ent s'allier les
unes aux autres par voie d'hybridite, sans que leur descendance perde
la faculte de se perpetuer." If we were to trust to external
differences alone, and give up the test of sterility, a multitude of
species would have to be formed out of the varieties of these three
species of Cucurbita. Many naturalists at the present day lay far too
little stress, in my opinion, on the test of sterility; yet it is not
improbable that distinct species of plants after a long course of
cultivation and variation may have their mutual sterility eliminated,
as we have every reason to believe has occurred with domesticated
animals. Nor, in the case of plants under cultivation, should we be
justified in assuming that varieties never acquire a slight degree of
mutual sterility, as we shall more fully see in a future chapter when
certain facts are given on the high authority of Gaertner and
Koelreuter.[752]

The forms of _C. pepo_ are classed by Naudin under seven sections, each
including subordinate varieties. He considers this plant as probably
the most variable in the world. The fruit of one variety (pp. 33, 46)
exceeds in volume that of another by more than two thousand fold! When
the fruit is of very large size, the number produced is few (p. 45);
when of small size, many are produced. No less astonishing (p. 33) is
the variation in the shape of the fruit; the typical form apparently is
egg-like, but this becomes either drawn out into a cylinder, or
shortened into a flat disc. We have also an almost infinite diversity
in the colour and state of surface of the fruit, in the hardness both
of the shell and of the flesh, and in the taste of the flesh, which is
either extremely sweet, farinaceous, or slightly bitter. The seeds also
differ in a slight degree in shape, and wonderfully in size (p. 34),
namely, from six or seven to more than twenty-five millimetres in
length.

In the varieties which grow upright or do not run and climb, the
tendrils, though useless (p. 31), are either present or are represented
by various semi-monstrous organs, or are quite absent. The tendrils are
even absent in some running varieties in which the stems are much
elongated. It is a singular fact that (p. 31), in all the varieties
with dwarfed stems, the leaves closely resemble each other in shape.

Those