ical, 1.5
inches in diameter and 18 inches gauge length, with squared ends
4 inches long joined to the cylindrical portion with a fillet.
The dimensions are carefully measured, and the usual data
obtained in regard to the rate of growth, proportion of late
wood, location and kind of defects. The weight of the
cylindrical portion of the specimen is obtained after the test.

_Making the test_: After the specimen is fitted in the machine
the load is applied continuously at the rate of 22 deg. per minute.
A troptometer is used in measuring the deformation. Readings are
made until failure occurs, the points being entered on the
cross-section paper. The character of the failure is described.
Moisture determinations are made by the disk method.

_Results_: The conditions of ultimate rupture due to torsion
appear not to be governed by definite mathematical laws; but
where the material is not overstrained, laws may be assumed
which are sufficiently exact for practical cases. The formulae
commonly used for computations are as follows:

5.1 M
(1) T = -------
c^{3}

114.6 T f
(2) G = -----------
a c

a = angle measured by troptometer at elastic limit, in
degrees.
c = diameter of specimen, inches.
f = gauge length of specimen, inches. _G_ = modulus of
elasticity in shear across the grain, pounds per square
inch.
M = moment of torsion at elastic limit, inch-pounds.
T = outer fibre torsional stress at elastic limit, pounds per
square inch.

SPECIAL TESTS

_Spike-pulling Test_

Spike-pulling tests apply to problems of railroad maintenance,
and the results are used to compare the spike-holding powers of
various woods, both untreated and treated with different
preservatives, and the efficiency of various forms of spikes.
Special tests are also made in which the spike is subjected to a
transverse load applied repetitively by a blow.

For details of tests and results see:

Cir. 38, U.S.F.S.: Instructions to engineers of timber tests,
p. 26. Cir. 46, U.S.F.S.: Holding force of railroad spikes in
wooden ties. Bul. 118, U.S.F.S,: Prolonging the life of
cross-ties, pp. 37-40.

_Packing Boxes_

Special tests on the strength of packing boxes of various woods
have been made by the U.S. Forest Service to determine the
merits of different kinds of woods as box material with the view
of substituting new ki