The annual summer shoots of this are of yellowish green, of the former variety of a reddish brown. Suckers are as abundant with the Dutch as with the French variety.

In France this Dutch variety is called Doucin;* also Doucain and Fichet They are used principally for half-standards.+ For in rich ground this variety I grows nearly as rampant as the strongest seedling; with the difference, though, that they very early become fruitful. Fifteen apple trees grafted on Doucin, which I received from Harlem, all bore the third year. This is the Doucin, which the English botanist and scholar Miller does so much justice in saying that less than twenty feet distance on a trellis will not do for it.

The third variety seems to be the "Hedge apple" of several German authors; but as I have seen none of them, I can say nothing about it. Von Munchhausen thinks it is identical with the Doucin, but this is not probable.

In the French paradise apple we have an incomparable subject for apple trees in pots. Its roots, not rampant and wild, being all fibrous or feeding roots, may be checked in any degree; for after the laws of nature, a curbed power of root causes early fruitfulness and shortened life; while, on the contrary, a wild, rampant growth insures longevity and late fruitfulness. For this reason, all trees planted on the side of a mountain, or in good but shallow soil resting on rock, become fruitful very early.

If, now, paradise apple stocks are to be prepared for pot-trees according to the rules of art, the principal rule for them, as well as for all other kinds of stocks for all other kinds of fruit, is : To induce a large number of fine feeding roots to spring from a short piece of the main root or neck. As the length of the main root can never be more than three or four inches for a pot eight or nine inches in depth, it happens often, that the stocks have but few roots on this length, which must, therefore, be produced by art, namely, by transplanting; for the more feeding roots we have on the short piece of main root, the more pleasure shall we derive from the merry growth of our trees.

Young stocks or suckers, not over half an inch thick, smooth and free from any hard crust, must, therefore, be chosen. Of these the main root is cut back to two or three inches; the small roots remaining over the cut are also cut back to one inch or one inch and a half; and now, when the top has also been cut back, the tree is planted in the nursery, one inch deeper than it stood before, to prevent it suffering from drought in the summer. In planting care must be taken that these fine roots are well spread horizontally, for, otherwise, these fine beard-roots are apt to cling to the main root, which often causes their decay. Crippled stocks with injured bark never succeed.

* See Du Hamel, Pomona Gellica, vol. ii., p. 43.

+ See Nouvelle maison rustique, vol ii., p. 149. Paris, 1721, (a work of two large quarto volumes).

+ For dwarf trees in the garden, or trellis, or as pyramids, the contrary holds good. Here it is necessary to have a forest of feeding roots to sustain the larger tree. For trellis, at least one foot of main root is required, and for pyramids, where the graft must absolutely come in the ground, if they shall be able to resist the winds, 18 inches are not too much.

When such a little tree has stood one summer in the nursery, and grown well, and it had already good roots when it was planted there, it may be budded the same summer or grafted the next spring. But if it had no fine roots when it was planted, it must be taken up and replanted next spring, when the new roots are also cut back again. In this case, a tree is to be expected that will have, as it were, a forest of feeding roots, and with them the properties necessary to make it a perfect pot-tree; for each cut root will throw out numerous smaller ones.

But often this replanting is not necessary, and many paradise suckers are found which have sufficient roots for our purpose.

If, however, our conjecture in this respect has proved erroneous, and the trees already grafted, are found to have too few roots in taking up, they must be replanted in the nursery, according to the above rules for stocks. In this case we not only save a year, but, for obvious reasons, the tree will sooner become fruitful. I transplant my trees intended for pots, every spring, in March, by which means they soon set many fruit buds; but of this in the chapter of vegetation.

But never must an amateur, who expects undisturbed pleasure for many years from his pot-trees, plant Paradise stocks, much less any other kind of stocks, in pots, to be grafted in them. Instead of gaining hereby, there is much loss; for, not calculating that the tree, grafted the same year, does not thrive so well, often very poorly, it exhausts the soil in the pot at least three years too soon, unnecessarily and without any compensation. The pot becomes too soon cramped with roots; and when, in the third or fourth year, the tree has formed a crown and fruit buds, it also often begins already to lose in vigor. It must, therefore, be transplanted, and, at least for one year, all pleasure and enjoyment are lost. Bat if the tree has first been two years in the nursery, and been carefully transplanted there every spring, I get a tree with a crown full of fruit buds, which will yield the finest fruit for five or six years.

It is therefore, a general rule to plant the stocks first in the nursery, to graft or bud them there, and then to leave them for two years, either undisturbed or transplanted every spring, for the development of their roots and crowns. The first year's growth of the graft is to be properly pruned back the next spring, by which we obtain in the second summer the proper branches for a crown, which we obtainneither so numerous nor strong in a pot. In the fall the tree may be potted, but better not till next spring, because we then save several months of trouble. But never must a tree that has been potted in the fall, nor any other newly planted fruit tree, be pruned at that season. The slow circulation of the sap is better sustained by the branches, and one runs no risk of losing just that bud by frost which by the next season's growth was destined to form a new branch.

After the second season's growth in the nursery, the little tree is generally well developed, and after potting often bears fruit the same year, though oftener only blossoms, which fall off again, because the tree is not yet well rooted in the pot. This was the general rule. But the case is widely different with the student, with whom pot culture is not the end, but the means. If he, solely for the study of varieties, wants to experiment with a new or unknown variety; if it is enough to see the fruit once or twice, and observe the manner of vegetation of the tree, to enable him to form a judgment and make comparisons, then two years saved are important. In this case the grafted trees may be potted the first year, or some Paradise stocks may be potted yearly in advance, to be grafted in the pot. Nay, if this has been done in the fall, they may be grafted the following spring, or even grafted and potted at the same time. In this case the regular shape of the tree, which can only be obtained in the nursery, is of no consequence; but instead it may be tried to develop the future fruit buds even the first season, by pinching in the young summer shoots.

He who has observed and meditated on the advantages and disadvantages of the vine fretter or vine pricker, (Curculiones longinoster, long-billed snout bugs,) will easily understand this artifice. These are mostly busy before July. After they have pricked off a vine, the next bud generally pushes out in a week, and forms a summer shoot. But not seldom, and with stone fruit and pears nearly always, not only the next, but several buds come into activity, which otherwise would have pushed only next year, or perhaps never.

If, now, a summer shoot of a healthy growing tree is pinched in before July, at about that bud where it would have had to be pruned next spring, we obtain fruit spurs one year earlier, which often blossom and bear fruit the next year.

In regard to grafting, I will only add, that this may be done as high up as where the crown is to form, and the best height for this is from 10 to 12 inches. For dwarf trees for the garden, on the contrary, the paradise stock must be grafted very near the ground.

If the paradise stock has been potted only the preceding fall, or only in the spring when it is grafted, the graft only makes one or two weak shoots. But if it has grown in the pot for one year previous to grafting, a bud will make a strong shoot, and a graft two or three. These shoots are permitted to grow till the middle of June, when they are pinched in one half. They will now make enough side shoots to form a crown the same year, and apples and pears will even often make fruit buds.

Summer pinching is for dwarf trees as important as pruning in spring, and is based on the same principles. Its purpose is to give trees the proper shape; to render the pruning easier; and to strengthen the necessary shoots.

[At this time, when the culture of fruit trees in pots is taking its place as a specialty, the above remarks from Diel will be read with much interest. Though written more than seventy years ago, they still have a degree of freshness, and agree mainly with the best practice of the present day. That some progress should have been made in this long interval was to have been expected. As we expect to give further extracts, our remarks may as well be reserved for the end. - Ed].