336 S. Reginald Price. 
plant’s economy to prevent this. It is rare for a corky periderm to 
be formed by a monocotyledonous root, bat the mucilaginous sheath 
is an excellent substitute. The mucilage, which retains water very 
persistently, must be quite dried before the root-hairs are injured. 
The sheath will also serve to protect the hairs and delicate cortex 
from mechanical injury. 
How respiration takes place under these conditions is rather a 
difficult question, but the intercellular space system is very well 
developed—far more so than in most desert plants—presumably 
for gaseous circulation. It is possible that the main interchange of 
gases with the outer air takes place at the base of the plant stock. 
The roots are not positively geotropic, and Massart suggests 1 
that the light determines the depth of the root-tip below the surface 
of the soil. 
II. OTHER SPECIES. 
The following species were examined as herbarium specimens 
only. Pieces of the roots were removed, soaked in alcohol, and 
treated with glycerine for microscopical examination of the root- 
hairs and the sand particles. In some cases also sections of the 
wiry steles were obtained, but usually no precise anatomical 
examination was possible. 
A RIS7 IDA OBTUSA Delile. 
This is also a perennial grass which inhabits the desert regions 
of Egypt, North Africa, Arabia, &c. 2 In habit it is very much 
smaller, and more tufted than A. pungens, while the roots are 
smaller and probably much shorter. The herbarium material 
collected by Dr. Moss and Mr. Tansley in Algeria shows two forms 
of the species, one with a more open habit than the type, and with 
pubescent leaf sheaths, the two characters together giving a 
distinctly different vegetative facies. The distinction does not, 
however, appear to be specific, and the anatomy of the two forms 
is closely similar if not identical. 
The roots show very much the same type of adaptation as 
those of A. pungens, but this is not quite so pronounced. There is 
a distinct sandy sheath into which the root-hairs project, and these 
latter scarcely extend beyond the sheath. Moreover, as in A. 
pungens, the hairs are persistent. 
The structure of the stele seems on the whole to be very 
similar to that already described for A. pungens. The sclerised 
1 Massart, l.c., p. 239. 
2 Boisser, “ Flora Orientalis,” Vol. V, Geneva, 1884, p. 494. 
