_
has been obtained. In particular, Messrs. Windebank and Kingsbury, of
Southampton, have succeeded in raising a set of plants in which the
flowers are very double and very attractive in a florist's point of
view. The corollas in these flowers are not merely duplicated, but from
their inner surface spring, in some cases, funnel-shaped or tubular
petals (p. 315), so regular in form as quite to resemble a perfect
corolla. These tubes are attached to the inner side of the tube of the
corolla, in the same way as are the stamens, these latter organs being,
it appears, absent. The carpels are present, but open at the top, and
bear numerous ovules, hence it was at first surmised that these plants
were obtained and perpetuated, by the application of pollen from single
flowers to these double-flowered varieties.

The raisers of this fine race however assert that "the double kinds are
all raised from the seed obtained from _single_ flowers; the double
blooms do not produce seed, as a rule, and even if they did yield seed,
and it were to germinate, the plants so raised would simply produce
single flowers." Semi-double flowers will produce seed, but it is
necessary that they should be fertilised with the pollen from the single
blooms. They rarely, however, if ever, produce really double flowers
when so fertilised, and the number of semi-double flowers, even, is
always small, the remainder, and, consequently, the larger part, proving
single. To obtain double varieties, the raiser fertilises certain fine
and striking single flowers, with the pollen of other equally fine
single blooms, and the desired result is obtained. This is Messrs.
Windebank and Kingsbury's _modus operandi_, the exact process or mode of
accomplishment being, however, a professional secret.[569]

From what has been said, as well as from other evidence which it is not
necessary to detail in this place, it may be seen that the causes
assigned by physiologists, and the plans proposed by cultivators for the
production of double flowers, are reducible to three heads, which may be
classed under Plethora, Starvation, and Sterility. These three seem
inconsistent one with the other, but are not so much so as they at first
sight appear to be.

Tho advocates of the plethora theory have much in their favour: for
instance, the greater frequency of double flowers among cultivated
plants than among wild ones. The great preponderance of double flowers
in plants derived from th