton, is suggestive 
enough to merit notice. 
The mesoderm also contains numerous fibrous strands, consisting of fusiform cells 
arranged longitudinally side by side withoverlapping ends (PI. XXX. figs. 13, 17). Occa- 
sionally one of these cells is met with prolonged into two processes at one end, and at 
intervals cells of quite a different character occur; these are larger, oval, and densely 
granular (two instances not very -well represented are shown in fig. 1 3). They are prolonged 
at one or both ends into a slender thread-like process, and take a comparatively deep stain 
with reagents. The fibrous strands sometimes run more or less parallel to the wall of 
one of the water-canals, but this is not always the case, and frequently it is quite 
impossible to correlate their direction with that of any other structure in the sponge. In 
