490 Proceedings of the Royal Irish Academy. 



acids, not destroyed by ignition, nor to any appreciable extent 

 charred : it is transparent and colourless, and easily stained by 

 eosin. The effect of ignition upon a siliceous spicule, or corpuscle, 

 after it has been boiled in potash, is very singular; it no longer 

 violently decrepitates when heat is applied, but entirely changes its 

 optical characters ; its transparency and glassy appearance vanish ; it 

 becomes nearly opaque, or rather translucent ; appears white and opa- 

 line by reflected light, and granular and yellowish -brown by trans- 

 mitted light. The beauty of the corpuscle is in no way affected by 

 its transformation : seen by reflected light, it makes a good " shew" 

 object. 



It is the film which remains after partial solution of a corpuscle 

 which is of importance to us now ; for though of scarcely any sub- 

 stance, and tender to the last degree, it is still sufficient to hold the 

 corpuscles together, and still retains the form of those parts of the 

 corpuscles from which the silica has been removed — parts to us of 

 critical interest, since they are those concerned in the union of the 

 corpiiscles. The skeleton, after boiling in potash, is transferred to 

 distilled water, and all traces of alkali removed ; it is next placed in 

 magenta, which stains the residual membrane, and then left to soak in 

 melted gelatine jelly. It thus becomes prepared for cutting in the 

 freezing microtome, according to the method already described by me.* 

 Paraffin would of course be preferable as a medium for embedding 

 were its use here practicable ; but owing to the mixture of very hard 

 and very soft parts, and the discontinuity of the slice when cut, one 

 has no choice, and must have recourse to the gelatine freezing process. 

 The slices of jelly containing the cut skeletal network may be dehy- 

 di^ated, and mounted in balsam, or at once put up in glycerine. It is 

 best to mount in both ways : the spicules show better in balsam ; the 

 connecting film in glycerine. Slices so prepared reveal the structure 

 of the skeleton beyond all doubt. It consists of a network of which 

 the nodes are mainly of one kind, though produced not by the union 

 of the ends of rods or rays, but chiefly by the centra of the corpuscles. 

 The ends of the rays abut upon these centra, and by means of their 

 foliaceous expansions actually embrace them, so that an exceedingly 

 dense and resistant skeleton results. The two kinds of nodes exist, 

 but are confused or coincident. 



Frequently one observes in a section places where a centrum has 

 been torn away ; and then the expanded ends of the rays can be seen 

 remaining in their original position, side by side, so as to form a kind 

 of cup, within which the centrum, now torn away, originally lay 

 grasped. 



The centrum, as it frequently emits rays from only one part 

 (first kind of corpuscle), receives rays only on the other ; and about 

 the place of union the shrub-like growth of spines occurs, concealing. 



4 Sollas. Q. Jour. Micr. Sci., N. S., xxiv., p. 163. 



