578. _Harmony between masses distant from each other_, in which harmonies may be established by the same general means as they are between neighbouring masses. The foliage in the masses being much more abundant than the flowers, the shades of their greens will not differ so much from each other as the colours of their flowers; consequently, the distant masses, however varied in their foliage, are always in harmony of form and colour, if they have been planted according to our rules; and if they are composed solely of ligneous plants, which lose their leaves in winter, or solely of evergreens, which do not lose them; but in the opposite case, that is to say, when the masses are formed, the one of deciduous trees, and the other of evergreens, some remarks must be made relative to the conditions of harmony, which are the more necessary, as even the most studied landscape gardens generally err in this respect; for a clump or a mass of evergreens is almost always out of keeping in a large space where masses of deciduous trees are found. To remedy this defect we must multiply the groups or masses of evergreens, so as to establish between all of them this same correlation which is required by deciduous trees; but it is not necessary that the trees should occupy a space equal to that occupied by ordinary trees; it is sufficient if their forms recur at suitable intervals. In a word, for evergreens to produce a good effect, they must compose a whole, which unites or intercalates with the general effect of the masses of deciduous trees. We may oppose allspice trees to pines, cedars to larches; different groups, composed of three or four trees only, suffice to harmonize a large space of ground occupied with two or three groups composed of half a hundred similar trees. 579. _There are certain cases in which the want of either perspective or harmony requires, in a large mass, a line of trees which is neither concentric with its circumference, nor identical with the central line, if there be one; such, for example, as the line P. R., which is found in the masses 1 and 2_ (Pl. 18); _for if the planting of this line be correct, and according to the preceding principles, the trunks, which define it, beyond the concentric plantings, must be in the points of intersection of the line P. R., with the concentric lines, and the central line, if there be one; and the trunks must be as much as possible at equal distances from each other._ Such a plan, always easy to make, will enable a landowner, when once his masses are planned, their concentric lines traced, and the species to be planted determined upon, to order from the nurseryman the exact number of each species he requires.