metry of a tree. It however by no means follows that, because out of two contiguous branches, one growing erect and the other forced into a downward direction, the latter may die, that all branches trained downwards will die. On the contrary, an inversion of their natural position is of so little consequence to their healthiness, that no effect seems in general to be produced beyond that of causing a slow circulation, and the formation of flowers. Hence the directing of branches downwards is one of the commonest and most successful contrivances employed by gardeners to render plants fruitful. Mr. Knight was the first to recommend the practice, in the following account of his recovery of an old and worthless Pear tree.

"An old St. Germain Pear tree, of the spurious kind, had been trained in the fan form, against a north-west wall in my garden, and the central branches, as usually happens in old trees thus trained, had long reached the top of the wall, and had become wholly unproductive. The other branches afforded but very little fruit, and that never acquiring maturity was consequently of no value; so that it was necessary to change the variety, as well as to render the tree productive. To attain these purposes, every branch which did not want at least twenty degrees of being perpendicular was taken out at its base; and the spurs upon every other branch, which I intended to re

T

tain, were taken off closely with the saw and chisel. Into these branches, at their subdivisions, grafts were inserted at different distances from the root, and some so near the extremities of the branches, that the tree extended as widely in the autumn after it was grafted, as it did in the preceding year. The grafts were also so disposed, that every part of the space the tree previously covered was equally well supplied with young wood.

"As soon, in the succeeding summer, as the young shoots had attained sufficient length, they were trained almost perpendicularly downwards, between the larger branches and the wall, to which they were nailed. The most perpendicular remaining branch upon each side was grafted about four feet below the top of the wall, which is twelve feet high; and the young shoots, which the grafts upon these afforded, were trained inwards, and bent down to occupy the space from which the old central branches had been taken away; and therefore very little vacant space remained any where in the end of the first autumn. A few blossoms, but not any fruit, were produced by several of the grafts in the succeeding spring; but in the following year, and subsequently, I have had abundant crops, equally dispersed over every part of the tree; and I have scarcely ever seen such an exuberance of blossom as this tree presents in the present spring." (Hort. Trans., ii. 78.)

The practice was then followed by Sir Joseph

Banks, whose fruit trees trained downwards over the walls of his garden at Spring Grove, and facing the high road, long excited the astonishment of passers by; and it has now been generally applied to other cases. What are called Balloon Apples and Pears, formed by forcing downwards all the branches of standard trees till the points touch the earth, are an instance of this; and they have the merit of producing large crops of fruit in a very small compass: their upper parts are, however, too much exposed to radiation at night, and the crop from that part of the branches is apt to be cut off. One of the prettiest applications of this

principle is that of Mr. Charles Lawrence, described in the Gardener's Magazine, viii. 680., by means of which standard Rose trees are converted into masses of flowers. The figure given in that work, and here reproduced (fig. 35.), represents the variety called the Bizarre de la Chine, "which flowered most abundantly to the ends of its branches, and was truly a splendid object."

The last object of training to which it is necessary to advert is that of improving the quality of fruit, by compelling the sap to travel to a very considerable distance. The earliest notice of this, with which I am acquainted, is the following by Mr. Williams of Pitmaston.

"Within a few years past," he says in 1818, "I have gradually trained bearing branches of a small Black Cluster Grape, to the distance of near fifty feet from the root, and I find the branches every year grow larger, and ripen earlier as the shoots continue to advance. According to Mr. Knight's theory of the circulation of the sap, the ascending sap must necessarily become enriched by the nutritious particles it meets with in its progress through the vessels of the alburnum; the wood at the top of tall trees, therefore, becomes shortjointed and full of blossom buds, and the fruit there situated attains its greatest perfection. Hence we find Pine and Fir trees loaded with the finest cones on the top boughs; the largest acorns grow on the terminal branches of the Oak, and the

finest mast on the high boughs of the Beech and Chestnut; so likewise apples, pears, cherries, &c., are always best flavoured from the top of the tree." (Hort. Trans., iii. 250, 251.) The merit of the Fontainbleau mode of training the Vine (fig. 36.), in which many of the stems are carried to very

considerable distances, seems to depend in some measure upon this principle; and there is a wellknown Black Hamburg Grape at Bath, growing in a garden formerly belonging to Mr. Farrant, the stem of which, owing to local circumstances, is necessarily conveyed to a very considerable distance before it is allowed to produce its bearing