“The operation of cutting back a branch or trunk to receive the graft always promotes the growth of numerous buds. What is to be done with the shoots that spring from these? Evidently they must be suppressed, for they would appropriate, to no good end, the sap intended for the graft. Nevertheless the suppression must be done cautiously. Let us not forget that what primarily causes the sap to ascend is the evaporation of moisture from the leaves. As long as the graft has not opened its buds and spread its leaves, it is well to let the young shoots of the stock remain untouched. They act as helpers, in that their foliage draws upward the juices extracted from the soil by the roots; so that, far from having an injurious effect at this time, their presence is most useful. But the day will come when the graft alone will suffice for this work of pumping up the sap, and then it is best to get rid of these messmates which, of heartier appetite than the graft, would soon starve it out. First the lower shoots of the stock are suppressed, then gradually those higher up, care being taken not to destroy the top ones until the graft has developed shoots two or three decimeters long.” CHAPTER XXXIX GRAFTING (Concluded) “The part of a plant or tree above ground and the part under ground are mutually dependent, the development of one implying a corresponding development in the other. If there is a superabundance of foliage, the roots will be unable to furnish it sufficient nourishment; on the other hand, if the roots are unduly vigorous, there will be too much sap for the foliage—an excess of nourishment which, there being no use for it, will encumber the plant and be injurious to it. Hence if the trunk to be grafted is strong it must have several grafts, in order that the number of buds to be nourished may be in right proportion to the number of nourishing roots. “To this end the trunk is cut, not obliquely as for a single graft, but horizontally. Then it is split all the way across, following a line that passes through the central pith, and two grafts are implanted in the cleft, one at each end. It is evident that not more than two can be placed in the same cleft, because the bark of the graft must of necessity come in contact with the bark of the stock to insure inter-communication and coalescence between the sap-canals of the two. If the size of the stock requires more than two grafts, instead of splitting the trunk diametrically several times, it is preferable to make lateral clefts which, leaving the center untouched, cause less danger to the solidity of the stock.