Barley Grain in Relation to Localized Water Absorption. 385 
pass from the vascular supply in the ovary wall to the cells of the endo¬ 
sperm. The tissues of the pericarp and ovule are continuous ; indeed, this 
elongated tract is to be regarded as the base of the ovule—the extended 
chalaza—from the flanks of which the integuments originate. The whole 
structure recalls very strongly the arrangement found in Lepidocarpon and 
Lepidostrobus (8). On the ovular side of the vascular bundle, and lying 
immediately between the points of origin of the tegmen, is a group of cells 
of glandular character with somewhat thickened walls and yellowish 
homogeneous contents. In the developing grain the contents appear to be 
of an oily nature. Radiating from this group, towards the centre of the 
developing grain, is a sheaf-like mass of elongated cells which serve to 
distribute supplies to the endosperm. In the matured grain these cells are 
represented by a mass of cell-walls, the cavities of the cells having been 
obliterated by the compression due to the swelling and subsequent drying 
of the grain. The sheaf-like mass of cells frequently shows two distributing 
tracts corresponding to the longitudinal halves of the endosperm on either 
side of the furrow, and it must further be remarked that the mass extends 
deepest into the grain where the grain has the greatest circumference (see 
Figs. 3 and 4). 
The epidermis of the nucellus can be traced immediately within the 
tegmen up to the point of origin of the latter , but not across the chalazal 
tract itself 
It has been usual to regard the whole of the above-described tissues 
as nucellar in their origin. This interpretation was due to Holzner and 
Lermer, and is repeated in the work of Brown and Morris and others. 
A. Brown, in his first communication, suggested that the epidermis of the 
nucellus functioned as the selective membrane, because the deposit of silver 
chloride was seen to stop at the base of the sheaf of cells (sheaf-like mass 
of cell-walls) in the region of the furrow. As the nucellar epidermis does 
not delimit this sheaf-like mass of cell-walls, it cannot be held responsible 
for obstructing the passage of the salts at this point. 
A structure known as the ‘ embryonic appendage’ must be mentioned 
here. This is a group of cells forming the apex of the root-sheath, which, 
lying immediately beneath the micropyle, makes real contact between the 
embryo and the tegmen. The cells differ from those of the root-sheath 
proper in that they have no contents and are capable of very rapid 
swelling when placed in contact with water. Although the actual origin 
of these cells has not been proved yet, the writer is of the opinion that they 
may represent a suspensorial group. 
(b) Reaction of the various parts of the integumentary system to reagents . 
Having obtained a knowledge of the various layers and their positions, 
it was an easy matter to locate any of them in sections which had not been 
C c 3 
