526 MR. L. LEWTON-BRAIN ON THE ANATOMY 
organs, it is arranged in bands just within the outer epidermis. This is particularly well 
seen in Aira flexuosa (Pl. 36. fig. 6), where the small bands under the lower epidermis 
with the marginal bands and the band at the summit of the single ridge on the upper- 
side form an almost complete hollow cylinder of mechanical tissue. 
The reed-grasses, again, have broad.leaves, and on account of their tall stems the 
leaves are exposed to winds. more than is usual in water-grasses. Accordingly we find 
the mechanical tissue unusually well developed. 
Again, the maritime sand-grasses are constantly exposed to terrifie storms and winds, 
and we should naturally expect, on this account, an exceptional development of the 
mechanical tissue. But we might also expect these exceptional strains to be met simply 
by a stronger development of the system of girders and marginal bands, as is actually 
the ease in Elymus arenarius. We have also to explain the presence of the sub- 
epidermal bands in Psamma arenaria and Agropyrum junceum. We must remember 
that under dry conditions the leaves of these grasses are rolled up, and that dry 
conditions are very frequent for them. Thus, during a greater part of their life, these 
leaves are not bilateral, but cylindrical organs, and the subepidermal band is thus the 
mechanically best arrangement of the stereome, forming as it does a hollow cylinder of 
strengthening tissue at the periphery of the organ. 
I have already mentioned Kerner's suggestion that these subepidermal bands of thick- 
walled cells aid in the rolling up of the leaf. 
Chlorophyll-containing Tissue.—In the great majority of grass-leaves the chlorophyll- 
tissue appears in transverse section to be composed simply of a mass of closely packed, 
irregularly shaped, but more or less isodiametrical, cells with small intercellular spaces 
(Pl. 40. figs. 58, 67, 73, mes.). 
In longitudinal section, however, the arrangement of the chlorophyll-cells is seen to 
be a much more definite one (figs. 74, 75, &c.). Under the upper epidermis we find a 
regularly arranged layer of cells (fig. 74, p.!), which, by reason of its position, the 
arrangement of its cells, and probably its function, invites comparison with the palisade- 
tissue of the leaves of Dicotyledons; the cells are, however, nearly isodiametrie. Under 
the lower epidermis we get a very similar palisade-layer (figs. 74, 75, p.2). Then, 
stretching from the upper palisade-layer to the lower, we get branched chains of cells 
(figs. 74, 75) with much larger air-spaces. "These chains, where they would be broken 
by a vascular bundle, become attached to the outer bundle-sheath. These features are 
much more striking in some cases than in others, but in the small number of leaves that 
I have examined they can always be made out. A more complete examination of the 
assimilating tissue of grass-leaves would probably bring to light many interesting facts. 
Occasionally the cells both of the * palisade” and “spongy” mesophyll have very 
curious shapes, and further investigation would probably show that this is common. 
A much more exceptional form of assimilating tissue is met with in a few British 
grasses, as Poa rigida and Setaria viridis. In these leaves the assimilating tissue is 
arranged in rings around the vascular bundles (Pl. 40, figs. 65 & 72). These grasses are 
all more or less xerophytic in habitat, and the arrangement of the chlorophyll-cells in 
rings, which according to Volkens is also found in Egyptian desert grass-leaves, is 
