On Training Fruit Trees. 
485 
training for the top, as the standards, or technically riders, being 
planted as the permanent trees, the dwarfs can be planted between 
them in the usual way; but instead of their being the permanent, they 
become the temporary ones, and are to be cut away or removed as 
the ‘^‘inverted fan” requires room. 
Figure 84, a, is the representation of a tree upon the old fan form, 
in which Pears are usually trained. I have some of them, but am 
reducing them to the equi-distant pendant form, by training the top 
of the most central branch (as at a') horizontally each way under the 
coping, and to furnish shoots to bring regularly downwards as they 
progressively go on. Some of them are of ordinary kinds, conse¬ 
quently I inverse-graft them, as represented in the figure, and that I 
may neither lose much time, nor walling, whilst the top shoots are 
in training, I select the most central branches upon each side of the 
tree, between the top and the bottom of the wall, as at h b, and insert 
inverse-grafts of choice varieties of Pears; and as they are annoyed 
by the old branches, I reduce the spurs, and eventually cut them 
entirely away, as is represented by one side of the figure, where they 
are only shown in skeleton. When the upper pendent leaders become 
too much crowded, by being intermixed with the grafts underneath, I 
lastly cut them out with the branch into which they were inserted, 
close to the perpendicular stem of the tree, when they will soon have 
the appearance of figure 80. 
Figure 84, li, is the representation of a tree on the old horizontal 
form, which I am changing into better sorts, by the inverse method of 
grafting, and have made a selection of the top and central horizontals 
for the purpose of grafting; so by that means, and the admirable 
method of inverse pendant grafting, I can see no plan so well fitted 
for an expeditious, choice, and fruitful renovation. I cut off all the 
spurs about Christmas, from such shoots as I select for grafting; 
and insert the grafts when the sap will admit of the bark rising 
freely. I reduce the old horizontals as the shoots from the grafts 
require room, as upon figure 84, c. The two sets of shoots from 
the top and central horizontals, may remain, if thought proper, and 
be shortened from crowding upon those beneath, for they can be 
kept in an orderly and fructiferous state by letting the su})er-abundant 
