308 Oliver —On Sarcodes sanguined , Torr . 
tracheides, for the most part with scalariform thickenings ; 
there are a few, in the protoxylems, with spirals. The endo- 
dermis is difficult to define, and I doubt if such a layer can 
really be distinguished here. The succulent cortex is well 
stocked with starch, throughout, from bundle-cylinder to 
epidermis. The epidermis does not form a continuous mantle, 
with cells fitting everywhere closely edge to edge, but consists 
of a layer of tube-like cells — the long axis of each cell being 
at right-angles to the surface of the root (e, Fig. 44). Each 
epidermal cell is laterally not in contact with its neighbours, 
but is isolated by the ingrowth of hyphae from the investing 
mycelial sheath above mentioned. The epidermal cells at 
their first origin are in contact, but early in their history the 
hyphae penetrate between them, and closely invest each epi- 
dermal cell on every side except its base (cf. Figs. 44 and 45). 
In Fig. 44 the epidermal cells are seen in most cases in section ; 
two only (, a and b ) show a side-wall lying in the plane of 
section with the hyphae creeping over them. The nuclei 
of the epidermal cells are rod-like, or even dumb-bell shaped, 
and lie in the basal third of each cell. They are placed for 
the most part at right-angles to the long axes of the epi- 
dermal cells (Fig. 44, n). Starch occurs in the epidermis very 
rarely, though in quantities in the subjacent cortex. Outside 
is the layer of fungal mycelium constituting the mycorhiza. 
This layer is particularly well exhibited in Sarcodes , where 
it attains a thickness of *2—25 mm. The inner part, abutting 
on and penetrating between the epidermal cells, forms a very 
compact pseudo-parenchyma. In the outer layers of this 
denser stratum are embedded layers of flattened, dead cells, 
coloured deep brown with tannin. These layers, as will be 
shown below, are the remains of the root-cap. In an ordinary 
plant, in the absence of an investing fungal layer, the outer 
layers of cells of the root-cap fall away and the cap is renewed 
from within by a generating layer. Here, however, from the 
manner in which the hyphae are interwoven around them, 
this does not happen. Indeed, on the oldest roots, this same 
sheath — the survival of the root-cap — remains still enclosed 
