years. One piece, 3 in. x 14 in., was nearly as dry after 21⁄2 years' seasoning as after 51⁄2 years. These blocks were stored in the loft of a blacksmith's shop which was unheated except for the fire of the forge. The Forest Service records show losses in the weight of Western larch bridge stringers (8 in. x 16 in. in section) throughout a period of 31⁄2 years and of Western hemlock and redwood throughout a period of 3 years (Figs. 44 and 46).

Losses of weight, especially with large timbers, are so gradual after the greater portion of the moisture has been evaporated that it is impossible to fix any particular moisture per cent. to be designated "air dry." The air dry condition depends, of course, on the humidity of the atmosphere; it changes for each climate and season, and varies from day to day.

The moisture contents of small air dry blocks (11⁄2 in. x 11⁄2 in. x 8 in.) of various species of wood determined at New Haven, Conn., are shown in Table 10. These blocks had been seasoned for over a year under cover of a shed and the determinations were made during a period of clear weather.* Under the conditions of this test, the air dry wood contained from 13 to 15 per cent. moisture. In a drier climate the moisture content may go as low as 12 per cent. or lower.

TABLE 10.-MOISTURE CONTENT OF SMALL BLOCKS THOROUGHLY AIRSEASONED AT NEW HAVEN, CONN.

For most uses structural timber would be considered air dry when it had lost 75 per cent. or more of the total moisture loss possible by air seasoning. As a rule the time required to reach this condition would not be excessive for commercial practice.

It is apparent from the information here presented that, as a rule, timber weighing considerably more than the theoretical air dry wood must be considered "air dry" and used.

SEASONING AFTER TREATMENT.

Fig. 54 shows the gain in weight of lodgepole pine ties during treatment with 3 per cent. (approximately) zinc chloride solution and the

The specimens were of uniform size and were lying side by side on a rack during the entire time of seasoning. The half-inch discs from which the moisture determinations were made were cut one-half inch from the ends of the blocks on October 14, the weather having been clear for five consecutive days previously. There were two specimens of each species, the figures given being the average of the two.

loss during subsequent seasoning. The ties, which were air seasoned when placed in the cylinder, were subjected to steaming as a part of the treating operation, and gained, on an average, approximately 60 Its per tie. When last weighed, a little more than 2 months after they were treated, they were 3 lbs. heavier per tie than before the solution was injected. The same test was applied to ties treated without steam; the result was the same except that the increase in weight was 10 s. less per tie. All of this increase was lost by the end of the seasoning

Fig. 54-Gain in Weight of Lodgepole Pine Ties During Treatment
With Zinc Cchloride Solution and Loss During
Subsequent Seasoning.

period employed in the first case. The results in both cases are based on two-truck loads of from 28 to 30 ties each.

Red oak ties, treated at the Forest Products Laboratory (Madison, Wis.), lost 61 per cent. of the weight gained during treatment. These ties were treated with approximately 46 lbs. of 3 per cent. zinc chloride solution per tie and were piled about 5 months in the winter and spring (from February to June). Hard maple ties under the same conditions gained 63 lbs. (2.5 per cent. solution) per tie and lost 78 per cent. of this amount. There are, however, no data which afford a comparison between these losses and the losses from green untreated ties under similar conditions.

Under some conditions at least the zinc chloride apparently retards the evaporation of water very appreciably, as shown by the fact that loblolly pine ties treated with a 2.5 per cent. solution and seasoned 8 months at Lafayette, Ind., then weighed 3.1 lbs. more per cu. ft. than before they were treated.*

Forest Service Circular 39.

SHRINKAGE.

The drying of wood is accompanied by a shrinkage of its volume which begins usually when all water has been evaporated* from the cell cavities, and the cell walls themselves begin to dry out. When this condition is reached, the moisture content is, as a rule, less than 30 per cent., but the moisture content in a large stick is not evenly distributed and the outer portions dry first, so some shrinkage occurs almost as soon as seasoning begins. This is seen in Fig. 55, which shows the per cent. of green area in cross-section of Douglas fir, Western hemlock and Western larch beams as compared with loss of moisture. It will be noted that the reduction of proportional green area becomes more pronounced as the beams approach an air dry condition.

Table II shows the linear shrinkage in the radial and tangential directions for small blocks of a number of species; the shrinkage from the green to the air dry, or approximately air dry condition, and from the green to the oven dry state is shown separately. In all cases the greater part of the shrinkage occurred after the blocks were placed in the oven. These points are significant since they show that partial air seasoning has very little effect in preventing the subsequent shrinkage of timbers, and that complete air seasoning is not sufficient if the wood is later to be subjected to further drying, as by use in artificially heated structures.

Shrinkage tangentially is nearly twice as great as radially. Longitudinal shrinkage is so small that it may be disregarded. The shrinkage in circumference of air-seasoned poles was extremely small, being less than 1 per cent.; this was due largely to the fact that the poles were not sufficiently dry when the tests ended to cause much shrinkage, and it was due also, perhaps, partly to the checking which occurred.

TABLE 11.-RADIAL AND TANGENTIAL SHRINKAGE OF VARIOUS SPECIES*

Average moisture)

content

Species

No.of Size of
tests specimen

Green

Air dry

Average Shrinkage

Green to air-dry Green to oven dry
Radial Tangential Radial Tangential

Inches Percent. Per cent. Percent. Per cent. Percent. Per cent.

*Tests made by the Forest Service at the Seattle, Washington, Laboratory.

*Eucalyptus, which begins to shrink at once with any loss of moisture from the green wood, seems to be an exception to this rule. Forest Service Bulletin 70 and Circular 108.