102 THE VEGETABLE WORLD. 
openings, so admirably organised for carrying out the vital 
phenomena of which it is the seat, phenomena which we shall 
study more minutely by-and-by. 
We find in the nervure of the leaves, vessels of divers orders 
bound together by cellules of various forms, which constitute the 
fibrous part of this bundle. The structure of the nervure, whic 
is most complex, becomes more simplified in proportion as it 18 
divided and diminished in its dimensions. 
If from plants with aerial leaves we pass to plants with leaves 
which float upon water, or whose natural position is to be sub- 
merged, we shall see the structure of these organs modifying 
themselves according to the medium in which nature has placed 
them. These curious modifications have been carefully studied 
in our own day by M. Adrien Brongniart. 
The leaves of the Water Lily (NWymphea), which float upon the : 
water, present, it is true, both epidermis and parenchyma, differing 
very slightly from those with aerial leaves, but the lower epidermis 
in contact with the water has no stomates. The submerged leaves 
of the Pond Weed are generally very thin and quite destitute of 
any epidermis, and consequently of stomates also. They are 
channelled with air-cells, which have no communication with each 
other, but are formed of polyhedrous or many-sided cellules, com- 
pressed and gorged with green fluid. These air-cells have 2° 
analogy with those of aerial leaves; they can only be considered 
as reservoirs of air furnished by the plant itself, and calculated, 2° 
doubt, to reduce its weight. They are floating apparatus, which 
seem to play a part analogous to the natatory vessels of the fish- 
The Lanunculus aquaticus present at once, as we have seen, the 
aerial leaves which float on: the surface of the water, and the much- 
divided leaves which are submerged. The aerial leaves, furnished 
with an epidermis provided with stomates, present a parenchyma oe 
whose structure scarcely differs from that of the aerial leaves 
already described. Aquatic leaves have no epidermis properly 5° 
called, but only isk } t stlul i agai st 
the other, tituting tl I nine y uniformly dense, hollowed 
here and there with isolated aeriform cavities, as in Fig. 128. 
_ We cannot conclude our study of leaves without sa + ial : 2 
# 
- 
