38 FRUIT GARDEN. 
Preliminary Operations 
may be classed under the heads Propagation, Planting, 
Training, and Protection of Blossom. 
Propagation by Seed.—Although fruit-trees are fur- 
nished with all the natural means of reproduction, it is not 
in general expedient to attempt to propagate them by the 
sowing of seed. This method is found to be equally tedious 
and precarious, requiring the labor of a good many years, 
and very rarely producing an exact copy of the fruits from 
which the seeds are taken. The chief reason of the varia- 
tion is pretty obvious; the blossoms of different varieties 
of the same species of fruit are commonly expanded, at the 
same period of time, in the neighborhood of each other, and 
the pollen of one kind is thus extremely apt to be trans- 
ferred, by the agency of bees and other insects, to the 
stigma of another kind. If, therefore, we desire to pro- 
cure uncontaminated seed of an excellent variety, such as 
the Ribston apple, we ought to encircle the blossom-bud 
with a fine gauze bag, sufficiently wide to allow the blos- 
som to expand, and not remove the covering till the fruit 
be fairly set. Another source of variation is to be found 
in the influence of the stock upon the graft, which is real, 
though not easily detected, except in extreme cases (such 
as grafting Scotch apples upon stocks of the Russian trans- 
parent, and finding the former acquiring the transparent cha- 
racter). To obviate this the tree should stand on its own 
bottom, or be struck from a cutting. All our present ad- 
mired fruits are regarded as seminal varieties obtained 
from the wild inhabitants of the forests; they have been 
trained into an artificial condition, and when sown seem to 
have a tendency to resume their original constitution. In 
