The anatomy of the common forms of rocks in any particular district is soon learned. The sketcher has seldom to deal with more than two or three kinds, and will quickly come to recognise their characteristic shapes, whether splintered like slate, bedded like limestone, or contorted like granite, and the metamorphic rocks. On the other hand, nothing is more individual than the shape and colours of particular rocks, so that the sketcher will learn for himself all that he could be told beforehand of the rocks in his neighbourhood, and will always be discovering facts about them of which no one could have forewarned him.

With regard to trees it is different. There will be a great many species within a short circuit, and it is often more important to distinguish the kind than the individual. Foreground studies of the trunk, branches, and leafage of each important tree should be made until the sketcher knows how to recognise its characteristics at any distance. He will then perceive at once the distinguishing curves, and the accidents that break and vary them will be full of meaning for him. He will see that the curve of the maple differs from that of the oak or willow or elm, and that when the normal curve is departed from it indicates peculiar circumstances, or, perhaps, some accident that has happened to the tree. Such facts, quickly apprehended and noted down, greatly add to the interest of a sketch. But trees grown in masses, in groves or plantations, lose much of their characteristic forms, and the part of each that comes into view falls into some large, sweeping line. It is, then, of importance, while following with the brush these leading lines, to indicate with each touch one of the many slight departures from them which give them variety and life.

Herbage of all sorts, except in the foreground, it is generally safe to paint exactly as bare earth would be painted, that is, by values, with strict attention to relief, perspective, and the changes of local colour. One will find that the character of many kinds of vegetation may be given in this way without any drawing, or even indication, of indi-vidual plants. But in foregrounds it is seldom possible to simplify things so much. There not only should care be taken to indicate by some sort of handling the general character of the vegetation, whether poor or luxuriant, uniform or varied, but this will be found insufficient without considerable drawing of individual plants. If one has not made a good many careful preliminary studies, the more interesting a foreground may be the more should he refuse to have anything to do with it. For the effort that will be required to draw in sufficient detail to make a satisfactory foreground will use up all the sketcher's time, and his patience and interest in the subject as well; the effect which attracted him will vanish, meanwhile, and he will gain by an afternoon's work only a poor and unreliable foreground study, when he might have made a thorough one in half the time.

In order to be of use in painting comprehensive landscape views one should make a mental classification of the varieties of " foreground stuff " that he oftenest meets with, and, after making close studies, should attempt rapid but correct sketching of each variety. It may be sufficient for the purposes of the landscapist to divide all such vegetation (distinct from bushes and trees) into the four classes of grasses, vines, large-leaved plants, like the dock and most cultivated roots, and plants of upright habit, like thistles, wild carrot, and hawkweed. But it will be better to go a little further and learn enough botany to distinguish, by their general characteristics, not by their flowers, the following botanical classes: - Ranunculaceae, or Buttercup family; Compositae, or Aster and Thistle family; Liliaceae, or Lily family; Cruciferae, or Kale family, and, perhaps, a few more. It is of little use studying the botany of forest trees, because their relations as given by the botanists are often either very obscure or else obvious to everybody. No one will confound an oak with a pine, or fail to see the similarity of the fir to the hemlock. The landscapist may content himself with making studies, under different aspects, of the species and varieties that he recognises as most common or most picturesque.

Most of these studies may best be begun in winter when the branches, being bare of leaves, may be traced from trunk to spray. The ramification of coniferous trees, which do not lose all their leaves, is so regular and striking that it can be perceived at any season. Pines are particularly worthy of study. Of the others, the oak is very irregular, and has clumsy, knotted blanches; the beech shows a strong tendency to spread horizontally in its lower portion, while the main branches grow upward at an angle from the stem more acute as they are farther from the ground, and carve like a whip-lash. The birch and several other trees have nearly upright branches and drooping twigs, while the poplar and aspen show an upright tendency in both branches and twigs. In summer, the disposition of the masses of Foliage follows that of the branches, but with differences occasioned by sparseness or luxuriance of growth as much as by the form and size of the leaf. It results from this that individuals of the same species and variety will look very different from one another, and the same individual, even, will present different forms at different seasons.