to this fact and the custom of sawing ties from coarse heartwood timber, the seasoning experiments on Douglar fir show a small loss of moisture in per cent of green weight as compared with other species.

Four hundred ties were cut each month of the year from November, 1903, to December, 1904, except that in September, 1904, none were cut. Of these, 200 ties were marked, weighed, and shipped each month to Pasco. A new series was started at Tacoma in December, 1904, and continued until July, 1905, to act as a check on the first series through the rainy season. As a rule, the ties were weighed within three and one-half days from the time the trees were felled; hence the first weight is the approximate green weight.

A 7 by 2 pile was adopted at both Tacoma and Pasco, with 50 ties in each pile. Two piles of each series of 200 ties were faced north and south and two piles east and west.

A monthly meteorological table (Table 1), prepared by the United States Weather Bureau station at Tacoma, was used to show whether the ties seasoned under normal conditions and to determine what effect climatic changes had upon the amount and rate of seasoning. No such data were obtainable at Pasco.

The green weight of Douglas fir, based on the data obtained from 4,000 ties seasoned at Tacoma, shows a seasonal variation of 4.7 pounds per cubic foot. The timber weighed 37 pounds per cubic foot in July and 41.7 pounds in February and March. A thousand feet, board measure, of the July timbers weighed 3,082 pounds, of the February and March timbers 3,474 pounds. The yearly average was 3,284 pounds per thousand feet. These results correspond very

closely to the rough green weight of 3,300 pounds adopted by the lumber associations in the Northwest.

Table 2 shows the green and seasoned weight per cubic foot and per thousand feet, board measure, of Douglas fir. The seasoned weight is based on eleven months' seasoning. The column on the right gives the time required for the various months' cuts to reach their minimum weight.

TABLE 2.-Green and seasoned weight of Douglas fir, Tacoma, Wash., based on eleven months' seasoning.a

a Compare time required to reach minimum weight with the amount of rainfall during that time in meteorological table.

The oven-dry weight is 28.3 pounds per cubic foot.

POSITION OF THE PILE WITH REGARD TO WIND DIRECTION.

The seasoning experiments at Tacoma and Pasco showed that there is little difference in the rate of seasoning of timbers stacked in piles which face different directions. The greatest difference for any single month was 2 per cent, while the average difference for a year's cut, based on eleven months' seasoning, was 0.05 per cent at Tacoma and 0.02 at Pasco.

In view of the slight difference between the two forms of piles, the results at both Tacoma and Pasco have been based on a total of 200 ties for each month.

RATE OF SEASONING.

The oven-dry weight of Douglas fir used as a basis in this experiment was taken from data established by Professor Hunt at Berkeley, Cal. His results show a variation from 22.1 to 36.2 pounds per cubic

a The weight determined by drying small pieces of the wood in an oven at 100° C. until a constant weight is obtained. For further discussion of this see Forest Service Circular 115, Second Progress Report on the Strength of Structural Timber.

34211-Cir. 146--08-2

foot. The average for all grades, 28.3 pounds per cubic foot, was taken as the basis.

The standard dimensions of sawed ties are 7 by 9 inches by 8 feet, and therefore each tie contains 3.5 cubic feet. Several hundred of the experimental ties were carefully measured and found to conform with the specifications; hence, the weight of each tie was divided by 3.5, the number of cubic feet it contained, and all results expressed in weight per cubic foot.

The results given in Tables 3 and 4 furnish a comparison of the maximum and minimum rates of seasoning of Douglas fir under two entirely different climatic conditions. While the weight per cubic foot and moisture content of the green timber seasoned at each station was practically the same, there is a wide difference in the rate of seasoning. On account of the uniform rate of seasoning of certain months' cuts at Pasco, it has been possible to make four groups and thereby simplify the results. The following four groups were therefore made: (1) October, November, December, and January; (2) February, March, April, and May; (3) June and July; (4) August and September." It is impossible to group the monthly cuts at Tacoma.

TABLE 3.-Rate of seasoning of Douglas fir ties, Pasco, Wash.

a No cut was received during September, but the results would undoubtedly correspond with those of August.

b Oven-dry weight is 28.3 pounds per cubic foot.

[Cir. 146]

TABLE 4.-Rate of seasoning of Douglas fir ties, Tacoma, Wash.

Three important conclusions may be drawn from seasoning timbers under different climatic conditions: (1) Timbers cut during different months will eventually reach approximately the same weight per cubic foot in regions with a minimum rainfall; but in humid climates the maximum loss in weight corresponds closely to the seasonal variation in the green weight. These results failed to substantiate the theory that the varying physical characteristics of timber cut at different seasons determine the minimum air-seasoned weight. (2) Rainfall is the chief factor in determining the rate of seasoning. (3) The rate of seasoning determines, to a great extent, the resultant behavior of the timber as to checking.

CHECKING.

In order to determine the effect of the rate of seasoning upon the condition of the timber, frequent photographs were taken of each month's cut. Two kinds of checks were observed-radial and small honeycomb. Honeycomb checks occur on timbers which season slowly during the first few months. Radial checks occur on timbers

which season rapidly during the first few months, that is, timbers which are cut during the period from June to September. A study of these two forms of checks indicates that honeycomb checks cause little damage to the tie, while radial checks are very destructive and often make the tie unfit for service. Because of the excessive checking of Douglas fir, many eastern concerns will not use this timber in large dimensions. In view of these seasoning experiments, it is believed that this objection can be eliminated and the market of Douglas fir extended. Cutting should be limited to the months from November

to April, inclusive, and the timber seasoned for at least ninety days before being shipped through the arid region east of the Cascade Mountains.

TREATMENT OF TEST TIES (CREOSOTE).

Four hundred and fifty of the ties seasoned at Tacoma were selected for treatment. These were treated by the straight creosote method. After treatment the ties were again seasoned four months before being placed in the track.

In order to determine the absorption by increase in weight the ties were weighed before and after treatment, then ninety days later. The increase in weight during treatment was 8.6 pounds per cubic foot. After ninety days seasoning an average of 2.4 pounds of this increase had been lost, leaving the net increase per cubic foot 6.2 pounds. When the ties were placed in the track they contained an average of about 6 pounds of creosote per cubic foot.