sla 7! 
any rate is foreign and this is essential. It is unknown how the 
surrounding layers of the prosthema are formed; but it seems to me 
very probable that the mother-scleroblast of the crepis is responsible 
for it. Little is known about such secondary deposits; still less than 
about the primary -spicopal, which is in contact with the axial 
thread and very probably stands under its influence. A slight irre- 
gularity in the axial thread is followed by the same irregularity in 
the spicopal. One can convince oneself easily of this fact by care- 
fully examining microscopic preparations of spicules. Invariably 
some abnormalities are found. Take e.g. a preparation of styli; 
generally there are some which show slight thickenings in the axial 
thread near the base. The layers of spicopal follow the thickening 
most minutely. Often between the normal styli, others are found 
which are obviously arrested in growth so that they do not termi- 
nate in a sharp point, but are rounded off. In such cases one can 
repeatedly see that at that end the spicopal has formed exactly 
the same curved lamellae as, normaliter, at the base. Such layers 
are thicker, the earlier normal growth has stopped. These are patho- 
logical products. We get the impression that the silica, which is still 
present in the scleroblast, is used up, «/so if the axial canal is shut. But 
this secondary deposit is generally more irregular. We may suppose 
that similar processes are going on in desmoids, only on still larger 
scale. The reason of such abnormal development may be sought in 
the poor condition into which P/akina comes after it has been over- 
grown by Hymeniacidon. We know of several analogous cases of 
secondary deposit of spicopal. As far as 1 am aware little attention 
has been paid to it. Examples we have e.g. in sterrasters, spherasters, 
sterrospirae ; but also, I believe, in the spines of acanthostyli. Ster- 
rasters and spherasters are both polyaxon spicula; the primary 
spicopal is deposited orn the axial threads, with some form of 
oxyaster as result. If then the axial canals are shut, the secretion of 
silica goes on for a while, with the result that the centre becomes more 
and more one mass of spicopal. This mode of growth is for both 
kinds of spicules fundamentally the same; only in sterrasters it goes 
farther. For these spicules at least it has been proved that the primary 
spicopal as well as the secondary is secreted by a single cell. Sterro- 
spirae *) on the other hand are monaxon spicules. As in sterrasters 
secondary silica is secreted after the closing of the central canal, so it 
happens in sterrospirae, albeit in another way. Why the secondary 
spicopal in acanthostyli and spinispirae is deposited in concentric 
lamellae and finally as conical spines vertical on the axis, whereas 
1) Cf. Vosmaer 1902, p. 111. 
