ion, becomes
one of the causes of the second great trouble in wood seasoning,
namely, the difference in the shrinkage along the radius and that
along the rings or tangent. This greater tangential shrinkage appears
to be due in part to the causes just mentioned, but also to the fact
that the greatly shrinking bands of summer-wood are interrupted along
the radius by as many bands of porous spring-wood, while they are
continuous in the tangential direction. In this direction, therefore,
each such band tends to shrink, as if the entire piece were composed
of summer-wood, and since the summer-wood represents the greater part
of the wood substance, this greater tendency to tangential shrinkage
prevails.

The effect of this greater tangential shrinkage effects every phase of
woodworking. It leads to permanent checks and causes the log or piece
to split open on drying. Sawed in two, the flat sides of the log
become convex; sawed into timber, it checks along the median line of
the four faces, and if converted into boards, the latter checks
considerably from the end through the center, all owing to the greater
tangential shrinkage of the wood.

Briefly, then, shrinkage of wood is due to the fact that the cell
walls grow thinner on drying. The thicker cell walls and therefore the
heavier wood shrinks most, while the water in the cell cavities does
not influence the volume of the wood.

Owing to the great difference of cells in shape, size, and thickness
of walls, and still more in their arrangement, shrinkage is not
uniform in any kind of wood. This irregularity produces strains, which
grow with the difference between adjoining cells and are greatest at
the pith rays. These strains cause warping and checking, but exist
even where no outward signs are visible. They are greater if the wood
is dried rapidly than if dried slowly, but can never be entirely
avoided.

Temporary checks are caused by the more rapid drying of the outer
parts of any stick; permanent checks are due to the greater shrinkage,
tangentially, along the rings than along the radius. This, too, is the
cause of most of the ordinary phenomena of shrinkage, such as the
difference in behavior of the entire and quartered logs, "bastard"
(tangent) and rift (radial) boards, etc., and explains many of the
phenomena erroneously attributed to the influence of bark, or of the
greater shrinkage of outer and inner parts of any log.

Once dry, wood may be swelled again to i