If the new growth or shoot to be employed as a cion is slender and feeble, then the base of the cion may be of two-year-old wood, leaving just a bud or two on the upper end of the one-year shoot. But it will seldom be necessary to employ such cions in grafting the chestnut, although it may occur when seeking to secure wood for propagation, from very old trees which have made only a feeble annual growth. [Illustration: FIG. 15. STOCK.] [Illustration: FIG. 16. CION.] [Illustration: FIG. 17.] [Illustration: FIG. 18.] =Cleft Grafting.=--This method is employed principally upon stocks or branches of trees too large for splicing. The stock is first cut off at the point where it is desirable to insert the cion; then split with a knife, being careful to divide it, so that the edges will be kept smooth, and not rough and ragged (Fig. 15). When the knife blade is withdrawn, the cleft may be kept open with a hard wood wedge, if the stock is too large to admit of opening it with the point of the knife when ready to insert the cion. The cion may be three or four inches long, containing two or more buds; the lower end is cut wedge-shape, as shown in Fig. 16, and slightly the thickest on the side to be set against the bark of the stock. In stocks of an inch or more in diameter, two cions, one on each side, may be inserted (Fig. 17), and if both grow one should be cut away, else the tree, in later years, will be very likely to divide or break apart at this point. In stocks of an inch or less in diameter, one cion is sufficient, the top of the stock to be cut off with an upward slope, as shown in Fig. 18. After the cions are inserted, the entire exposed surface of the wood must be covered with grafting wax or waxed paper, and usually both may be employed with benefit. All the various forms of grafting in the open air, as described in my work on the "Propagation of Plants," may be employed on the chestnut, but the two here given will probably answer just as well as others for those who may have occasion to propagate this tree. =Success in Grafting.=--The question has been asked many times, and will, no doubt, be frequently repeated, "What percentage of cions should one accustomed to grafting make grow?" As there are no statistics upon which to base an answer to the question, I can only give my own personal experience, and this leads me to say that seventy-five per cent may be considered an excellent, if not a high average. In some seasons this has been exceeded by at least ten per cent, while in others it has fallen as much or more below, with no apparent reason for the difference. Ninety-five per cent of the cions may push their buds, or even make a growth of several inches, then begin to die off; consequently, the time to count your successfully grafted trees is in the autumn, and not in spring or midsummer, as it is to be feared some are in the habit of doing when making a report upon what they call success in grafting nut trees.