than one way, since on the stalks of some plants appear leaves
which are already more or less colored long before they approach
inflorescence; others are fully colored when near inflorescence. Nature
also goes over at once to the corolla, sometimes by skipping over the
organs of the calyx, and in such a case we likewise have an opportunity
to observe that leaves of the stalk become transformed into petals. Thus
on the stalk of tulips, for instance, there sometimes appears an almost
completely developed and colored petal. Even more remarkable is the
case when such a leaf, half green and half of it belonging to the stalk,
remains attached to the latter, while another colored part is raised
with the corolla, and the leaf is thus torn in two.

"The relationship between the petals and stamens is very close. In some
instances nature makes the transition regular--e.g., among the Canna
and several plants of the same family. A true, little-modified petal is
drawn together on its upper margin, and produces a pollen sac, while the
rest of the petal takes the place of the stamen. In double flowers
we can observe this transition in all its stages. In several kinds of
roses, within the fully developed and colored petals there appear other
ones which are drawn together in the middle or on the side. This drawing
together is produced by a small weal, which appears as a more or less
complete pollen sac, and in the same proportion the leaf approaches the
simple form of a stamen.

"The pistil in many cases looks almost like a stamen without anthers,
and the relationship between the formation of the two is much closer
than between the other parts. In retrograde fashion nature often
produces cases where the style and stigma (Narben) become retransformed
into petals--that is, the Ranunculus Asiaticus becomes double by
transforming the stigma and style of the fruit-receptacle into real
petals, while the stamens are often found unchanged immediately behind
the corolla.

"In the seed receptacles, in spite of their formation, of their special
object, and of their method of being joined together, we cannot fail to
recognize the leaf form. Thus, for instance, the pod would be a simple
leaf folded and grown together on its margin; the siliqua would consist
of more leaves folded over another; the compound receptacles would be
explained as being several leaves which, being united above one centre,
keep their inward parts separate and are joined on