320 MR. L. LEWTON-BRAIN ON THE ANATOMY 
midrib forms the greater part of the leaf, the other ribs have disappeared, and the whole 
of the lateral parts of the leaf have become mere appendages of the midrib. By a still 
further increase of the midrib and an even greater reduction of the lateral parts, we 
should arnive at the type met with in Aira flexuosa. 
Perhaps one ought to describe the extremely small leaves of such grasses as Mibora 
verna (Pl. 36. fig. 1), Aira caryophyllea, Aira precoz (Vl. 36. fig. 5), &c. as a fifth type. 
These grasses are generally annuals, with a short vegetative period, and are characteristic 
of dry exposed situations. "The leaves are very short and narrow, and their structure 
is extremely reduced. 
HISTOLOGY. 
Epidermis —The epidermis is, as might be expected, the most variable of the tissues 
of grass-leaves. Under this heading, we have to consider not only the ordinary cells of 
the epidermis, but also the motor cells, hairs and other emergencies, and the stomata. 
The epidermal cells generally appear more or less square in transverse section (Pl. 40. 
figs. 53-56); in longitudinal section they appear as enormously elongated rectangles 
(Pl. 40. fig. 74), the long axis being thus parallel to the long axis of the leaf. Occasion- 
ally we get interspersed among these long cells short cells of various forms. The cells, 
as is well known, are arranged in rows, running along the long axis of the leaf. 
Grob (1896) has worked out the histology of the epidermis of grass-leaves pretty 
thoroughly, though unfortunately he is not concerned with many of the British species. 
The form of the outer wall of the epidermal cell varies considerably. Most generally, 
perhaps, it is slightly arched outwards (Pl. 40. figs. 53, 55, 73). A common feature is 
for the upper epidermal cells to be arched, while those of the lower epidermis are much 
flatter (Pl. 40. figs. 53-56); the lower epidermis is also commonly more strongly 
cuticularized than the upper. These features are, of course, best seen in the xerophytic 
grasses, e. g. T'riodia decumbens (Pl. 40. figs, 58, 59), but are also well marked in many 
other grasses, Hierochloé borealis (Pl. 40. fig. 66), Lolium temulentum (Pl. 40. figs. 55, 56), 
Festuca duriuscula (Pl. 40. fig. 62), Ee Another fairly common feature is for the 
epidermal cells which adjoin the bands of mechanical tissue to become smaller and 
to have thicker walls than the remaining cells (Pl. 40. figs. 62, 72) ; in fact they take on 
the appearance of sclerenchyma-cells, and generally the outer wall is much thicker than 
the others. The arching outwards of the outer wall of the epidermal cell is sometimes 
much exaggerated, and the cell in fact becomes papillose, as is well seen, for example, in 
the cells of the upper epidermis of Glyceria fluitans (Pl. 40. fig. 63); in Mibora verna 
again (Pl. 40. figs. 60, 61) the cells of both the upper and the lower epidermis are much 
arched, and are, moreover, very feebly cuticularized. These features are in harmony with 
the general feebleness of the tissues of this leaf, and depend upon its extreme smallness. 
Mention must also be made of the very exceptional epidermal cells of Setaria viridis 
(Pl. 40. fig. 65), which are most unusually large—each cell, in fact, forms from one- 
fourth to one-third of the thickness of the leaf. I cannot understand the function of 
these large cells. It has been suggested to me that they serve for water-storage; but 
this is, I think, negatived by their position, and by the fact that neither epidermis is at 
