554 W. WOODLAND. 
sides (usually, but not always, one on each side) of the primary 
rod at the centre of the perforated plate—the ‘“ corpuscule 
calcaire fondamental” ; occasionally, however, but very 
rarely they are situated more peripherally. Also in every in- 
stance the nuclei are placed rather more towards that side of 
the perforated plate on which the “stool” is borne than to- 
wards the inner aspect of the spicule. No nuclei ever adhere 
to the “stool” itself, and this develops entirely unaided in 
this respect into the form shown in fig. 55, where the spicule 
is viewed from a lateral aspect. 
The whole spicule is in Thyone, as in Cucumaria, entirely 
enveloped by cytoplasm—is contained in a protoplasmic sac 
—and each of the nuclei is, as usual, situated in a more or 
less aggregated portion of it lying to one side of the spicule 
limb. Black granules are only rarely present in the sclero- 
blasts of young spicules, and are apparently quite absent in 
those of adult spicules. 
The small irregular spicules of T. fusus [variety (c) above] 
undoubtedly arise in one cell, as is also the case with the small 
superficial spicules of Cucumaria. When first observing these 
latter I was, as the reader is already aware, strongly disin- 
clined to believe this, and I expressed myself very tentatively 
upon the subject. But later observations on additional speci- 
mens of Cucumaria and my still more recent examination of 
the similar spicules of Thyone banish all doubt upon the sub- 
ject. I have not, in Thyone, been able to discover the primary 
rod stage of development of these small spicules (the youngest 
forms which I have observed being represented by figs. 57— 
61), but this undoubtedly occurs. As the spicule grows, it 
assumes an irregularly-branched form, the ramifications of 
which, however, do not anastomose (figs. 62—65), as in the 
larger types of spicule to form perforated plates, and at a 
certain stage of growth the initial nucleus divides into two 
(figs. 63 and 64). Still later the nucleus further divides 
forming three and four scleroblasts (fig. 65)—a “ scleroblast”’ 
consisting of a nucleus with its associated mass of protoplasm 
—but I have never observed more than four. In this type of 
