ILLUSTRATIONS. Page. FIG. 1. Comparative rates of seasoning of cross-arms composed of varying amounts of heartwood and sapwood.... 2. Common method of piling cross-arms in commercial practice.... 3. Modified form of piling cross-arms in commercial practice... 12 14 14 4. Ideal form of piling cross-arms, 20 by 20, giving free circulation of air. 5. Comparative rates of seasoning of cross-arms piled 28 by 28 and 20 by 20.... 15 16 6. Recommended method of roofing cross-arm piles with boards..... 17 17 8. A 20 by 20 pile of cross-arms with a roof constructed of the arms themselves-not recommended............. 18 9. Comparative rates of seasoning of soaked and unsoaked cross-arms... [Cir. 151] 19 (4) THE PRESERVATIVE TREATMENT OF LOBLOLLY PINE CROSS-ARMS. INTRODUCTION. THE DEFICIENCIES OF THE PRESENT METHOD. All who are familiar with wood preservation know that the ease with which the preservative can be injected depends upon the character and condition of the wood. Porous species, such as red oak and certain pines, absorb the liquid more readily than such dense woods as tamarack and hemlock. In a single species, timber cut from opengrained fast-grown trees takes treatment more readily than that from trees which grow slowly, with a consequent close grain; porous sapwood receives the liquid more readily than the denser heartwood, the pores of which are more or less filled with oils and other infiltrated organic substances; and wood is easier to treat when dry than when a large amount of water in the cells presents a mechanical resistance to the entrance of the preservative. In small timbers, which receive an absolute penetration, the difference in absorption is more apparent than in the treatment of large dimension stuff, merely the outer layers of which are impregnated. The cross arms manufactured throughout the South are cut mainly from rapidly grown, open-grained loblolly pine. No distinction is made as to species, however, and shortleaf and longleaf are frequently included. At present cross arms cut from various parts of the tree are mixed indiscriminately; hence cross arms cut from heartwood are treated with those cut from sapwood, and those which have reached an air-dry condition with freshly cut or rafted material. As a consequence, the amount of the preservative forced into the wood varies greatly; the dry sapwood material receives more than is required and the heartwood not enough. PRELIMINARY TESTS. These considerations induced the Forest Service, in February, 1903, working in cooperation with the American Telephone and Telegraph Company and the Norfolk Creosoting Company, to make a series of [Cir. 161] six experimental runs at the latter company's plant near Norfolk, Va. The principal objects of this test were to demonstrate the inequality of the treatments received by different classes of material and to furnish a basis for more detailed work in the future, with a view to rendering the treatments more uniform and economical. For creosoting under pressure, the timber is piled in small trucks or buggies, hauled into large, air-tight horizontal cylinders, and the doors tightly closed. Live steam is then admitted for varying lengths of time, usually for from three to six hours, according to the character and dryness of the wood. The steam is then expelled and a vacuum drawn for several hours to exhaust the air and water remaining in the wood cells. Creosote (dead oil of coal tar) is then allowed to run into the cylinder from one of the storage tanks. When the oil appears in the overflow pipe, indicating that the cylinder is full, force pumps are started, and the resulting pressure is maintained in the cylinder until the desired amount of oil per cubic foot of timber has been forced in. This amount has previously been calculated for the entire load, and when the indicator on the storage tank shows that the required amount of oil has been drawn from the tank the supply is shut off, and that which remains in the cylinder is forced back into the storage tank. The treated load is left to drip in the cylinder for a short time, and is then hauled out and unloaded. GRADING THE ARMS. Before the treatment was begun 14,000 arms were separated into three classes, in accordance with the relative amounts of heartwood and sapwood each arm contained. Class A was composed of arms with 75 per cent or more of heartwood; Class B, 75 per cent or more of sapwood; and Class C contained less than 75 per cent of either heartwood or sapwood. No attempt was made to separate the arms according to the amount of moisture they contained, hence nearly green and partially air-dry arms were represented in the same class. Six runs were made in these preliminary experiments. In order to ascertain the exact absorption of oil, a number of arms of each class were weighed before treating and so marked that they could readily be identified and weighed after the treatments were finished. A record of the runs is given in Table 1. [Cir. 151] Load No. 1 was made up in accordance with specifications which may be taken fairly to represent ordinary commercial practice. No attempt was made to restrict the load to any one class of arms, but, on the contrary, all three classes were represented; and care was taken to make the load as nearly as possible representative of the usual conditions for treatment. The preliminary steaming of four hours was at about 20 pounds pressure to the square inch. Twelve pounds of creosote to the cubic foot was required. In the same load the variation in the absorption of oil was in some cases from nearly 27 pounds per cubic foot, the maximum for the sap class, to about 1 pounds, the minimum for the heart class; a total difference of over 25 pounds per cubic foot. The average absorption per cubic foot for the sap class exceeded the specified amount by about 5 pounds, whereas that of the heart class was over 7 pounds too light. Even the maximum amount absorbed by any one heart arm was 1 pound less than the specifications called for. The sap class absorbed nearly twice as much oil as the mixed class and 3.7 times as much as the heart arms. Load No. 2 was made up of partially seasoned arms. The customary steam bath was omitted, and the wood was warmed merely by passing superheated steam for thirty minutes through the coils in the bottom of the cylinder. The absorption of this load averaged more than 3 pounds per cubic foot in addition to the amount required by the specifications. Load No. 3 was made up of material similar to that in load No. 2. It was given a bath of live steam for three hours. The steam was then blown off and the temperature maintained for one hour by the superheated steam in the coils. A four-hour vacuum was drawn before the oil pressure was applied. The absorption of this load averaged over 5 pounds per cubic foot more than the specifications require. It should be noted that the time required for this load (treated according to the regular specifications) was eight hours and twelve minutes, whereas load No. 2 required only two hours and thirty-one minutes. Load No. 4 was composed of arms representing all three classes. The preliminary treatment consisted of six hours steam, one hour blow-off, and five hours vacuum. That the wood absorbed the oil comparatively slowly is shown by the fact that a pressure for fiftyfive minutes at 50 pounds was required to force the desired amount of oil into the wood. By reference to the table it will be noted that the variation in the amount of oil absorbed by the several grades in this load is comparatively slight. This may be due to the fact that live steam has a greater influence upon green heartwood than upon heartwood partially seasoned. On the other hand, sapwood becomes more receptive of oil when air seasoned than when subjected to the bath of live steam. Load No. 5 was composed of partially seasoned heart arms. The wood absorbed the oil very slowly and showed the difficulty of forcing large quantities of oil into the dense heartwood. Twenty-four per cent of the total amount of oil absorbed during the entire treatment was taken up by the wood within fifteen minutes after the oil was first turned in and before the pumps were started. It required nearly ten times as long to force the remaining 76 per cent of oil into the wood as it did for the wood to absorb without pressure the 24 per cent. The average absorption was 10.3 pounds per cubic foot, as compared with 4.76 pounds for the heart arms in load No. 1, in which the arms of the three classes were mixed indiscriminately. Load No. 6 was made up of partially seasoned heart and sap arms (Class C). The average absorption was 12.07 pounds per cubic foot. The variation between the minimum and maximum absorption per cubic foot was due to the fact that in some arms the percentage of heartwood largely exceeded that of the sapwood, and vice versa. It is therefore evident that the maximum amount was absorbed by an arm having a large percentage of porous sapwood, while the minimum amount was taken by an arm in which the heartwood predominated. These preliminary tests, although necessarily incomplete, served to bring out strongly several important points: They clearly demonstrated that the present system of treating heart arms and sap arms together, without regard to the amount of moisture they contain, results in an excessive and detrimental absorption of oil by the sapwood and an insufficient penetration of the heartwood; that a lighter treatment than 12 pounds of oil to the cubic foot is probably sufficient for sapwood; that an economy could be effected by regulating |