s multiplication,
as well as that produced by roots and layers, seems merely to operate as
an extension of the life of an individual, and not as a reproduction of
the species such as happens by seed. All plants increased by grafts or
layers retain precisely the peculiar qualities of the individual to
which they owe their origin, and, like an individual, they have only a
determinate existence; in some cases longer, and in others shorter.[812]
It seems now admitted by horticulturists, that none of our garden
varieties of fruit are entitled to be considered strictly permanent, but
that they wear out after a time;[813] and we are thus compelled to
resort again to seeds; in which case there is so decided a tendency in
the seedlings to revert to the original type, that our utmost skill is
sometimes baffled in attempting to recover the desired variety.

_Varieties of the cabbage._--The different races of cabbages afford, as
was admitted, an astonishing example of deviation from a common type;
but we can scarcely conceive them to have originated, much less to have
lasted for several generations, without the intervention of man. It is
only by strong manures that these varieties have been obtained, and in
poorer soils they instantly degenerate. If, therefore, we suppose in a
state of nature the seed of the wild _Brassica oleracea_ to have been
wafted from the sea-side to some spot enriched by the dung of animals,
and to have there become a cauliflower, it would soon diffuse its seed
to some comparatively sterile soils around, and the offspring would
relapse to the likeness of the parent stock.

But if we go so far as to imagine the soil, in the spot first occupied,
to be constantly manured by herds of wild animals, so as to continue as
rich as that of a garden, still the variety could not be maintained;
because we know that each of these races is prone to fecundate others,
and gardeners are compelled to exert the utmost diligence to prevent
cross-breeds. The intermixture of the pollen of varieties growing in the
poorer soil around would soon destroy the peculiar characters of the
race which occupied the highly manured tract; for, if these accidents so
continually happen, in spite of our care, among the culinary varieties,
it is easy to see how soon this cause might obliterate every marked
singularity in a wild state.

Besides, it is well known that, although the pampered races which we
rear in our gardens for use or ornament ma