case constantly
change their situation. Objections which are based on the proof of the
prevalence of self-fertilisation are not, therefore, pertinent. At first
sight another point of view, which has been more recently urged, appears
to have more weight.

W. Burck (Burck, "Darwin's Kreuzeungsgesetz... ", "Biol. Centralbl".
XXVIII. 1908, page 177.) has expressed the opinion that the beneficial
results of cross-fertilisation demonstrated by Darwin concern only
hybrid plants. These alone become weaker by self-pollination; while
pure species derive no advantage from crossing and no disadvantage from
self-fertilisation. It is certain that some of the plants used by Darwin
were of hybrid origin. (It is questionable if this was always the case.)
This is evident from his statements, which are models of clearness and
precision; he says that his Ipomoea plants "were probably the offspring
of a cross." ("Cross and Self fertilisation" (1st edition), page 55.)
The fixed forms of this plant, such as Hero, which was produced by
self-fertilisation, and a form of Mimulus with white flowers spotted
with red probably resulted from splitting of the hybrids. It is true
that the phenomena observed in self-pollination, e.g. in Ipomoea, agree
with those which are often noticed in hybrids; Darwin himself drew
attention to this.

Let us next call to mind some of the peculiarities connected with
hybridisation. We know that hybrids are often characterized by their
large size, rapidity of growth, earlier production of flowers, wealth of
flower-production and a longer life; hybrids, if crossed with one of the
two parent forms, are usually more fertile than when they are crossed
together or with another hybrid. But the characters which hybrids
exhibit on self-fertilisation are rather variable. The following
instance may be quoted from Gartner: "There are many hybrids which
retain the self-fertility of the first generation during the second
and later generations, but very often in a less degree; a considerable
number, however, become sterile." But the hybrid varieties may be more
fertile in the second generation than in the first, and in some hybrids
the fertility with their own pollen increases in the second, third,
and following generations. (K.F. Gartner, "Versuche uber die
Bastarderzeugung", Stuttgart, 1849, page 149.) As yet it is impossible
to lay down rules of general application for the self-fertility of
hybrids. That the beneficial influenc