6 
euticula of gigantie thickness, when compared to the subjacent 
layer of cells that secrete it. 
In addition to this transparent substance the cells forming the 
deep layers of the integument have to furnish the caleareous ma- 
terial for the spieulae. The mode of growth of the latter is explained 
by the preparations in the following way. The truncated hinder 
end of each spicule is surrounded by a kind of small bag of 
cells, with very distinet nuclei; this bag appears more bulky in 
the thinner, not yet full-grown spicules (fig. 7), whereas it more 
tightly fits round the base of the spieule when this is somewhat 
thicker and more massive (fig. 8). In this bag the secretion of 
carbonate of lime and thereby the growth of the spicule appears 
to take place. The cells constituting its walls remain in direct 
connexion with those of the matrix of the integument by a string 
of connective tissue forming a sort of funiecle upon which the bag 
and the spieule with it, are implanted !). All the spieules, even 
those of the more superficial layers, being in this way connected 
with the deep cellular layer the whole of these cellular connecting 
strings traverse the interspicular substance in a radial direction, a 
free path being left open for every one of them as is shown in 
tangential sections through the integument (fig. 30,p). The nuclei 
in these connecting strings are very distinet and it would often 
appear as if they were superficially applied to them. Towards the 
centre these strings have a more fibrillar aspect. 
The smallest and evidently youngest stages of spicules corresp- 
onding in size to not more than a couple of cells are found in 
or quite close to the deep cellular matrix of the integument and 
we may conclude that it is here that all of them take their origin. 
As gradually the first formed increase in size and the layer of 
chitinous interspicular substance, keeping them together, increases 
in bulk, new sets of spieules continually develope in the cellular. 
matrix. The older, more superficial spicules become separated from 
this cellular matrix and their growth would probably have stopped, 
1) In Chitonellus (25) the connexion between the spicules and the cellular 
. matrix appears to be very much the same (ride Explan. of the plates). 
