Chapter II: Part I: Principles or Science of Horticulture

Horticulture, apart from the mechanical details connected with the maintenance of a garden and its appurtenances, may be considered as the application of the principles of plant physiology to the cultivation of plants from all parts of the globe, and from various altitudes, soils and situations. The lessons derived from the abstract principles enunciated by the physiologist, the chemist and the physicist require, however, to be modified to suit the special circumstances of plants under cultivation. The necessity for this modification arises from the fact that such plants are subjected to conditions more or less unnatural to them, and that they are grown for special purposes which are at variance, in degree at any rate, with their natural requirements.

The life of the plant (see PLANTS) makes itself manifest in the processes of growth, development and reproduction. By growth is here meant mere increase in bulk, and by development the series of gradual modifications by which a plant, originally simple in its structure and conformation, becomes eventually complicated, and endowed with distinct parts or organs. The reproduction of the higher plants takes place either asexually by the formation of buds or organs answering thereto, or sexually by the production of an embryo plant within the seed. The conditions requisite for the growth, development and reproduction of plants are, in general terms, exposure, at the proper time, to suitable amounts of light, heat and moisture, and a due supply of appropriate food. The various amounts of these needed in different cases have to be adjusted by the gardener, according to the nature of the plant, its "habit" or general mode of growth in its native country, and the influence to which it is there subjected, as also in accordance with the purposes for which it is to be cultivated, &c. It is but rarely that direct information on all these points can be obtained; but inference from previous experience, especially with regard to allied forms, will go far to supply such deficiencies. Moreover, it must be remembered that the conditions most favourable to plants are not always those to which they are subjected in nature, for, owing to the competition of other forms in the struggle for existence, liability to injury from insects, and other adverse circumstances, plants may actually be excluded from the localities best suited for their development. The gardener therefore may, and does, by modifying, improve upon the conditions under which a plant naturally exists. Thus it frequently happens that in our gardens flowers have a beauty and a fragrance, and fruits a size and savour denied to them in their native haunts. It behooves the judicious gardener, then, not to be too slavish in his attempts to imitate natural conditions, and to bear in mind that such attempts sometimes end in failure. The most successful gardening is that which turns to the best account the plastic organization of the plant, and enables it to develop and multiply as perfectly as possible. Experience, coupled with observation and reflection, as well as the more indirect teachings of tradition, are therefore of primary importance to the practical gardener.

We propose here to notice briefly the several parts of a flowering plant, and to point out the rationale of the cultural procedures connected with them (see the references to separate articles at the end of article on BOTANY).

_The Root._--The root, though not precluded from access of air, is not
directly dependent for its growth on the agency of light. The
efficiency of drainage, digging, hoeing and like operations is
accounted for by the manner in which they promote aeration of the
soil, raise its temperature and remove its stagnant water. Owing to
their growth in length at, or rather in the immediate vicinity of,
their tips, roots are enabled to traverse long distances by
surmounting some obstacles, penetrating others, and insinuating
themselves into narrow crevices. As they have no power of absorbing
solid materials, their food must be of a liquid or gaseous character.
It is taken up from the interstices between the particles of soil
exclusively by the finest subdivisions of the fibrils, and in many
cases by the extremely delicate thread-like cells which project from
them and which are known as root-hairs. The importance of the
root-fibres, or "feeding roots" justifies the care which is taken by
every good gardener to secure their fullest development, and to
prevent as far as possible any injury to them in digging, potting and
transplanting, such operations being therefore least prejudicial at
seasons when the plant is in a state of comparative rest.

_Root-Pruning and Lifting._--In apparent disregard of the general rule
just enunciated is the practice of root-pruning fruit trees, when,
from the formation of wood being more active than that of fruit, they
bear badly. The contrariety is more apparent than real, as the
operation consists in the removal of the coarser roots, a process
which results in the development of a mass of fine feeding roots.
Moreover, there is a generally recognised quasi-antagonism between the
vegetative and reproductive processes, so that, other things being
equal, anything that checks the one helps forward the other.

_Watering._--So far as practical gardening is concerned, feeding by
the roots after they have been placed in suitable soil is confined
principally to the administration of water and, under certain
circumstances, of liquid or chemical manure; and no operations demand
more judicious management. The amount of water required, and the times
when it should be applied, vary greatly according to the kind of plant
and the object for which it is grown, the season, the supply of heat
and light, and numerous other conditions, the influence of which is to
be learnt by experience only. The same may be said with respect to the
application of manures. The watering of pot-plants requires especial
care. Water should as a rule be used at a temperature not lower than
that of the surrounding atmosphere, and preferably after exposure for
some time to the air.

_Bottom-Heat._--The "optimum" temperature, or that best suited to
promote the general activity of roots, and indeed of all vegetable
organs, necessarily varies very much with the nature of the plant, and
the circumstances in which it is placed, and is ascertained by
practical experience. Artificial heat applied to the roots, called by
gardeners "bottom-heat," is supplied by fermenting materials such as
stable manure, leaves, &c., or by hot-water pipes. In winter the
temperature of the soil, out of doors, beyond a certain depth is
usually higher than that of the atmosphere, so that the roots are in a
warmer and more uniform medium than are the upper parts of the plant.
Often the escape of heat from the soil is prevented by "mulching,"
i.e. by depositing on it a layer of litter, straw, dead leaves and the
like.

The _Stem_ and its subdivisions or branches raise to the light and air
the leaves and flowers, serve as channels for the passage to them of
fluids from the roots, and act as reservoirs for nutritive substances.
Their functions in annual, biennial and herbaceous perennial plants
cease after the ripening of the seed, whilst in plants of longer
duration layer after layer of strong woody tissue is formed, which
enables them to bear the strains which the weight of foliage and the
exposure to wind entail. The gardener aims usually at producing stout,
robust, short-jointed stems, instead of long lanky growths defective
in woody tissue. To secure these conditions free exposure to light and
air is requisite; but in the case of coppices and woods, or where long
straight spars are needed by the forester, plants are allowed to grow
thickly so as to ensure development in an upward rather than in a
lateral direction. This and like matters will, however, be more fitly
considered in dealing hereafter with the buds and their treatment.

_Leaves._--The work of the leaves may briefly be stated to consist of
the processes of nutrition, respiration and transpiration. Nutri tion
(assimilation) by the leaves includes the inhalation of air, and the
interaction under the influence of light and in the presence of
chlorophyll of the carbon dioxide of the air with the water received
from the root, to form carbonaceous food. Respiration in plants, as in
other organisms, is a process that goes on by night as well as by day
and consists in plants in the breaking up of the complex carbonaceous
substances formed by assimilation into less complex and more
transportable substances. This process, which is as yet imperfectly
understood, is attended by the consumption of oxygen, the liberation
of energy in the form of heat, and the exhalation of carbon dioxide
and water vapour. Transpiration is loss of water by the plant by
evaporation, chiefly from the minute pores or stomata on the leaves.
In xerophytic plants (_e.g._ cacti, euphorbias, &c.) from hot, dry and
almost waterless regions where evaporation would be excessive, the
leaf surface, and consequently the number of stomata, are reduced to a
minimum, as it would be fatal to such plants to exhale vapour as
freely in those regions as the broad-leaved plants that grow in places
where there is abundance of moisture. Although transpiration is a
necessary accompaniment of nutrition, it may easily become excessive,
especially where the plant cannot readily recoup itself. In these
circumstances "syringing" and "damping down" are of value in cooling
the temperature of the air in hothouses and greenhouses and increasing
its humidity, thereby checking excessive transpiration. Shading the
glass with canvas or washes during the summer months has the same
object in view. Syringing is also beneficial in washing away dirt and
insects.

_Buds._--The recognition of the various forms of buds and their modes
of disposition in different plants is a matter of the first
consequence in the operations of pruning and training. Flower-buds are
produced either on the old wood, _i.e._ the shoots of the past year's
growth, or on a shoot of the present year. The peach, horse-chestnut,
lilac, morello cherry, black currant, rhododendron and many other
trees and shrubs develop flower-buds for the next season speedily
after blossoming, and these may be stimulated into premature growth.
The peculiar short, stunted branches or "spurs" which bear the
flower-buds of the pear, apple, plum, sweet cherry, red currant,
laburnum, &c., deserve special attention. In the rose, passion-flower,
clematis, honeysuckle, &c., in which the flower-buds are developed at
the ends of the young shoot of the year, we have examples of plants
destitute of flower-buds during the winter.

_Propagation by Buds._--The detached leaf-buds (_gemmae_ or
_bulbils_), of some plants are capable under favourable conditions of
forming new plants. The edges of the leaves of _Bryophyllum calycinum_
and of _Cardamine pratensis_, and the growths in the axils of the
leaves of _Lilium bulbiferum_, as well as the fronds of certain ferns
(e.g. _Asplenium bulbiferum_), produce buds of this character. It is
a matter of familiar observation that the ends of the shoots of
brambles take root when bent down to the ground. In some instances
buds form on the roots, and may be used for purposes of propagation,
as in the Japan quince, the globe thistle, the sea holly, some sea
lavenders, _Bocconia_, _Acanthus_, &c. Of the tendency in buds to
assume an independent existence gardeners avail themselves in the
operations of striking "cuttings," and making "layers" and "pipings,"
as also in budding and grafting. In taking a slip or cutting the
gardener removes from the parent plant a shoot having one or more buds
or "eyes," in the case of the vine one only, and places it in a moist
and sufficiently warm situation, where, as previously mentioned, undue
evaporation from the surface is prevented. For some cuttings, pots
filled with light soil, with the protection of the propagating-house
and of bell-glasses, are requisite; but for many of our hardy
deciduous trees and shrubs no such precautions are necessary, and the
insertion of a short shoot about half its length into moist and gritty
ground at the proper season suffices to ensure its growth. In the case
of the more delicate plants, the formation of roots is preceded by the
production from the cambium of the cuttings of a succulent mass of
tissue, the _callus_. It is important in some cases, _e.g._ zonal
pelargoniums, fuchsias, shrubby calceolarias, dahlias, carnations,
&c., to retain on the cutting some of its leaves, so as to supply the
requisite food for storage in the callus. In other cases, where the
buds themselves contain a sufficiency of nutritive matter for the
young growths, the retention of leaves is not necessary. The most
successful mode of forming roots is to place the cuttings in a mild
bottom-heat, which expedites their growth, even in the case of many
hardy plants whose cuttings strike roots in the open soil. With some
hard-wooded trees, as the common white-thorn, roots cannot be obtained
without bottom-heat. It is a general rule throughout plant culture
that the activity of the roots shall be in advance of that of the
leaves. Cuttings of deciduous trees and shrubs succeed best if planted
early in autumn while the soil still retains the solar heat absorbed
during summer. For evergreens August or September, and for greenhouse
and stove-plants the spring and summer months, are the times most
suitable for propagation by cuttings.

_Layering_ consists simply in bending down a branch and keeping it in
contact with or buried to a small depth in the soil until roots are
formed; the connexion with the parent plant may then be severed. Many
plants can be far more easily propagated thus than by cuttings.

_Grafting_ or "_working_" consists in the transfer of a branch, the
"graft" or "scion," from one plant to another, which latter is termed
the "stock." The operation must be so performed that the growing
tissues, or cambium-layer of the scion, may fit accurately to the
corresponding layer of the stock. In _budding_, as with roses and
peaches, a single bud only is implanted. _Inarching_ is essentially
the promotion of the union of a shoot of one plant to that of another
of the same or allied species or variety. The outer bark of each being
removed, the two shoots are kept in contact by ligature until union is
established, when the scion is completely severed from its original
attachments. This operation is varied in detail according to the kind
of plant to be propagated, but it is essential in all cases that the
affinity between the two plants be near, that the union be neatly
effected, and that the ratio as well as the season of growth of stock
and scion be similar.

The selection of suitable stocks is a matter still requiring much
scientific experiment. The object of grafting is to expedite and
increase the formation of flowers and fruit. Strong-growing pears, for
instance, are grafted on the quince stock in order to restrict their
tendency to form "gross" shoots and a superabundance of wood in place
of flowers and fruit. Apples, for the same reason, are "worked" on the
"paradise" or "doucin" stocks, which from their influence on the scion
are known as dwarfing stocks. Scions from a tree which is weakly, or
liable to injury by frosts, are strengthened by engrafting on robust
stocks. Lindley has pointed out that, while in Persia, its native
country, the peach is probably best grafted on the peach, or on its
wild type the almond, in England, where the summer temperature of the
soil is much lower than that of Persia, it might be expected, as
experience has proved, to be most successful on stocks of the native
plum.

The soil in which the stock grows is a point demanding attention. From
a careful series of experiments made in the Horticultural Society's
Garden at Chiswick, it was found that where the soil is loamy, or
light and slightly enriched with decayed vegetable matter, the apple
succeeds best on the doucin stock, and the pear on the quince; and
where it is chalky it is preferable to graft the apple on the crab,
and the pear on the wild pear. For the plum on loamy soils the plum,
and on chalky and light soils the almond, are the most desirable
stocks, and for the cherry on loamy or light rich soils the wild
cherry, and on chalk the "mahaleb" stock.

The form and especially the quality of fruit is more or less affected
by the stock upon which it is grown. The Stanwick nectarine, so apt to
crack and not to ripen when worked in the ordinary way, is said to be
cured of these propensities by being first budded close to the ground,
on a very strong-growing Magnum Bonum plum, worked on a Brussels
stock, and by then budding the nectarine on the Magnum Bonum about a
foot from the ground. The fruit of the pear is of a higher colour and
smaller on the quince stock than on the wild pear; still more so on
the medlar. On the mountain ash the pear becomes earlier.

The effects produced by stock on scion, and more particularly by scion
on stock, are as a rule with difficulty appreciable. Nevertheless, in
exceptional cases modified growths, termed "graft-hybrids," have been
obtained which have been attributed to the commingling of the
characteristics of stock and scion (see HYBRIDISM). Of these the most
remarkable example is _Cytisus Adami_, a tree which year after year
produces some shoots, foliage and flowers like those of the common
laburnum, others like those of the very different looking dwarf shrub
_C. purpureus_, and others again intermediate between these. We may
hence infer that _C. purpureus_ was grafted or budded on the common
laburnum, and that the intermediate forms are the result of
graft-hybridization. Numerous similar facts have been recorded. Among
gardeners the general opinion is against the possibility of
graft-hybridization. The wonder, however, seems to be that it does not
occur more frequently, seeing that fluids must pass from stock to
scion, and matter elaborated in the leaves of the scion must certainly
to some extent enter the stock. It is clear, nevertheless, from
examination that as a rule the wood of the stock and the wood of the
scion retain their external characters year by year without change.
Still, as in the laburnum just mentioned, in the variegated jasmine
and in _Abutilon Darwinii_, in the copper beech and in the
horse-chestnut, the influence of a variegated scion has occasionally
shown itself in the production from the stock of variegated shoots. At
a meeting of the Scottish Horticultural Association (see _Gard.
Chron._, Jan. 10, 1880, figs. 12-14) specimens of a small roundish
pear, the "Aston Town," and of the elongated kind known as "Beurre
Clairgeau," were exhibited. Two more dissimilar pears hardly exist.
The result of working the Beurre Clairgeau upon the Aston Town was the
production of fruits precisely intermediate in size, form, colour,
speckling of rind and other characteristics. Similar, though less
marked, intermediate characters were obvious in the foliage and
flowers.

Double grafting (French, _greffe sur greffe_) is sufficiently
explained by its name. By means of it a variety may often be
propagated, or its fruit improved in a way not found practicable under
ordinary circumstances. For its successful prosecution prolonged
experiments in different localities and in gardens devoted to the
purpose are requisite.

_Planting._--By removal from one place to another the growth of every
plant receives a check. How this check can be obviated or reduced,
with regard to the season, the state of atmosphere, and the condition
and circumstances of the plant generally, is a matter to be considered
by the practical gardener.

As to season, it is now admitted with respect to deciduous trees and
shrubs that the earlier in autumn planting is performed the better;
although some extend it from the period when the leaves fall to the
first part of spring, before the sap begins to move. If feasible, the
operation should be completed by the end of November, whilst the soil
is still warm with the heat absorbed during summer. Attention to this
rule is specially important in the case of rare and delicate plants.
Early autumn planting enables wounded parts of roots to be healed
over, and to form fibrils, which will be ready in spring, when it is
most required, to collect food for the plant. Planting late in spring
should, as far as possible, be avoided, for the buds then begin to
awaken into active life, and the draught upon the roots becomes great.
It has been supposed that because the surface of the young leaves is
small transpiration is correspondingly feeble; but it must be
remembered, not only that their newly-formed tissue is unable without
an abundant supply of sap from the roots to resist the excessive
drying action of the atmosphere, but that, in spring, the lowness of
the temperature at that season in Great Britain prevents the free
circulation of the sap. The comparative dryness of the atmosphere in
spring also causes a greater amount of transpiration then than in
autumn and winter. Another fact in favour of autumnal planting is the
production of roots in winter.

The best way of performing transplantation depends greatly on the size
of the trees, the soil in which they grow, and the mechanical
appliances made use of in lifting and transporting them. The smaller
the tree the more successfully can it be removed. The more
argillaceous and the less siliceous the soil the more readily can
balls of earth be retained about the roots. All planters lay great
stress on the preservation of the fibrils; the point principally
disputed is to what extent they can with safety be allowed to be cut
off in transplantation. Trees and shrubs in thick plantations, or in
sheltered warm places, are ill fitted for planting in bleak and cold
situations. During their removal it is important that the roots be
covered, if only to prevent desiccation by the air. Damp days are
therefore the best for the operation; the dryest months are the most
unfavourable. Though success in transplanting depends much on the
humidity of the atmosphere, the most important requisite is warmth in
the soil; humidity can be supplied artificially, but heat cannot.

_Pruning_, or the removal of superfluous growths, is practised in
order to equalize the development of the different parts of trees, or
to promote it in particular directions so as to secure a certain form,
and, by checking undue luxuriance, to promote enhanced fertility. In
the rose-bush, for instance, in which, as we have seen, the
flower-buds are formed on the new wood of the year, pruning causes the
old wood to "break," i.e. to put forth a number of new buds, some of
which will produce flowers at their extremities. The manner and the
time in which pruning should be accomplished, and its extent, vary
with the plant, the objects of the operation, i.e. whether for the
production of timber or fruit, the season and various other
circumstances. So much judgment and experience does the operation call
for that it is a truism to say that bad pruning is worse than none.
The removal of weakly, sickly, overcrowded and gross infertile shoots
is usually, however, a matter about which there can be few mistakes
when once the habit of growth and the form and arrangement of the buds
are known. Winter pruning is effected when the tree is comparatively
at rest, and is therefore less liable to "bleeding" or outpouring of
sap. Summer pruning or pinching off the tips of such of the younger
shoots as are not required for the extension of the tree, when not
carried to too great an extent, is preferable to the coarser more
reckless style of pruning. The injury inflicted is less and not so
concentrated; the wounds are smaller, and have time to heal before
winter sets in. The effects of badly-executed pruning, or rather
hacking, are most noticeable in the case of forest trees, the
mutilation of which often results in rotting, canker and other
diseases. Judicious and timely thinning so as to allow the trees room
to grow, and to give them sufficiency of light and air, will generally
obviate the need of the pruning-saw, except to a relatively small
extent.

_Training_ is a procedure adopted when it is required to grow plants
in a limited area, or in a particular shape, as in the case of many
plants of trailing habit. Judicious training also may be of importance
as encouraging the formation of flowers and fruit. Growth in length is
mainly in a vertical direction, or at least at the ends of the shoots;
and this should be encouraged, in the case of a timber tree, or of a
climbing plant which it is desired should cover a wall quickly; but
where flowers or fruit are specially desired, then, when the wood
required is formed, the lateral shoots may often be trained more or
less downward to induce fertility. The refinements of training, as of
pruning, may, however, be carried too far; and not unfrequently the
symmetrically trained trees of the French excite admiration in every
respect save fertility.

_Sports or Bud Variations._--Here we may conveniently mention certain
variations from the normal condition in the size, form or disposition
of buds or shoots on a given plant. An inferior variety of pear, for
instance, may suddenly produce a shoot bearing fruit of superior
quality; a beech tree, without obvious cause, a shoot with finely
divided foliage; or a camellia an unwontedly fine flower. When removed
from the plant and treated as cuttings or grafts, such sports may be
perpetuated. Many garden varieties of flowers and fruits have thus
originated. The cause of their production is very obscure.

_Formation of Flowers._--Flowers, whether for their own sake or as the
necessary precursors of the fruit and seed, are objects of the
greatest concern to the gardener. As a rule they are not formed until
the plant has arrived at a certain degree of vigour, or until a
sufficient supply of nourishment has been stored in the tissues of the
plant. The reproductive process of which the formation of the flower
is the first stage being an exhaustive one, it is necessary that the
plant, as gardeners say, should get "established" before it flowers.
Moreover, although the green portions of the flower do indeed perform
the same office as the leaves, the more highly coloured and more
specialized portions, which are further removed from the typical
leaf-form, do not carry on those processes for which the presence of
chlorophyll is essential; and the floral organs may, therefore, in a
rough sense, be said to be parasitic upon the green parts. A check or
arrest of growth in the vegetative organs seems to be a necessary
preliminary to the development of the flower.

A diminished supply of water at the root is requisite, so as to check
energy of growth, or rather to divert it from leaf-making. Partial
starvation will sometimes effect this; hence the grafting of
free-growing fruit trees upon dwarfing stocks, as before alluded to,
and also the "ringing" or girdling of fruit trees, i.e. the removal
from the branch of a ring of bark, or the application of a tight
cincture, in consequence of which the growth of the fruits above the
wound or the obstruction is enhanced. On the same principle the use of
small pots to confine the roots, root-pruning and lifting the roots,
and exposing them to the sun, as is done in the case of the vine in
some countries, are resorted to. A higher temperature, especially with
deficiency of moisture, will tend to throw a plant into a flowering
condition. This is exemplified by the fact that the temperature of the
climate of Great Britain is too low for the flowering, though
sufficiently high for the growth of many plants. Thus the Jerusalem
artichoke, though able to produce stems and tubers abundantly, only
flowers in exceptionally hot seasons.

_Forcing._--The operation of forcing is based upon the facts just
mentioned. By subjecting a plant to a gradually increasing
temperature, and supplying water in proportion, its growth may be
accelerated; its season of development may be, as it were,
anticipated; it is roused from a dormant to an active state. Forcing
therefore demands the most careful adjustment of temperature and
supplies of moisture and light.

Deficiency of light is less injurious than might at first be expected,
because the plant to be forced has stored up in its tissues, and
available for use, a reserve stock of material formed through the
agency of light in former seasons. The intensity of the colour of
flowers and the richness of flavour of fruit are, however, deficient
where there is feebleness of light. Recent experiments show that the
influence of electric light on chlorophyll is similar to that of
sunlight, and that deficiencies of natural light may to some extent be
made good by its use. The employment of that light for forcing
purposes would seem to be in part a question of expense. The advantage
hitherto obtained from its use has consisted in the rapidity with
which flowers have been formed and fruits ripened under its influence,
circumstances which go towards compensating for the extra cost of
production.

_Retardation._--The art of retarding the period of flowering in
certain plants consists, in principle, in the artificial application
of cold temperatures whereby the resting condition induced by low
winter temperature is prolonged. For commercial purposes, crowns of
lily of the valley, tulip and other bulbs, and such deciduous woody
plants as lilac and deciduous species of rhododendron, while in a
state of rest, are packed in wet moss and introduced into cold-storage
chambers, where they may be kept in a state of quiescence, it desired,
throughout the following summer. The temperature of the cold chamber
is varied from the freezing-point of water, to a few degrees lower,
according to the needs of the plants under treatment. When required
for use they are removed to cool sheds to thaw, and are then gradually
inured to higher temperatures. The chief advantages of retarded plants
are:--(a) they may be flowered almost at will; (b) they are readily
induced to flower at those times when unretarded plants refuse to
respond to forcing. Cold-storage chambers form a part of the equipment
of most of the leading establishments where flowers are grown for
market.

_Double Flowers._--The taste of the day demands that "double flowers"
should be largely grown. Though in many instances, as in hyacinths,
they are less beautiful than single ones, they always present the
advantage of being less evanescent. Under the vague term "double" many
very different morphological changes are included. The flower of a
double dahlia, e.g. offers a totally different condition of structure
from that of a rose or a hyacinth. The double poinsettia, again, owes
its so-called double condition merely to the increased number of its
scarlet involucral leaves, which are not parts of the flower at all.
It is reasonable, therefore, to infer that the causes leading to the
production of double flowers are varied. A good deal of difference of
opinion exists as to whether they are the result of arrested growth or
of exuberant development, and accordingly whether restricted food or
abundant supplies of nourishment are the more necessary for their
production. It must suffice here to say that double flowers are most
commonly the result of the substitution of brightly-coloured petals
for stamens or pistils or both, and that a perfectly double flower
where all the stamens and pistils are thus metamorphosed is
necessarily barren. Such a plant must needs be propagated by cuttings.
It rarely happens, however, that the change is quite complete
throughout the flower, and so a few seeds may be formed, some of which
may be expected to reproduce the double-blossomed plants. By
continuous selection of seed from the best varieties, and "roguing" or
eliminating plants of the ordinary type, a "strain" or race of double
flowers is gradually produced.

_Formation of Seed--Fertilization._--In fertilization--the influence
in flowering plants of the male-cell in the pollen tube upon the
egg-cell in the ovule (see BOTANY)--there are many circumstances of
importance horticulturally, to which, therefore, brief reference must
be made. Flowers, generally speaking, are either self-fertilized,
cross-fertilized or hybridized. Self-fertilization occurs when the
pollen of a given flower affects the egg-cell of the same individual
flower. Cross-fertilization varies both in manner and degree. In the
simplest instances the pollen of one flower fertilizes the ovules of
another on the same plant, owing to the stamens arriving at maturity
in any one flower earlier or later than the pistils.

Cross-fertilization must of necessity occur when the flowers are
structurally unisexual, as in the hazel, in which the male and female
flowers are monoecious, or separate on the same plant, and in the
willow, in which they are dioecious, or on different plants. A
conspicuous example of a dioecious plant is the common aucuba, of
which for years only the female plant was known in Britain. When,
through the introduction of the male plant from Japan, its
fertilization was rendered possible, ripe berries, before unknown,
became common ornaments of the shrub.

The conveyance of pollen from one flower to another in
cross-fertilization is effected naturally by the wind, or by the
agency of insects and other creatures. Flowers that require the aid of
insects usually offer some attraction to their visitors in the shape
of bright colour, fragrance or sweet juices. The colour and markings
of a flower often serve to guide the insects to the honey, in the
obtaining of which they are compelled either to remove or to deposit
pollen. The reciprocal adaptations of insects and flowers demand
attentive observation on the part of the gardener concerned with the
growing of grapes, cucumbers, melons and strawberries, or with the
raising of new and improved varieties of plants. In wind-fertilized
plants the flowers are comparatively inconspicuous and devoid of much
attraction for insects; and their pollen is smoother and smaller, and
better adapted for transport by the wind, than that of
insect-fertilized plants, the roughness of which adapts it for
attachment to the bodies of insects.

It is very probable that the same flower at certain times and seasons
is self-fertilizing, and at others not so. The defects which cause
gardeners to speak of certain vines as "shy setters," and of certain
strawberries as "blind," may be due either to unsuitable conditions of
external temperature, or to the non-accomplishment, from some cause or
other, of cross-fertilization. In a vinery, tomato-house or a
peach-house it is often good practice at the time of flowering to tap
the branches smartly with a stick so as to ensure the dispersal of the
pollen. Sometimes more delicate and direct manipulation is required,
and the gardener has himself to convey the pollen from one flower to
another, for which purpose a small camel's-hair pencil is generally
suitable. The degree of fertility varies greatly according to external
conditions, the structural and functional arrangements just alluded
to, and other causes which may roughly be called constitutional. Thus,
it often happens that an apparently very slight change in climate
alters the degree of fertility. In a particular country or at certain
seasons one flower will be self-sterile or nearly so, and another just
the opposite.

_Hybridization._--Some of the most interesting results and many of the
gardener's greatest triumphs have been obtained by hybridization, i.e.
the crossing of two individuals not of the same but of two distinct
species of plants, as, for instance, two species of rhododendron or
two species of orchid (see HYBRIDISM). It is obvious that
hybridization differs more in degree than in kind from
cross-fertilization. The occurrence of hybrids in nature explains the
difficulty experienced by botanists in deciding on what is a species,
and the widely different limitations of the term adopted by different
observers in the case of willows, roses, brambles, &c. The artificial
process is practically the same in hybridization as in
cross-fertilization, but usually requires more care. To prevent
self-fertilization, or the access of insects, it is advisable to
remove the stamens and even the corolla from the flower to be
impregnated, as its own pollen or that of a flower of the same species
is often found to be "prepotent." There are, however, cases, e.g. some
passion-flowers and rhododendrons, in which a flower is more or less
sterile with its own, but fertile with foreign pollen, even when this
is from a distinct species. It is a singular circumstance that
reciprocal crosses are not always or even often possible; thus, one
rhododendron may afford pollen perfectly potent on the stigma of
another kind, by the pollen of which latter its own stigma is
unaffected.

The object of the hybridizer is to obtain varieties exhibiting
improvements in hardihood, vigour, size, shape, colour, fruitfulness,
resistance to disease or other attributes. His success depends not
alone on skill and judgment, for some seasons, or days even, are found
more propitious than others. Although promiscuous and hap-hazard
procedures no doubt meet with a measure of success, the best results
are those which are attained by systematic work with a definite aim.

Hybrids are sometimes less fertile than pure-bred species, and are
occasionally quite sterile. Some hybrids, however, are as fertile as
pure-bred plants. Hybrid plants may be again crossed, or even
re-hybridized, so as to produce a progeny of very mixed parentage.
This is the case with many of our roses, dahlias, begonias,
pelargoniums, orchids and other long or widely cultivated garden
plants.

_Reversion._--In modified forms of plants there is frequently a
tendency to "sport" or revert to parental or ancestral
characteristics. So markedly is this the case with hybrids that in a
few generations all traces of a hybrid origin may disappear. The
dissociation of the hybrid element in a plant must be obviated by
careful selection. The researches of Gregor Johann Mendel (1822-1884),
abbot of the Augustinian monastery at Brunn, in connexion with peas
and other plants, apparently indicate that there is a definite natural
law at work in the production of hybrids. Having crossed yellow and
green seeded peas both ways, he found that the progeny resulted in
_all yellow_ coloured seeds. These gave rise in due course to a second
generation in which there were three yellows to one green. In the
third generation the yellows from the second generation gave the
proportion of one pure yellow, two impure yellows, and one green;
while the green seed of the second generation threw only green seeds
in the third, fourth and fifth generations. The pure yellow in the
third generation also threw pure yellows in the fourth and fifth and
succeeding generations. The impure yellows, however, in the next
generation gave rise to one pure yellow, one pure green, to two impure
yellows, and so on from generation to generation. Accordingly as the
green or the yellow predominated in the progeny it was termed
"dominant," while the colour that disappeared was called "recessive."
It happened, however, that a recessive colour in one generation
becomes the dominant in a succeeding one.

_Germination._--The length of the period during which seeds remain
dormant after their formation is very variable. The conditions for
germination are much the same as for growth in general. Access to
light is not required, because the seed contains a sufficiency of
stored-up food. The temperature necessary varies according to the
nature and source of the seed. Some seeds require prolonged immersion
in water to soften their shells; others are of so delicate a texture
that they would dry up and perish if not kept constantly in a moist
atmosphere. Seeds buried too deeply receive a deficient supply of air.
As a rule, seeds require to be sown more deeply in proportion to their
size and the lightness of the soil.

The time required for germination in the most favourable circumstances
varies very greatly, even in the same species, and in seeds taken from
one pod. Thus the seeds of _Primula japonica_, though sown under
precisely similar conditions, yet come up at very irregular intervals
of time. Germination is often slower where there is a store of
available food in the perisperm, or in the endosperm, or in the embryo
itself, than where this is scanty or wanting. In the latter case the
seedling has early to shift for itself, and to form roots and leaves
for the supply of its needs.

_Selection._--Supposing seedlings to have been developed, it is found
that a large number of them present considerable variations, some
being especially robust, others peculiar in size or form. Those most
suitable for the purpose of the gardener are carefully selected for
propagation, while others not so desirable are destroyed; and thus
after a few generations a fixed variety, race or strain superior to
the original form is obtained. Many garden plants have originated
solely by selection; and much has been done to improve our breeds of
vegetables, flowers and fruit by systematic selection.

Large and well-formed seeds are to be preferred for harvesting. The
seeds should be kept in sacks or bags in a dry place, and if from
plants which are rare, or liable to lose their vitality, they are
advantageously packed for transmission to a distance in hermetically
sealed bottles or jars filled with earth or moss, without the addition
of moisture.

It will have been gathered from what has been said that seeds cannot
always be depended on to reproduce exactly the characteristics of the
plant which yielded them; for instance, seeds of the greengage plum or
of the Ribston pippin will produce a plum or an apple, but not these
particular varieties, to perpetuate which grafts or buds must be
employed. (M. T. M.; W. R. W.)