by growth, the remainder of the top may, by successive excisions, be entirely removed." [Illustration: Fig. 16.--RENEWAL OF THE TOP OF AN OLD TREE.] GRAFTING IN THE NURSERY is either done at or near the collar of the stock, or it is performed in-doors upon the roots or sections of roots of young stocks. The latter may be first described, as it constitutes the most extensive means of multiplying fruit trees. It is a sort of machinery, with division of labor, and appliances, that enable the operators to turn out immense numbers. Machinery has indeed been applied to the business; we have grafting apparatus to facilitate the work. The Minkler machine consists of a frame or gauge which regulates the angle of the slope, which is cut with a broad chisel that reduces the roots and scions to a condition for putting them together; by its use an immense number of grafts can be cut, and another hand binds them together with the waxed thread, without any tie. Mr. Robey's machine consists of a complicated shears to cut the slope and tongue at one operation, preparing the pieces for whip grafting. Mr. S.S. Jackson, of Cincinnati, has also invented an apparatus for this purpose, which proves to be very useful. ROOT GRAFTING.--The methods of performing the operation vary somewhat, but all agree in the object to be attained: the co-aptation of the scion with a piece of root. Some grafters use only the upper portion of the root, thinking the original collar of the seedling stock the only point at which the most perfect and successful union between the aërial and terrestrial portions of trees should or can be effected--theoretically this may be very well, but the practice constantly pursued, in myriads of cases, abundantly proves that the grafting need not be restricted to this part, and that a perfect union may be effected at any point of the root, and that this may even be inverted. The very common practice has been to take two or more cuts from the root, when it is of sufficient size and length; and though some of our best propagators restrict themselves to two cuts from each, others, who have experimented carefully, insist that the third section will average as well as the others. A lot of trees, worked especially for a test in this matter, gave the following results. In 1859 an average lot of roots and scions, about fifty in each lot, were treated as follows, White Pippin and Willow-leaf being used as scions:-- White Pippin--No. 1, being on the first cut of the root, had made a fair growth. No. 2, being on the second cut, were quite as good or better. No. 3, being on the third cut, were not quite so good as the others, the ground being partially shaded by a large tree. Another, of Willow-leaf--No. 1, on the 3d cut of root, very good growth. No. 2, on very slender roots, nearly as good. No. 3, only 1 inch of root to 1 inch of scion; not so good growth nor so good a strike, but shaded by a tree.