Many aquatic plants have two kinds of leaves; some more or less rounded, which float on the surface; and others cut up into narrow segments, which remain below. The latter thus present a greater extent of surface. In air such leaves would be unable even to support their own weight, much less to resist the force of the wind. In still air, however, for the same reason, finely-divided leaves may be an advantage, while in exposed positions compact and entire leaves are more suitable. Hence herbaceous plants tend to have divided, bushes and trees entire, leaves. There are many cases when even in the same family low and herb-like species have finely-cut leaves, while in shrubby or ligneous ones they more or less resemble those of the Laurel or Beech. These considerations affect trees more than herbs, because trees stand more alone, while herbaceous plants are more affected by surrounding plants. Upright leaves tend to be narrow, as in the case of grasses; horizontal leaves, on the contrary, wider. Large leaves are more or less broken up into leaflets, as in the Ash, Mountain-Ash, Horse-Chestnut, etc. The forms of leaves depend also much on the manner in which they are packed into the buds. The leaves of our English trees, as I have already said, are so arranged as to secure the maximum of light; in very hot countries the reverse is the case. Hence, in Australia, for instance, the leaves are arranged not horizontally, but vertically, so as to present, not their surfaces, but their edges, to the sun. One English plant, a species of lettuce, has the same habit. This consideration has led also to other changes. In many species the leaves are arranged directly under, so as to shelter, one another. The Australian species of Acacia have lost their true leaves, and the parts which in them we generally call leaves are in reality vertically-flattened leaf stalks. In other cases the stem itself is green, and to some extent replaces the leaves. In our common Broom we see an approach to this, and the same feature is more marked in Cactus. Or the leaves become fleshy, thus offering, in proportion to their volume, a smaller surface for evaporation. Of this the Stonecrops, Mesembryanthemum, etc., are familiar instances. Other modes of checking transpiration and thus adapting plants to dry situations are by the development of hairs, by the formation of chalky excretions, by the sap becoming saline or viscid, by the leaf becoming more or less rolled up, or protected by a covering of varnish.