316 MR. L. LEWTON-BRAIN ON THE ANATOMY 
remembered that in these cases the grass may often be recognized without an examination 
of the leaf-section. Taking, for example, the four species of Aira that I have examined: 
Aira caryophyllea and A. precox are ruderal grasses, Aira flexuosa is a heath-grass, and 
A. cespitosa is found in moist and shaded places. The first two species are very similar 
iu habit and cannot be separated, except when in flower; they are, however, easily 
distinguishable from A. flexuosa and from A. cespitosa, as also these two species one 
from the other, merely by the general habit and the form and size of the leaves. 
Turning now to the anatomical structure of the leaves, we find that both Aira cespitosa 
(Pl. 37. fig. 20) and A. flexuosa (Pl. 36. fig. 6) have very characteristic leaf-sections 
and are easily recognizable, but that the sections of A. caryophyllea and A. precoz 
(Pl. 36. fig. 5) are practically identical, though very different from those of the former 
species. This is by no means an isolated case, and it is evident that we must not rely 
too much on the leaf-sections for separating closely allied species. 
Pée-Laby divides the grasses into five main groups, depending on the following 
characters seen in transverse section:—(1) Whether or not the upper and lower leaf- 
surfaces are parallel; (2) the relative number of stomata on the two sides; (3) the 
arrangement of the chlorophyll-tissue ; (4) the degree of development of the motor 
cells, Thus his Group II. includes those grasses in which the leaves have the upper and 
lower surfaces not parallel, have a greater number of stomata on the upper face, have 
normal green parenchyma (7. e., without large lacunze and not arranged in rings around 
the bundles), and motor cells well developed. He claims that these groups include only 
grasses of similar biological characteristics; thus he says: “ It [?. e., the classification] 
will give an exact idea of their habitual situation, of their vegetative period, and to a 
certain extent of the soil which they require." This statement is, on the whole, correct, 
but, as will be seen later, it has numerous exceptions. 
Raunkiær (1895-99) * in his account of the natural history of Graminez takes into 
consideration the anatomy of the leaves, and under each genus mentions the chief 
peculiarities of the leaf-structure. He also devotes a section to a connected account 
of * Leaf-formation," discussing the structure of the leaves from the point of view of 
the adaptations shown in relation to the environment, and more partieularly to the 
conditions of water-supply and transpiration. He points out, however, and I think 
with justice, that some of the structural peculiarities of the leaves are not to be 
explained as being due to the influence of the environment. He says, for instance: 
“Naturally, we cannot explain all constructions directly by a consideration of the 
outer factors, though many can be explained, or at least understood, in this way. 
Others, which are of a more primitive phylogenetic nature, are outside the laws of 
adaptation.” He gives a number of diagrammatic figures of transverse sections of 
the leaves. 
Professor Marshall Ward (1901), in his recent book on * Grasses,’ devotes two chapters 
to the leaf-structure. In Chapter IV. he gives a fairly complete, though necessarily 
somewhat short, account of the chief anatomieal and histological features of the leaves 
* I am indebted to Professor Bayley Balfour for a translation of those parts of Raunkisr's work which bear on 
the present research. 
