exceed the oven-dry weight given in the tables by from 15 to 30 per cent., according to size and species. The oven-dry weight is obtained from the weight and from the volume of measurements of the timber at the time of test, in conjunction with subsequent moisture determinations. The shrinkage of the timber is not taken. into account. This is variable and at present is not accurately known. It is estimated that air-dry timber has lost about 5 per cent. of its green volume.

The actual weight of the wood at the time it was tested may be computed from the oven-dry weight and moisture per cent. Thus, a dry weight of 29.4 pounds per cubic foot, at a moisture per cent. of 19.4, gives a weight of 35.1 pounds per cubic foot at the time of the test. The measurements of volume were, in most cases, those of surfaced lumber.

The origin of the sticks and their stage of seasoning are carefully described. Photographs were taken and drawings made locating knots and showing the amount of heart, sap, wane, etc. The sticks were graded by an experienced lumber inspector.

The tables give the average, maximum, and minimum values. Results of tests on individual sticks will be published in connection with special reports.

SUMMARY.

A digest of the results of bending tests on large sticks given in the tables is shown in Table 1, which indicates the weight, strength, and stiffness of beams, such as are found on the market and used by engineers. More detailed information is given in the various other tables. The modulus of rupture represents fairly well the strength of the timber; the modulus of elasticity represents its stiffness. The strength of small, clear, green sticks cut from these beams is found in Table 11.

It should be noted that the strength values of wood usually quoted in handbooks are based on small, clear, well-seasoned sticks, the strength of which largely exceeds that of large structural timber.

TABLE 1.—Summary of the average bending strength of structural timber.

NOTE.-Figures written as subscripts to the figures for modulus of rupture indicate the number of sticks failing in longitudinal shear.

The moisture condition of the beams varied somewhat, between the different species given in Table 1. The moisture content of green timber also varies with the species-for instance, the maximum is about 37 per cent. of the dry weight in the case of Douglas fir heartwood and as high as 100 per cent. in the case of loblolly pine sapwood, so that the same moisture per cent. in these two woods does not represent an equal degree of seasoning. Again, Douglas fir seasons more rapidly in the dry climate of California than does loblolly pine in the moist climate of the Atlantic coast. It is surprising how much moisture is found in well-seasoned timber. Sticks of longleaf pine 10 by 12 inches in cross section after drying in a lumber yard at Washington, D. C., for one year contained 35 per cent. of moisture, and sticks of loblolly pine from Virginia, 8 by 8 inches in cross section, after drying in the same

place for two years and becoming almost black on the surface, contained 34 per cent. of moisture.

In small sticks the strength begins to increase after the moisture has been reduced to about 26 per cent.* The laws expressing the relation of strength and moisture in the cases of small sticks do not, however, necessarily apply to large sticks. Timbers of commercial size develop checks and other defects while seasoning, and these partially offset the increase in strength due to drying. However, in the case of select sticks the actual strength was in some cases increased from 10 to 25 per cent. by one year of careful seasoning.

LOBLOLLY PINE.

Loblolly pine has not only a wide range of structural merit, but also a wide distribution. It occurs in a belt along the Atlantic coast and the Gulf of Mexico, from Virginia to eastern Texas, extending inland from 50 to 300 miles.

Under the name of Virginia pine the timber cut in the northern portion of this belt is generally found on the markets in small sticks, 8 by 8 inches or 10 by 10 inches in cross section, almost entirely sapwood and of so rapid a growth that sometimes only four rings occur in 3 inches. This is second-growth timber, usually very knotty and of an inferior grade. The same species is also marketed under the name of North Carolina pine, and in that case it is generally forest-grown timber of large size, with a large proportion of heartwood, fairly free from knots, and possessing a high order of structural value.

In the forest loblolly pine is prolific, grows vigorously, and holds its place in competition with other species. It is the principal tree in the operations of those lumber companies in the Southern States which look upon their forest holdings as part of their capital and reap successive crops from them by conservative forest management. It is therefore a timber which engineers and architects may expect to find on the market for an indefinite period. The chief objection to it is that being largely sapwood it decays rapidly when exposed. Because of its open grain, however, it is a wood which may be treated very successfully with preservatives.

Table 2, reference numbers 1-5, gives the results of bending tests on loblolly pine obtained from a mill at Charleston, S. C., and on the market at Washington, D. C.

The loblolly pine listed in the table, under reference numbers 1,

*See Bulletin No. 70 Forest Service, "Effect of Moisture on the Strength and Stiffness of Wood."

2 and 3, was "North Carolina pine," cut from the holdings of the E. P. Burton Lumber Co., near Charleston, S. C. It is a timber that generally shows sapwood on all four faces, and is on this account of a "standard" or "square edge" grade according to the Standard Inspection Rules (1902) of the Georgia Sawmill Association. These rules have been revised since this timber was graded, so as to allow more sap in the merchantable grade. A number of sticks classed as "square edge" in the table would now be graded as merchantable. The timber is good structural material, such as is used in warehouses, mills, and other structures in which the conditions do not demand longleaf pine.

Some of the bending tests (Table 2) were made with the load applied at the center of the span, and the remainder with the load applied at points one-third of the span from each end. The modulus of rupture of the green North Carolina pine beams (Table 2, reference number 1) is 5,580 pounds per square inch, and the modulus of elasticity 1,426,000 pounds per square inch. In the case of the partially air-dried beams containing from 25 to 30 per cent. of moisture (reference number 2) these values are 5,650 pounds per square inch, and 1,435,000 pounds per square inch, respectively. The oven-dry weight of the timber in both of these groups is 31.2 pounds per cubic foot. The moisture per cent. was 27.7 for the partially air-dried and 48 for the green material. Diagrams I and II show the distribution of moisture throughout the cross section of beams under different conditions of seasoning.

In the case of the partially air-dried beams containing less than 25 per cent. moisture (reference number 3), the modulus of rupture is 5,690 pounds per square inch, and the modulus of elasticity 1,340,000 pounds per square inch. The average moisture was 21 per cent., and the oven-dry weight 31.2 pounds per cubic foot. Reference numbers 1 to 3 show that the seasoning ordinarily undergone by large loblolly pine beams has little, if any, effect upon their strength.

The 8 by 8 inch partially air-dried Virginia pine (Table 2, reference number 4) was cut in Stafford County, Va., and had been drying in the yard for two years. This material has a modulus of rupture of 5,180 pounds per square inch, a modulus of elasticity of 1,180,000 pounds per square inch, at a moisture per cent. of 22.4, and an oven-dry weight of 28.8 pounds per cubic foot.

The 8 by 8 inch green Virginia pine (reference number 5), was cut about March, 1903, and was tested about one month after cutting. The timber was "sap stained," but it has been shown that

this staining, or "bluing," does not impair the strength of the wood. The modulus of rupture of these beams is 3,490 pounds per square inch, the modulus of elasticity 744,000 pounds per square inch, the oven-dry weight 26.9 pounds per cubic foot, and the weight as tested. 43.7 pounds per cubic foot.

Table 11, reference number 1, shows a comparison of the strength values of small sticks and large parent beams of loblolly pine. The ratio of the strength values of the large beams to the small beams is 0.77 for the fiber stress at elastic limit, 9.71 for the modulus of rupture, and 0.99 for the modulus of elasticity. Table 12, reference numbers 1 and 2, gives the results of tests in compression parallel to grain. The crushing strength of North Carolina pine, partially air dry, and green, is 4,250 and 3,510 pounds per square inch, respectively. This is higher than the crushing strength of the Virginia pine, which is 2,950 pounds per square inch when partially air dry, and 2,140 pounds per square inch when green.

Table 13, reference number 1, gives the results of tests on loblolly pine in compression at right angles to grain. The average compressive strength at the elastic limit at right angles to grain is 469 pounds per square inch. The material was green, containing 57.1 per cent moisture.

Table 14, reference number 1, gives the results of shearing tests parallel to grain on small blocks. The shearing strength is 630 pounds per square inch, in material containing 83.2 per cent mois

ture.

MOISTURE DISTRIBUTION.

The moisture distribution in the cross section of North Carolina pine beams was determined by cutting sections 1 inch in thickness from near the center of the sticks and dividing the sections into nine parts in the directions shown in Diagrams I and II. The figures in the various parts of the sections show the percentage of moisture.

DIAGRAM I. Distribution of moisture in cross section midway of the length of North Carolina loblolly pine beams. Average of 10 sections taken from sticks 8 by 14 inches by 10 feet. The timber was air dried from 2 to 5 months.