460 MR. J. R, GREEN ON THE ORGANS OF 
most vigorous growth. The cells, as the gland is developing, 
are polygonal in form, have very thin delicate walls, and finely 
granular contents. On treatment with osmic acid the granules 
stain black; with iodine they do not stain, but the rest of the 
cell-contents are coloured yellowish brown ; with aniline colours 
the contents of these cells stain much more deeply than those 
of the neighbouring cells. The number of cells going to form a 
gland varies very much, the size of the gland depending almost 
entirely on the number present, as the cells of those of different 
species are of about the same size. In sections of the glands of 
Hypericum perforatum, H. calycinum, &c., there are from eight 
to fifteen generally visible, while in those of H. balearicum the 
number is too great to admit of ready counting (compare figs. 10, 
11,13 with fig. 22). When the full number of cells has been 
reached the gland is still increasing in size; the phenomenon of 
absorption takes place. just as in the resin-passages (compare 
figs. 11, 12 with figs. 6,7). At this time the secretion, which in 
the younger cells could be seen as small granules, has mueh 
increased in quantity and composes nearly all the contents of 
the cells. On absorption of the latter the ethereal oil occupies 
the cavity of the reservoir so formed. With increasing age the 
cells disappear more and more completely; but they are never 
all absorbed, for an external layer remains and forms a kind of 
epithelium. 
In the oldest reservoirs the internal walls of these cells appa- 
rently disintegrate. Unlike the resin-passages, these reservoirs are 
never obliterated by subsequent changes taking place in the leaves. 
A peculiar appearance which the glands of H. balearicum pre- 
sent is figured by Unger in his work before referred to. Cells 
of the glandular tissue left lining the cavity after absorption 
is nearly complete grow out into the interior of the reser- 
voir in the form of hairs or papille, which he figures as being 
multicellular. Ihave not seen in any that I have examined such 
large ingrowths as Unger figures, but have observed several con- 
sisting of a row of two or three cells, and have seen glands with 
hairs growing out from nearly the whole of the inner surface. 
When the process of absorption of the cells of the gland begins, 
as in other species, the central cells are the first to go. Their 
protoplasm has almost all disappeared when they begin to break 
up. The cells nearer the periphery of the gland are at this 
time younger and contain more protoplasm. When the pressure 
