s (clausa, Banksiana, etc.)
where in some leaves there is but one row of cells. But with the
examination of other leaves one or more cells of a second row will be
found with very thick walls. Among Hard Pines there is no Old World
species with a biform hypoderm. But there are a few American species
with uniform hypoderm (resinosa, tropicalis, patula and Greggii); while,
in some leaves of the few American Hard Pines with multiform hypoderm,
the uniform hypoderm is a variation.

THE GREEN TISSUE.

In this tissue are the resin-ducts, each with a border of cells,
corresponding in appearance and in chemical reaction with the cells of
the hypoderm and with thinner or thicker walls. With reference to the
green tissue the foliar duct may be in one of four positions.

1. External against the hypoderm fig. 24.
2. Internal against the endoderm fig. 28.
3. Medial in the green tissue, touching neither
hypoderm nor endoderm fig. 26.
4. Septal touching both endoderm and hypoderm,
forming a septum fig. 30.

Among the Soft Pines the external duct is invariable in the subsection
Paracembra. It is also characteristic of the Strobi, where it is
sometimes associated with a medial duct. In the Cembrae and the
Flexiles, however, the ducts are external in some species, or medial or
both in others, without regard to the affinities of these species.

Among the Hard Pines the external duct is characteristic of the Old
World, there being but two American Pines with this character (resinosa
and tropicalis). The internal duct is peculiar to Hard Pines of the New
World, its presence in Old World species being extremely rare. The
medial duct is common to species of both hemispheres, either alone or in
association with ducts in other positions (figs. 25, 27). The septal
duct is peculiar to a few species (oocarpa, tropicalis, and less
frequently Pringlei and Merkusii). I have also seen it in a leaf of P.
canariensis. The internal and septal ducts appear to be confined to the
species of warm-temperate or tropical countries.

The number of resin-ducts of a single leaf may be limited to two or
three (strobus, koraiensis, etc.), but in many species it is exceedingly
variable and often large (pinaster, sylvestris, etc.). Eighteen or more
ducts in a single leaf have been recorded. Such large numbers are
peculi