KNOWLEDGE OF MONOCOTYLEDONOUS SAPROPHYTES. 181 
phyll consists of about five layers of cells with delicate walls and 
small intercellular spaces. Many isolated raphide-mucilage cells 
occur in the layer beneath the outer epidermis. 3. The vascular 
bundles are closely approximated to the upper (inner) epidermis, 
being separated from it by, at the most, one layer of cells. They 
are small and simple. In the larger bundles, in transverse 
section, towards the lower (outer) epidermis there is an are of 
lignified sheath-cells. The zjIem is represented by a few narrow 
spiral vessels or tracheides. The pAlo/m is excessively feebly 
developed. 4. The inner (upper) epidermis consists of flat cells 
with thin walls. Stomata are wanting, or occur sparsely towards 
the free margins. On the other hand, the remnants of hairs are 
more numerous than on the outer face of the scale. 
The bracts are histologically similar. 
The subterranean scale-leaves are like the aerial leaves, and 
their stomata are widely open and numerous. Remains of 
glandular hairs are found in the inner face of the scales. 
Distribution of Starch in Young Scales of the Scape. 
In the smallest rudiments of the scales forming a bud there is 
no starch. In slightly older, developing, scales starch occurs 
throughout the mesophyll and epidermis (not in the hairs), but 
it is more abundant in the upper (inner) half of the leaf: that 
is, on the half towards the glandular hairs. In still older scales 
this difference is more marked ; the starch has deserted the lower 
(outer) half of the leaf, and none is found in the outer epidermis 
save in the guard-cells. On the other hand, there is a large 
quantity of starch in the inner epidermis. In yet older scales 
most of the starch is confined to the upper (inner) epidermis ; 
there is little or none elsewhere excepting in the guard-cells of 
the stomata. Finally, in mature leaves, in which the secretion 
from the hairs on the inner face has stopped, the starch has 
vanished from the inner epidermis. Thus it appears that in the 
scales the distribution of starch has relation to the mucilage- 
secreting hairs. There are very few hairs on the outer surface, 
but many on the inner face. There is constantly, then, a 
flow of carbohydrate material to the inner face. It will be noted 
that this assumption that the starch is consumed to manufacture 
mucilage is confirmed by the peculiar mode of disappearance of 
the starch. Histologically I showed that conduction of material 
into the hairs must take place from the adjoining epidermal cells, 
LINN. JOURN.—BOTANY, VOL. XXXI. o 
