The Evolution of Fruits. 11
order to perpetuate their species. Thus we see such reduction
both in the palms and in the oak.
As an elongated floral receptacle represents a lower type of
structure than one horizontally expanded, and this again is
earlier than the arrangement by which the ovary appears to
be sunk in the receptacle, owing to an absence of separation
for some distance between the axial and lateral structures,
hypogynous insertion, as it is technically termed, probably
preceded perigynous, and perigynous preceded epigynous.
The comparative anatomy of a group tells us far more as
to the structure and evolution of fruits than we can learn
from ontogeny, or the study of individual development. All
foliar organs originate alike as a mere succulent, cellular
papilla; and in vegetable ontogeny stages are so readily
masked that whorled carpels seldom show any trace of
primitively spiral arrangement, nor can we learn much from
their early stages as to the true nature of ovules. Ontogeny
is perhaps most useful in discriminating the origin of those
layers of the pericarp which are known by the ambiguous
terms, epicarp, mesocarp, and endocarp. Probably the
carpels of the primitive fruit would so far retain their foliar
character as to wither into a dry, parchment-like consistence,
and burst by the consequent contraction. Cellular structures
in this state are generally slightly hygroscopic, so that we
have here the origin of many cases of violent dehiscence
facilitating the dispersal of seed.
If, bearing in mind these general principles, we glance at
the distribution of the different varieties of fruits through the
natural orders, we are at once struck by the fact that the
fruit-type is a character of high phylogenetic importance.
The caryopsis of the grasses, the silicle of Crucifers, the
hesperidium of the orange tribe, the drupe and the pome in
two sections of the Rosacea, the legume of the Leguminosae,
the cremocarp of the Umbelliferae, the pepo of the gourd tribe,
the cypsele of the Compositae, and the nut of the Cupulifera,
are familiar instances of this. It occurred to me, therefore,
to review the natural orders from this stand-point, not
disregarding other characters, but paying particular attention