358 Gibson . — On Siliceous Deposit in the 
the small intercellular spaces between the constituent cells of 
the median cluster. It fills up these spaces completely and 
spreads as a very delicate film over the walls which are 
adjacent to each other. I can find no evidence of mineraliza- 
tion on the outer sides of these cells, facing the lacuna proper. 
Moreover, if the trabeculae be destroyed by concentrated 
sulphuric acid, the mineralization remains, and has the form 
of a number of concave shells or hemispheres united by their 
bases. Sometimes the basal cell next the cortex has a mineral 
deposit running up its external wall for a short distance, more 
commonly the deposit ceases abruptly at the base. I have 
never found any deposit on the pericycle nor on the cuticu- 
larized trabecular cells arising from it. Further, the cortical 
cells have a mineral deposit in their walls continuous with the 
surface layer in the lacuna. The small intercellular spaces 
between the cells of the innermost cortex are filled with the 
same deposit. As a rule, in var. compact^ only the three cell- 
layers next the lacuna are so mineralized; the cell-walls of the 
outer cortex contain no mineralization (Fig. 5). 
As one would expect, the younger branches have considerably 
less development of the mineral, although it can be traced 
right up the stem almost to the merismatic region, being 
coincident in its appearance with the lacunar space. Fig. 3 
shows a surface-view of the cortex lining the lacuna of a young 
branch about a quarter of an inch from its apex. Here it will 
be seen that the plates appear to arise at first close to or 
immediately over the vertical walls of the cortical cells. Two 
types of plates may be distinguished at this stage: (1) those 
which arise immediately over the vertical walls and spread 
out equally to either side, and (3) those which arise to one side 
of the vertical wall and develop over the cell-surface on that 
side only. The former are generally oblong and rather narrow, 
thickest just over the vertical wall and thinning off to either 
side. In sectional outline they appear as extremely obtuse- 
angled triangles. The second type of plate is thick at one 
(the outer) side, and thins off towards the centre of the cell. 
These plates have the appearance of razor-blades. 
