. CONE-TISSUES AND SEEDS]

THE WOOD. Plate VII.

With the exception of the medullary rays, a very small proportion of the
whole, the wood of Pinus, as seen in cross-section (fig. 82), is a
homogeneous tissue of wood-tracheids with interspersed resin-ducts. In
tangential section the medullary rays appear in two forms, linear,
without a resin-duct, and fusiform, with a central resin-duct. In radial
section the cells of the linear rays are of two kinds, ray-tracheids,
forming the upper and lower limits of the ray, characterized by small
bordered pits, and ray-cells, between the tracheids, characterized by
simple pits.

The walls of the ray-tracheids may be smooth or dentate; the pits of the
ray-cells may be large or small. These conditions admit of four
combinations, all of which appear in the medullary rays of Pinus, and of
which a schematic representation is given in Plate VII. These
combinations are

Ray-tracheids with smooth walls. Soft Pines.
Ray-cells with large pits Subsection Cembra fig. 80.
Ray-cells with small pits Subsection Paracembra fig. 81.

Ray-tracheids with dentate walls. Hard Pines.
Ray-cells with large pits Group Lariciones fig. 83.
Ray-cells with small pits Other Hard Pines fig. 84.

This, the simplest classification of Pine-wood, is not without
exceptions. P. pinea of the Hard Pines resembles, in its
wood-characters, P. Gerardiana and P. Bungeana of the Soft Pines. The
dentate ray-tracheids of P. longifolia are not always obvious. The
tracheids of P. luchuensis, according to Bergerstein (Wiesner Festschr.
112), have smooth walls. My specimen shows dentate tracheids. There is
also evidence of transition from small to large pits (I. W. Bailey in
Am. Nat. xliv. 292). Both large and small pits appear in my specimen of
P. Merkusii.

Of other wood-characters, the presence or absence of tangential pits in
the tracheids of the late wood establishes a distinction between Soft
and Hard Pines. These pits, however, while always present in Soft Pines,
are not always absent in Hard Pines. The single and multiple rows of
resin-ducts in the wood of the first year may prove to be a reliable
sectional distinction, but this character has not been sufficiently
investigated to test its constancy. The wood-characters, therefore,
however decisive they may be for establishing the phylogenetic relations
of different genera, must be employed in th