BOOK IV. PROPAGATION BY GRAFTING. 395 



dulla appears in the section of the stock, and ends as much above it, upon the opposite 

 side. The scion, which should not exceed in diameter half that of the stock, is then to be 

 divided longitudinally, about two inches upwards from its lower end, into two unequal 

 divisions, by passing the knife upwards just in contact with one side of the medulla. 

 The stronger division of the scion is then to be pared thin at its lower extremity, and in- 

 troduced, as in crown-grafting, between the bark and wood of the stock ; and the more 

 slender division is fitted to the stock upon the opposite side. The scion consequently 

 stands astride the stock, to which it attaches itself firmly upon each side, and which it 

 covers completely in a single season. Grafts of the apple and pear rarely ever fail in 

 this method of grafting, which may be practised with equal success with young wood in 

 July, as soon as that has become moderately firm and mature. 381 



2034. A subvariety o/ saddle-grafting (Jig. 381.), applicable 

 to very slender shoots, is practised by Knight, who gives the 

 rationale and manipulation in his usual masterly manner. As 

 this mode has rarely " or never been properly executed, it will 

 be necessary that I describe the motion of the sap as I conceive 

 it to be, at the period when grafts are most advantageously in- 

 serted. The graft first begins its efforts to unite itself to the 

 stock just at the period when the formation of a new internal 

 layer of bark commences in the spring ; and the fluid, which 

 generates this layer of bark, and which also feeds the inserted 

 graft, radiates in every direction from the vicinity of the me- 

 dulla, to the external surface of the alburnum. The graft is of 

 course most advantageously placed when it presents the largest 

 surface to receive such fluid, and when the fluid itself is made to 

 deviate least from its natural course. This takes place most 

 efficiently, when a graft of nearly equal size with the stock is 

 divided at its base and made to stand astride the stock, and 

 when the two divisions of the graft are pared extremely thin, at 

 and near their lower extremities, so that they may be brought 

 into close contact with the stock (from which but little bark or 

 wood should be pared off) by the ligature. I have adopted this 

 mode chiefly in grafting cherry-trees, and I have rarely ever seen 



a graft fail, even where the wood has been so succulent and immature as to preclude 

 every hope of success by any other mode." (Hort. Trans, v. 147.) 



2035. Shoulder, or chink-grafting, is performed with a shoulder, and sometimes also 

 with a stay at the bottom of the slope. It is chiefly used for ornamental trees, where the 

 scion and stock are of the same size (Jig. 380, a, b, c, d). 



2036. Root-grafting (Jig. 380. ti) is sometimes performed in nurseries on parts of 

 the roots of removed trees, when the proper stocks are scarce ; and in which case, the 

 root of the white thorn has been resorted to as a stock both for the apple and pear. In 

 general, however, a piece of the root of the tree of the same genus is selected, well fur- 

 nished with fibres, and a scion placed on it in any of the ordinary ways for small stocks. 

 Thus united, they are planted so deep as to cover the ball of clay, and leave only a few 

 eyes of the scion above ground. Some gardeners have thought, that in this way, the 

 plant must preserve a near resemblance to the parent tree ; but Abercrombie remarks, 

 that though it is an expeditious way of obtaining a new plant, such a graft cannot be 

 materially different from a cutting or layer. 



2037. A variety of root-grafting, practised by Knight, is thus described. " Trans- 

 planting, many years ago, some pear-stocks from a seed-bed, of which the soil was soft 

 and deep, I found that the first emitted roots of many of them descended a foot or more 

 perpendicularly into the earth, before they divided into any lateral ramifications : and as 

 I did not like to replant the young trees, with such an inconvenient length of perpendi- 

 cular root, I cut off about six inches from each. The amputated parts were then accu- 

 rately fitted and bound, as in splice or whip-grafting, to scions of pear-trees, which were 

 selected as nearly as possible of the same size ; and the roots, with their attached branches, 

 were deposited in the ground as cuttings, so deep, that the whole of the root, and about 

 an inch of the scion, were covered. The soil was then drawn up with the hoe on each 

 side of the plants, which were placed in rows, so that one bud only of each graft was 

 above the soil, and another just within it. These grafts succeeded perfectly well ; and I 

 have subsequently repeated the same experiment with equal success upon the apple, the 

 plum, and the peach. In the greater part of these experiments, the roots were perfectly 

 cleansed from mould by washing, before they were fitted to the graft, and were then 

 placed in wet moss, till a sufficient number were ready to be carried to the nursery ; a 

 common dibber only was employed in planting them ; but the mould was washed into 

 the holes with water, to close it well round the roots, and to supply the place of the clay 

 used in other methods of grafting." (Hort. Trans, vol. i. p. 239.) A variation of this 



