910 FEUIT CULTURE IN IfOEEIGN COUNTEItS. 



to 12 feet liigh, depeDding on the size of the garden or inclosed space. 

 It should be covered with some sort of slightly projecting coping, as 

 tiles or slate, and should be of suflEicient thickness to insure its stability. 

 Here, where the whole soil is nearly underlaid with soft chalk stone, 

 the cost of the wall is a mere trifle— the trench for the foundation and 

 that for the trees fuinish nearly enough stone for the purpose. 



(2) After the wall is built some sort of trellis should be attached to 

 it upon which to tie and train the trees. The poorer gardeners employ 

 a plain wooden diamond-shaped trellis, made of slats nailed against 

 the wall. The walls of the best gardens, however, are provided with 

 trellises of iron wire, luid each strand is furnished hero and there with 

 thumbscrew swivels so that they may be drawn taut when required. 

 This is much the neater, jnore desirable, and, in the long run, more 

 economical system. 



(3) A trench should be dug at some slight distance from the wall for 

 the reception of the new plantation. The distance from the wall and 

 the dimensions of the trench will depend on the height of the wall as 

 regulating the height of the tree and the kind of fruit-tree to be planted, 

 say, ordinarily, IJ feet from the wall and 2 feet deep by 3 or even 2J 

 feet wide. This trench should now be filled with carefully prepared 

 upper soil, or loam, mixed with a suitable quantity of stable manure. 

 The tree, or vine, is now planted, in the month of ^November or Febru- 

 ary, with the stem slanted through the side of the trench and towards 

 the wall, and, its first bent being given it, attached to the trellis. 



Here let me stop a moment to say that the consensus of opinion 

 among gardeners here seemed tu be decidedly in favor of seedling trees 

 in preference to budded or grafted stocks. Of course, if a bough can 

 not be gotten where it is desired to have it by natural means, that is, 

 if a wood bud cau not be found at a point where a limb is required, 

 this want may be supplied by a bud artificially introduced, or by a graft. 

 Still, the French gardener prefers to deal with a seedling in the first 

 instance, holding that it is easier to give it the desired shape, and that 

 the buds for the production of lateral and other boughs may be more 

 certainly counted on than is the case with grafted or budded stocks. 



Having now arranged the preliminaries it becomes a question of what 

 shape shall be given to the trees. For my own part, I much prefer the 

 rather simpler forms, say the U forms and their modillcatious, or the 

 candelabra forms. The best of all the forms, howevei-, both as to beauty 

 and utility, I regard the palm-shape and its modifications. There are 

 other very good shapes, the names of which 1 have been unable to get. 

 My distinct preference, also, is for the forms which have vertical or ob- 

 lique branches, for these reasons : The sap of a tree has always a tend- 

 ency to mount, and under ordinary conditions the best fruit will be 

 found on the ends of the branches or the i^eriphery of the tree. There- 

 fore, the oblique forms follow more nearly nntrammeled nature, and 

 thus yield better general results. As a jtroof of the foregoing may be 



