mands some little variation and play in the surface line; and a bank that is slightly undulated will look much less formal and unnatural.

In conformity with the shaping of the ground line or face of a bank, its retiring or narrowest parts should be the lowest, while the fullest parts are also the highest. And if, as is most proper and beautiful, a straight line drawn through the face of the bank, parallel with the back, or taking its general direction, would leave no two of the different swells or bays at exactly the same distance from it; so a similar horizontal line should show the like irregularity. Even two or three inches of difference in all the various hollows or elevations of surface will have a powerful effect in carrying off every kind of dulness, and imparting a graceful freedom. Should a bank be tolerably broad, there may be partial undulations across as well as along its surface in the wider parts. Then, the front swells should be made roundish, and fall away to hollows connected with the depressions in the narrower parts, rising again to a more flattened elevation towards the back. Or, for variety, the back part may be kept highest, and slope away into the same low as before, treating the projecting front as a flattened arm of the general mass.

What I have thus described, in terms which might lead to the supposition that considerable variations of surface were contemplated, refers simply to such as can be compassed on a comparatively small bank, and on a scale of inches in change of height rather than feet. Nothing need be less beautiful because it is small; and the rules of taste embrace the least as well as the greatest things. Nor will there be a seeming pretension about the adaptation of such rules as this, in reference to the smallest mounds.

Perhaps the most influential characteristic of an artificial bank is its being well tailed out into the ground, and by a decided under curve. There can be no resemblance to nature without this. It gives the very crowning stroke of finish and grace. But as this point has been more than once previously insisted on, it does not demand further pressing.

Much of the success of any efforts to vary and undulate banks of earth, will turn upon the way in which they are planted, and the turf is brought up their faces. The boldest swells require to

be as boldly planted; that is, with the tallest description of plants admissible. The smaller elevations and the hollows can be similarly treated; thus making the entire range a series of undulations on the surface of the plants, as well as that of the ground; the first corresponding in a great degree to the last. Along the fronts, also, the plants should come much lower down on the fuller parts, so as to increase the effect of their fulness; and any weeping specimens, or such as naturally send forward their branches in a more horizontal direction, should be placed here with the same object. In the hollows or bays, on the contrary, the planting should retire nearly to the upper surface of the mound, the turf, of course, following the line of planting, within a foot or two, in both cases. Grass may even be carried over the edge of the mound in some of the hollows, and so far across it as just to leave room for a few shrubs to cover the wall or fence that may happen to be behind. Or, if there be nothing to conceal, some of the lowest hollows may have a glade of grass carried entirely across them, which will greatly relieve and lighten a lengthy range.

According to the several purposes for which mounds are used, should be their ordinary treatment. If for covering boundary fences, they ought to be almost entirely planted with taller or dwarf things, and also be continuous. Where they are placed between parallel walks, to separate them from each other, if they are of any length, several of their hollows can be turfed through, leaving a low specimen plant or two on the grass, irregularly, in one or two of them. They may also have more of undulation than those of the former class. If a mound be made to furnish a good view of the garden, or a prospect of the neighbouring country, from its summit, great height should never be attempted in a small place; and it should, if possible, form part of a range, that it may not appear too conspicuous and unconnected. The breadth must constantly bear some proportion to the height, or it will seem glaringly artificial, and a mere conceit. Besides, it will be difficult to convey a walk to its summit, unless there is some breadth to wind around.

Such a mound as the last named may be partially planted with close tufts or clusters of shrubs, to cover the walk, and shut in some parts of the view. A few low trees, more sparingly dotted about, will contribute to give it stability and

P

character. The walk should, of course, be quite narrow, and may ascend by a zig-zag route on one side only, or by curving round the entire face of the mound. It might appropriately be composed, in its steeper parts, of easy flights of rustic steps.

3. Among the trees adapted to associate with different styles of buildings, there are three distinct classes, easily recognisable by the particular shape their heads and branches assume. The first and largest group produces roundish and clustering heads, when their full growth is attained. The Oak, the Ash, and the Elm are familiar examples. Another set, much more thinly scattered, send out their branches horizontally throughout their whole height. The Cedar of Lebanon, the varieties of Fir, (not Pine,) the common Yew less perfectly, the Larch, and the deciduous Cypress in its usual state, will illustrate this section; though the Scotch Fir, when quite old, is admirably flat-headed. The third tribe, which has very few members, consists of upright or fastigiate trees. The Lombardy Poplar is the commonest instance; though the upright Elm is another very good example. If such as have pointed or spiry heads be included, many of the second class will come within this also,-the Firs especially. Larch, and several round-headed trees, in their younger state, before the upper branches get dense and spreading, will give a pretty clear idea of spiry-topped trees.

Mr. Repton, in his "Sketches and Hints" on landscape gardening, lays it down as a general principle, that round-headed trees harmonise best with Gothic forms of architecture, and trees of spiry shape (fig. 155) with Grecian buildings; on the ground that the horizontal lines which prevail in the latter style, and the perpendicular in the former, are best exhibited and relieved by contrast with vegetable forms of an opposite character. Without questioning the soundness of the rule, which appears quite unexceptionable, it may be doubted whether, in the case of Grecian and Italian structures, at least, the appropriateness of the Fir and Cypress tribe is not the result of association; as the Cedar of Lebanon, the branches of which are purely horizontal, is the most magnificent of all accompaniments for any variety of Grecian architecture, but is not at all suited for either of the forms of Gothic. And so, perhaps, the old ancestral Elms and Oaks, in which many an English Gothic house is often embosomed, (fig. 156,) may, by the commonness and antiquity

of the usage, have given a propriety to the employment of that kind of tree in relation to all similar edifices.

Still, if it be admitted that certain descriptions of trees throw

Fig. 156.

out and accord with the forms of certain architectural styles, it may, for aught that appears to the contrary, be fairly assumed that the use of those trees in such situations had its origin in their fitness, or supposed fitness, for the purpose; and that, though they may be peculiar to any given country in which the style of building to which they are now allied preponderates, or has at some former period prevailed, that very style may have been founded on its adaptation to the natural characteristics of the country, trees not being among the least significant of these. Possibly I may not be far wrong in accounting for the connexion which has, somehow, sprung up between particular classes of trees and particular styles of buildings, by a reference to the character of the leaves rather than the distinctive lines of the branches or heads. Light, thin, and feathery leaves characterise all the plants that look best when in contact with the varied class of Grecian structures,-the heaviness (or rather massiveness) and regularity of Grecian forms demanding some such contrast and mitigation. Gothic buildings, on the other hand, already light and playful, full of variety, and abounding in small decorations, require more of the depth and breadth of foliage for which round-headed trees are conspicuous, to bring out their elegance, and impart, at the same time, a more substantial character.

Practically applying the subject, however, the very upright and the very horizontal forms of trees, such as have spiry heads, and those of which the upper branches are nearly flat,-may occupy any desired position in the neighbourhood of Grecian, Roman, or Italian houses; while shrubs of similar (but chiefly upright) habits, and such as are pruned or trained into standards, with formal heads, will suit gardens laid out in any kindred style. In addition to those which have been named, I may note the Deodar Cedar, the Araucaria imbricata, and the Hemlock Spruce, with the Cypress tribe, and especially the Irish Yew, for gardens, as deserving of particular praise. Larch, Birch, Acacias, the Purple Beech, (though principally for its colour,) the Fernleaved Beech, the Turkey Oak, and the Lime, will further be suitable for the same style of erection.

On the principle above suggested, besides the numerous species of round-headed trees, of which the Sycamore for westerly districts may be specially signalled out, and the Horse Chestnut and Spanish Chestnut for more inland counties, all the broader-leaved