formulas, which are arithmetically equivalent and state the result in percentage of the dry weight:

1. Subtract the last weight from the first weight; multiply the difference by 100; divide by the dry weight, and the result is the percentage of moisture content in the specimen at the first weighing.

2. Multiply the first weight by 100; divide the product by the last weight; subtract 100 from the quotient, and the remainder is the same percentage as obtained above.

In the operation of progressive kilns it is customary to make moisture tests of the lumber only when it gets near the dry end of the kiln, as tests at earlier stages are not particularly useful in governing the control of the kiln. In compartment kilns it is desirable to know the ruling moisture content of the stock at all stages of the process; and indeed the conditions of heat and moisture are changed according to the schedules, which are commonly preferred and used, when the lumber reaches a certain moisture content and not after

FIGURE 59.A short kiln sample is provided in the load which is cut back about a foot when the first sample is taken, and an inch or two before cutting each successive sample. The usual sample is shown at a, while b shows such a sample resawed to separate the shell from the core, so that separate tests may be made on each

an arbitrary time period, thus keeping the schedule in exact alignment with the condition of the stock. For these frequent moisture tests it is not convenient to pull boards out of the piles; and a system of kiln samples is used for the preliminary tests, though the final test of dryness may be made on boards pried out of the side of the load. The kiln sample consists of a short board, 2 to 4 feet long, sawed off of one of the representative boards in the load, and two or three samples may be selected, representative of the heaviest, lightest, and average stock. These samples are put in the side of the load where they can be conveniently withdrawn; and sometimes a thick sticker course is provided in the middle or lower half of the load to give room for the insertion of the sample from the side of the car, or in end-piled loads a special space may be left between boards for the insertion of the kiln sample from the end. (Fig. 60.) It is also best to end-coat the sample with a suitable heavy paint to reduce end evaporation; and even with this precaution the sample will dry differently from the long lumber in the load, sometimes faster and sometimes slower, the difference often amounting to 3 per cent. For

this reason the final test is usually made from a long board; the kiln sample gives results close enough for control of the schedule.

To obtain a test specimen, the kiln sample is removed from the kiln and the specimen sawed from it, some distance from the end, the fresh end again end-coated and the sample returned to its place in the load, to supply further specimens for later tests.

For repeated test work a short cut is often used. When the first specimen is cut from the kiln sample, the remainder is also carefully weighed on accurate scales. It may weigh 10 pounds, and the sample taken at the time may show a moisture content of 25 per cent. Assuming that this is the average moisture content of the

FIGURE 60.-Here are shown recesses built in the side of a kiln load to take kiln samples, which are blocked up to give ventilation from below. Photograph courtesy of Forest Products Laboratory

kiln sample at 10 pounds weight, then its corresponding dry weight would be 8 pounds; and it is only necessary to weigh the kiln sample from time to time to learn its further reduction in moisture content without cutting further small test specimens.

It is possible to avoid figuring percentages by the use of a suitable chart showing the moisture percentage for any combination of original and dry weights within a given range. One of the most comprehensive forms of such a chart has been compiled by John E. Schmidt, kiln operator for the Hammond Lumber Co., Samoa, Calif., and has been copyrighted and published by the Moore Dry Kiln Co., of Jacksonville, Fla., and North Portland, Oreg., which has volunteered to send a copy on good paper to any reader of this report who applies for it, mentioning the report. This is a large

wall chart 38 inches wide and 25 inches high, with green weights from 1.00 to 5.00 running down the side and dry weights from 0.50 to 4.00 running across the top; and the equivalent moisture percentage for any two weights is shown where the line from the side crosses the column traced down from the top of the chart. The weight units may, of course, be any unit of weight provided the two are in the same kind of unit. Two charts of dew point and relative humidity useful in dry-kiln operation also appear. This chart is also available printed on heavy paper with linen cloth backing at $2 a copy, which is stated to be the cost of production in that form.

A scale known as the Troemroid Scaleometer is manufactured by a scale company and sold through various dry-kiln companies; it has two cylindrical charts mounted on the scale, one showing moisture percentages from 2 to 30 per cent and the other moisture content from 30 to 90 per cent. These charts read to the capacity of the scale, for which the test pieces must weigh between 134 and 2 ounces. The scale is doubly calibrated, in common fractions and in decimal fractions.

A number of so-called self-computing scales are sold which give the reading in moisture-content percentage instead of in weight, by various ingenious arrangements. Most of them do not carry the reading to a point of great accuracy; for precise work a good precision scale is preferable.

At one California pine sawmill the dryness of the lumber is judged during the course of kiln-drying by taking hygrometer readings inside the load instead of taking test samples; and whenever this hygrometer reading approaches a certain minimum of relative humidity the lumber is considered to be approaching the desired stage of 9 per cent moisture content. It is apparent that a hygrometer in that position would be influenced considerably by the moisture in the lumber in its effect upon the amount of moisture in the air.

CASEHARDENING TESTS AND SAMPLES

What is known as the casehardening test is really a test for shrinkage differences, which often exist before they have reached the final stage of casehardening. When a sample is cut for the moisture test a similar piece is sawed off-1 or 2 inches. This piece is then split from one edge, on a band saw, into several tongues which are left fastened together at the back. (Fig. 61.) If the board is in a condition of great dryness and shrinkage tension on the outside, the outer prongs will at once curve outward. If drying has proceeded to where the surface has become set in a partly shrunken condition, and the inside has also seasoned to a greater shrinkage than the outside, producing shrinkage tension in the interior, it is the condition known as casehardening, and the outer prongs of the test piece will as soon as cut turn inward because of the shrinkage tension on their inner surfaces. Such a condition may be relieved by thoroughly steaming the lumber, causing the surface portion to become plastic, so that when it again surface seasons it may shrink to a further extent and thus become better equalized with the interior. All these problems of manipulation in kiln-drying are more delicate with some

hardwoods than with any of the softwoods. Figure 62 shows a condition of casehardening stress in a southern pine board that may not have been suspected before resawing it, when the interior tension caused the heavy cup shown.

A kiln sample which is used for weighing in its complete form is, of course, not available for cutting off casehardening samples.

FIGURE 61.-This sketch shows the usual method of cutting a casehardening test section into prongs and the various ways in which the prongs immediately bend as influenced by existing stresses. If the sample is taken in the early stages when the outside is attempting to shrink and restrained by the interior, this exterior tension will pull the outer prongs outward, as at b. A little later interior shrinkage will have balanced conditions and the prongs will remain straight, as at c. When, however, interior shrinkage has gone far enough to produce interior tension which is resisted by the shell, the sample taken will develop interior warping of the outer prongs as at d, the extent indicating the degree of casehardening. All these effects appear promptly after resawing the sample; other changes on further drying after resawing are produced by uneven moisture content and consequent uneven additional shrinkage, the prongs bending toward the side which was least dry at the time of cutting

DRY-KILN CONTROL READINGS AND INSTRUMENTS

The basis instrument for determining heat and moisture conditions in a dry kiln is a hand thermometer, or rather two thermometers on one mounting; and even when registering or recording instruments are used the hand instrument is still basic for testing the calibration on the more complicated and expensive instruments and keeping them adjusted to correct readings.

One of the pair of bulbs in the hand instrument is an ordinary thermometer. The matched bulb differs in that it is covered with

a wick leading into a water cup on the mount, which keeps the wick constantly wet. Evaporation is constantly occurring from this damp wick, and the latent heat absorbed in the process has a cooling effect on the thermometer bulb, which therefore registers a lower temperature than the dry bulb. The rate of evaporation and therefore the cooling effect depends upon the temperature of the air to which the wet bulb is exposed (hotter air, of course, causing a faster evaporation) and upon the percentage of relative humidity in the air (drier air also causing faster evaporation). At any given dry-bulb temperature, a given difference of reading on the wet-bulb thermometer indicates a given relative humidity in the air, which can be read from a chart covering the desired range of variation in both readings. The two factors necessary to an accurate reading are that there should be enough air circulation when the reading is taken so that the wet bulb will not become blanketed by a layer of stagnant air, and that the wet-bulb wick should not become clogged with deposits from the evaporated water, thus reducing the evaporative effect from it.

Such an instrument will, of course, change its reading very quickly upon being taken out of the kiln and is difficult to read in place inside the kiln. Another refinement therefore consists in using tubes with a constriction in the bore sufficient to prevent the liquid falling in the tube as it contracts in the bulb, so that the maximum reading secured inside the kiln remains evident until the liquid in the tube is shaken down again.

Where a number of hand hygrometers are used for dry-kiln control safety requires that they be checked about every six months for possible changes in calibration, as thermometers change somewhat with gradual molecular changes in the glass, or the tube may get out of proper adjustment with the scale when this is not etched directly on the glass. A standard chemical thermometer, graduated in Fahrenheit scale up to 220°, may be used as a standard, both thermometers being immersed in water of various temperatures to compare the readings; and wet-bulb thermometers under such conditions, of course, give the same readings as dry. With a hand thermometer correction of the calibration is usually impractical; instead, the amount of error at various temperatures is ascertained and recorded and therefore added to or subtracted from readings made with that instrument. Recording thermometers usually have a general adjustment for calibration to bring it fairly in line, though this may still leave local errors at some points of the scale. An error of several degrees in thermometers might lead to added degrade in the product and would be far from its best operating condition.

It is, of course, important that the most severe conditions be known in any kiln if degrade is to be minimized. Usually, however, the readings taken are representative of more or less average air conditions inside the kiln, so as to record average air conditions in the kiln proper. The reading is taken at a point convenient to reach from the door. The instrument may be hung on a bracket holding it out 12 inches or more from the wall, or set on the lumber in the nearest load. The air reaching the instrument may have been cooled somewhat by the nearness of the door or wall, or it may be that the air current at that point is chiefly of air that has already passed through the lumber and perhaps is a little cooler and moister than