immersion for different periods up to thirty days. The general results are shown in figure 9.

0

20 40 60 80 100 120 140 160 TIME SEASONED-DAYS

FIG. 9. Comparative rates of seasoning of soaked and unsoaked cross-arms. However, it should be remembered that in these experiments the longest period of immersion was one month, since it was not believed

that the beneficial effect of soaking could be such as to warrant longer immersion. Had the arms been soaked for several months it is possible that wholly different results would have been obtained. So far as the present purpose is concerned, however, such results would have been of little practical value.

It may therefore be concluded that immersion before seasoning is not justified if it involves additional time and expense. The beneficial effects on subsequent seasoning, except, possibly, where the soaking is continued for prolonged periods, are slight; and soaked arms do not appreciably surpass unsoaked ones in the absorption of preservative.

SHRINKING, CHECKING, AND WARPING.

Shrinking, checking, and warping, which commonly take place during the seasoning of timber, are due to variations in the moisture content of the wood. A cubic foot of green sapwood of loblolly pine, weighing about 60 pounds, contains approximately 32 pounds of water. In the seasoning of the timber to an air-dry condition, to a weight of, say, 34.5 pounds per cubic foot, about 25.5 pounds of water are evaporated. This great change in the moisture content manifests itself by reducing the bulk of the timber. Generally this is accompanied by more or less checking and warping.

THEIR EFFECTS ON INSULATOR-PIN HOLES.

In Bulletin 70 of the Forest Service it is explained that water may exist in wood in two states, either as imbibed moisture, which is absorbed by the cell walls, or as free water, which merely fills the pores or cavities of the wood like honey in a comb. In wet and in green wood the water exists in both conditions. In the process of seasoning the free water is evaporated before the amount of moisture in the cell walls is reduced, and as its evaporation brings about no change in the substance of the wood, no shrinkage or other effect takes place. According to one theory the cell walls are made up of exceedingly minute particles of material called mycellæ, which are invisible even by the aid of a microscope. When comparatively dry wood is subjected to moisture the water readily passes into these mycellæ and forces the particles farther apart, causing a consequent increase in the bulk of the wood. Similarly, when the evaporation of the free water is complete, the moisture begins to be drawn from the cell walls, the particles are drawn together, and shrinkage takes place.

• Effect of Moisture on the Strength and Stiffness of Wood, Tiemann.

To determine whether the shrinkage in the cross-arms during air seasoning is sufficient to reduce appreciably the diameter of the holes for the insulator pins, measurements were made on several hundred arms of the heart, sap, and intermediate classes. It was found that for all practical purposes this shrinkage can be disregarded.

The objection to air seasoning is sometimes made that, although the area of the holes is not materially reduced, the shape is so distorted as to prevent the ready entrance of the pin. Theoretically, there is some basis for this fear, as shrinkage across the grain is much greater than longitudinally, and hence there might be a tendency for the holes to assume an elliptical shape. Close observation, however, failed to show that this possibility need be considered in practice.

TREATING.

The object of the experiments in the treatment of the cross-arms was to devise a process which would insure a uniform, efficient, and cheap impregnation.. The preliminary work had shown that under the system in common use the sap-wood portions of the arms received an excess of oil; this not only added to the cost of treatment but proved an actual detriment, because the oil exuded in warm weather and dripped upon passers-by. On the other hand, the heartwood portions of the arms received a merely superficial penetration. It is true that heartwood is naturally more durable than sapwood, yet this natural durability is by no means equal to that of sapwood thoroughly impregnated with a good preservative. This is especially true in such species as loblolly pine, because the heartwood is by no means durable. As the surface of most of the arms is composed of both heartwood and sapwood, it is apparent that both classes of wood should be thoroughly preserved. The decay of the heartwood portion, no matter how small, in most cases renders the entire arm unfit for service.

In the preliminary treatments it was shown that the preservative is absorbed much more readily into the sapwood portions than into the heartwood, and that the application of pressure is necessary to force the preservative into the more impervious heartwood. This pressure may be applied by force pumps, as in the closed-cylinder processes, or by the atmosphere, as in the open-tank method. In either case the sapwood is thoroughly saturated with the preservative before the heartwood has received more than a superficial painting. In order to obtain a uniform penetration, without excess of oil, it is necessary to continue the pressure until the heartwood has received the desired penetration, and then to extract and recover the superfluous oil from the treated wood. The finding of a method to do this constituted the principal object of the experiments. The first series

of experiments was undertaken to secure the desired penetration. The extraction of the superfluous oil constituted the object of the latter tests.

EFFECT OF THE PRELIMINARY PROCESSES OF LIVE STEAM AND VACUUM.

The preliminary bath of live steam, which is commonly applied in the treating plants of this country, is undesirable from several points of view. It is expensive; it adds from three to six hours to the length of the treatment; and, unless carefully controlled, there is serious danger of its detracting from the strength of the timber. There is also reason to believe that the seasoning it promotes, even in green timber, has been greatly overrated. In the case of seasoned and partially seasoned loblolly pine, experiments showed that the steam bath uniformly increased the moisture content of the wood. The following is one of a number of experiments which were made to test this point:

Ten 5-foot cross-arms were selected, each containing about 0.45 cubic foot. Three species of southern pine were represented-longleaf, shortleaf, and loblolly. The longleaf and loblolly arms were made up of both green and seasoned timber. All the arms were weighed immediately before treatment and placed in an experimental cylinder so arranged that all factors in the treatment could be carefully controlled. Live steam was then turned on and maintained for three hours at 20 pounds pressure. To assist in raising the temperature within the cylinder, steam was circulated through the heating coils arranged along the bottom. At the end of the steam bath the arms were quickly withdrawn and reweighed. The results are shown in Table 6.

TABLE 5.-The effect of live steam upon the moisture content of pine cross arms.

In every case the steam bath increased the moisture content of the wood. This was true in the green arms as well as in those which had been seasoned to an approximately air-dry condition. The strength of partially seasoned timber is decreased up to a certain

point in proportion to the increase of the moisture content. As
creosote retards the evaporation as well as the absorption of water,
the original strength of the arms is not regained for long periods.

The principal benefit of the steam bath, at least with such species
as loblolly pine, seems to consist in the thorough warming and expan-
sion of the wood structure. The poor heat conductivity of wood,
however, renders it doubtful whether more than the outer layers can
be affected by steaming at safe temperatures and for ordinary lengths
of time.

Moreover, the added moisture in the wood fibers tends to retard
the absorption of the creosote. Probably the same objection can not
be raised if aqueous solutions are used as preservatives, but with oils
it is important that the fibers should be as nearly dry as possible.
This view was substantiated by the result of the experimental treat-
ments, which showed that the penetration in the steamed arms did
not average as high as in the arms which had not been steamed. The
durations of the steam baths varied from one to ten hours at 20

pounds pressure. The results were comparatively uniform, and
showed little additional effect from the longer steaming periods. The
above observations, however, do not necessarily apply to a wood of
extremely resinous content, such as longleaf pine "fatwood." The
effect of steam on such woods involves wholly different considerations.
It was also found that the preliminary vacuum could be abandoned
without appreciably affecting the ease with which the timber absorbed
the preservative. For timber as porous as air-dry loblolly pine the
three or four hours which are usually required for the vacuum can
be saved, with a considerable reduction in the expense of the treat-
ments. Indeed, it is probable that the effects of the vacuum as a
preparatory process have also been overestimated. However, even
in the case of air-dry pine, the vacuum is desirable as furnishing an
effective means for quickly drawing the preservative from the storage
tanks to the treating cylinder.

Up to this point the experimental treatments had shown (1) that
cross-arms of air-dry sapwood loblolly pine will quickly become thor-
oughly saturated with warm creosote (at about 125° to 140° F.)
without the application of pressure; (2) that to saturate the heart-
wood it is necessary to apply pressure, and (3) that the steam bath
and vacuum are unnecessary as preparatory processes.

A large saving in the time required for the treatment had been effected, but the more important point of bringing about an economy in the use of oil yet remained to be accomplished. This last object. was fully attained by drawing a vacuum after the desired penetration had been secured.