“Grafting by approach is analogous to layering, with this difference, that the tree to be grafted takes the place of the soil that receives the layer. In layering we induce the growth of adventitious roots by partly burying in the ground a branch or shoot still adhering to the stock that nourishes it. When, acted upon by the soil, roots have started in sufficient number, the shoot is gradually cut loose until at last it is quite severed from the parent stock. In grafting by approach it is also proposed to make a branch, a shoot, a tree-top, while still united to its own stem or stock, take root, so to speak, not in the ground, but in the substance of a neighboring tree. “Let us suppose that two shrubs are growing close together and that we wish to engraft on one of them a twig or shoot of the other. The parts to be placed in contact receive each a longitudinal gash that penetrates to the marrow, or even deeper, and the two gashes are made of equal length. These parts are then brought together, care being taken to make the young and growing portions in the one exactly meet those in the other; that is to say, the inner layer of bark in each, with the channel traversed by the elaborated sap, is carefully fitted to the corresponding part in its neighbor. The whole is thereupon made fast with a ligature, and the two wounds are left to the slow operation of vital forces. Fed by its own stem or trunk, from which it is not yet separated, the shoot to be transplanted mingles its sap with the sap of its neighbor; on both sides there are new growths to cicatrize the wounds, while the two parts gradually coalesce until, sooner or later, the graft becomes incorporated with its future support. And now the graft must be weaned; that is, it must, little by little, be deprived of the sustenance furnished by its own stock. This is accomplished as in simple layering, by gradually cutting through the shoot below the point of union. As soon as the graft is thought to be getting all its nourishment from the new stem, it is completely severed from the mother tree. This mode of grafting, the most elementary of all, sometimes takes place accidentally and unassisted. In a hedge or any dense growth of bushes, if two branches chance to come into close and prolonged contact, there will be at this point, first, a slight abrasion and then a complete wearing away of the bark until the two raw surfaces end, it may be, in growing together. It is not improbable that natural occurrences of this kind furnished man with his first notions of grafting.