34 BULLETIN 1136, U. S. DEPARTMENT OF AGRICULTURE 
the tangential ones. If, however, a board is neither radial nor 
tangential, the difference between radial and tangential shrinkage 
will cause “ diamonding ”; its pairs of adjacent sides and edges then 
will no longer be at right angies to each other. (Pl. 138.) The nor- 
mal curvature of the annual rings makes it impossible to find boards 
that are truly radial or truly tangential; in ordinary sawmill opera- 
tion a large number are cut in which the rings are tangential at the 
center of the end section and are at an angle of from 30° to 45° to 
the broad faces at the edges. In boards of this kind, the difference 
between radiai and tangential shrinkage causes cupping; the edges 
of the board turn away from the heart of the log, flattening out the 
curvature of the annual rings. 
CASEHARDENING 
Above the fiber-saturation point, changes in the moisture content 
do not produce changes in the dimensions of the wood. Below this 
point, loss in moisture causes shrinkage and gain causes swelling. 
When a moisture gradient passes through the fiber-saturation point, 
that is, when part of the piece is above the fiber-saturation point and 
part below, unrestrained shrinkage is impossible and consequently 
shrinkage stresses are set up in drying. Similarly when the slope of 
a moisture gradient, below the fiber-saturation point, is changed, 
stresses are set up by the nonuniform shrinking or swelling that takes 
place. ! 
As already suggested, when surface layers of a board that is drying 
pass the fiber-saturation point they tend to shrink. In order to 
succeed, however, they must squeeze together ali of the green wood 
inside, since it has not yet reached the fiber-saturation point and is 
therefore not ready to shrink of its own accord. The first result is 
that the outer layers, in trying to squeeze the core of such a board, 
ereate in it a state of compression and in themselves a corresponding 
state of tension. Consider, in illustration, a rubber band pulling 
together a bundle of papers. The band is stretched, and the papers 
are compressed. The only difference between the surface wood and 
the rubber is that the tension is put into the rubber by actually 
stretching it, whereas the tension is produced in the outer layers of 
the wood by preventing them from shrinking. The same thing occurs 
if a piece of wet leather is kept from shrinking as it dries. 
If a piece of wood has been dried under such restraint that 
a tension stress has been caused in it, and if the restraint is then 
removed, the wood will spontaneously contract enough to relieve the 
stress. However, it will not shrink so much as it would have done 
if it had been free to shrink during the drying, even though the con- 
traction that does occur may be sufficient to relieve the stress entirely. 
Such wood has acquired a tension set. Similarly, when a piece of 
wood below the fiber-saturation point is made to absorb moisture but 
is not permitted to swell, it will acquire a compression set and 
although when the restraint is removed it will spontaneously swell 
somewhat, the amount of expansion will not be so great as that which 
would have occurred if the piece had been allowed to swell without 
restraint in the first place. Further, if dried again, this time without 
restraint, it will shrink to dimensions smaller than the original dry 
