200 MR. P. GROOM—CONTRIBUTIONS TO THE 
cells surrounded by five rows of cells (Pl. VI. fig. 7). The 
superficial cells may decrease in number higher up the pro- 
tuberance. Clinging to the hairs are particles of humus, around 
or against which the hairs may coil or flatten themselves, 
Mycorhizal hyphæ penetrate by means of the hairs, and enter 
the cortex of the rhizome vid the axial cells of the outgrowth 
(Pl. VI. fig. S). By no other channel do the mycorhizal hyphe 
penetrate the plants: hence the general epidermal cells contain 
no hyphe. The hairs and the superficial cells of the protuberance 
are feebly cuticularized. Below the lower scale (s") every hair 
appears to be permeated by mycorhizal hyphe. Above the lower 
scale the protuberances are many of them larger, and their hairs 
are much more elongated, straight, and more slender; these long 
hairs are not penetrated by hyphe nor do particles of humus 
cling to them. At the same level other protuberances are just 
similar to those described as occurring lower down; they have 
more irregular broader and shorter hairs containing hyphe. Here,. 
then, the direct effect of the mycorhizal hyphe (or of absorbing 
fluid?)is to shorten and broaden the absorbing hairs. For I 
assume that these hairs throughout function as root-hairs, since 
they are intimately associated with particles of humus and with 
mycorhizal hyphe. 
Within the epidermis succeed about ten layers of cortical 
parenchyma. (There is no trace of an exodermis.) In some 
spots three to four outer layers of parenchyma contain myco- 
rhizal hyphe forming typical coiled mycelia (Pl. VI. fig. 8). 
These cells contain no starch or very little, and the hyphe can 
be directly traced from them out to the protuberances. The rest 
of the cortical cells near the insertion on the tuber contain small’ 
starch-grains ; or some of them are raphide-mucilage cells. The 
intercellular spaces are of medium size. Ascending towards the 
scales the starch diminishes. 
There is a curious endodermis, irregular and with thin walls. 
The typical dots are to be seen mainly, but not exclusively, on 
the radial walls of these cells (Pl. VI. fig. 9). Within there 
is an interrupted layer of pericycle. The vascular bundles are 
abnormal. There is a clear tendency to a radial arrangement, 
but the bundles of xylem are placed nearer the centre of the 
stele than are the pAloém-bundles; and groups of phloém stand 
both between and radially outside the bundles of xylem. 
The bundles of xylem are feebly developed, being represented 
by about three groups of tracheides which have transverse pits- 
