270 
THE VOYAGE OF H.M.S. CHALLENGER. 
The test is large in amount relatively to the size of the colony, and it is 
rather hard and quite opaque. The test cells are few in number. They are seen 
best in the layer of test which forms the upper surface of the colony (PI, XXXV. 
fig. 2, t.m.), and which is in most places free from spicules. The test cells are mostly 
fusiform, and are placed with their long axes parallel to the surface. 
The spicules are very abundant, but not large. The general form is stellate (PI. XXXV, 
figs. 2, 3,sp.), but they are characterised by their irregularity. The rays are of unequal 
length, and are frequently as thick near the apex as at the base, or even thicker (PI. XXXV. 
fig. 4). In some cases smaller spicules with very short rounded rays, or having merely a 
knobbed spherical form, are seen (PI, XXXV. fig. 4), 
As this species, on account of the comparatively simple structure of the test, and the 
large number of very distinct spicules present, seemed to be a favourable object for an 
investigation into the relation between the calcareous spicules and the surrounding test 
matrix and the mode of formation of the spicules, I cut a large number of thin sections 
and examined them after various methods of treatment, with the following results ; — 
When a section of the test is treated, under the microscope, with a drop of hydro- 
chloric acid, the spicules are, seen to be dissolved out with effervescence, and after a short 
time they competely disappear, leaving a large number of rounded or polygonal vacuoles 
in the test (PI. XXXV. fig. 10) which were not visible before. In many of these vacuoles 
distinct thickenings are seen at the angles, and in some cases the angles are produced 
outwards to form a rudely stellate figure (PI. XXXV. fig. 10, — notice the vacuoles 
marked i, 2 , 3 ). 
When a specimen prepared in this manner is stained with aniline blue solution a 
delicate membrane bounding the vacuole is brought into view by its taking on the stain 
more deeply than the surrounding test matrix, and the angular points, which were just 
visible before, now appear as distinct thickenings on this membrane (see PI. XXXV. 
fig. 5). 
AVhen a small piece of the test is teased (before decalcification) so as to release some 
of the spicules, and these are then stained in aniline blue, in some cases a delicate 
membrane enclosing the spicule can be made out. It is seen best where it stretches across 
from one ray of the spicule to the adjacent one, as it sometimes does without dipping 
completely into the angle between (see PI. XXXV. figs. 8, 9). 
When a section of the test is treated with very dilute hydrochloric acid, which is 
washed off when the spicules are only partially solved, and is then stained in aniline 
blue or in eosine, spicules in all stages of solution are found, and they are seen to lie 
each in one of the rounded or polygonal vacuoles which have made their appearance, 
and the outlines of which are distinctly stained, and it is also now seen that the 
thickened angles of the vacuoles correspond to the apices of the rays of the contained 
spicules (see PI. XXXV. fig. 7, 1 and 2). . 
