Search Documents

By J. M. Davidson

The potassium mineral polyhalite (K2SO4.MgSO4.2CaSO4.2H2O) found in large deposits in western Texas and southeastern New Mexico is associated with varying amounts of halite (sodium, chloride, NaCl) and anhydrite (calcium sulphate, CaSO4). Although the anhydrite, because of its low solubility and slow rate of solution, is relatively inert, the halite is dissolved immediately in any process solution with which the polyhalite comes in contact. Limited quantities of sodium chloride in the liquors used for the hot extraction of polyhalite materially improve the results obtained, but only a very small sodium chloride content may be desired in the entering polyhalite in other procedures, such as Bureau of Mines process. In all of the processes suggested for the recovery-of potassium sulphate from polyhalite the control of the sodium chloride content will be desirable, if not absolutely essential to efficient operation.

Jan 1, 1934

By Irwin Roman

IT has been known for many years that when a wire is moved in a magnetic field, an electromotive force is developed which is proportional to the rate at which the wire is moved in a direction perpendicular to the magnetic intensity and to the magnetic intensity itself. If the ends of the wire are connected to a current indicator, the current behaves exactly as though the source of electromotive force were a galvanic cell. If a wire is wound circularly and rotated uniformly around a diameter not parallel to the intensity, the current flowing through the wire will be alternating of sinusoidal shape. If, instead of a single loop, there are several turns wound in a short helix, the electromotive force developed will be approximately proportional to the number of turns, the area of the helix and to the speed of rotation. The maxi-mum electromotive force will be devel-oped when the axis of rotation is perpen- dicular to the field; if this is not so, only the component of the intensity that is perpendicular to the axis of rotation is effective. This is the principle of the usual earth inductor. With a few turns of wire and comparatively slow rotation, FIG. 1.-MAGNETIC GRADIOMETER. it is possible to produce a considerable deflection of a galvanometer. If two similar coils are rotated together at two different points for which the magnetic field intensities perpendicular to the axes of rotation are different, the electromotive forces developed will each depend on its corresponding intensity and be a measure of that intensity. If the inten-sities are sufficiently different so that the electromotive forces may be compared, the two intensities may be compared. This is the principle

Jan 1, 1934

By R. H. Lynn, T. C. Craig

During the year 1933, in the Texas Panhandle, 113 oil wells were completed, adding 33,337 bbl. to the daily potential of the field. The field potential on Jan. 1, 1933, as determined by the Texas Railroad Commission, was 149,260 bbl., with a daily allowable production of 41,000 bbl. The potential was raised gradually, until on July 16, 1933, it reached 155,344 bbl. with a daily allowable of 55,000 bbl. This potential and allowable was maintained over a period from July 16, 1933, to Sept. 8, 1933, when the potential was reduced to 136,893, with a daily allowable of 43,727 bbl. Since Sept. 8, 1933, there has been a gradual increase, until on Jan. 1, 1934, the field potential reached 159,740 bbl. with a daily allowable of 42,000 bbl. Of this amount 26,429 bbl. arc "marginal oil" and 2682 bbl. are exempt on account of water. Deepening of Wells During the year 64 wells were deepened, increasing their total oil production 9631 bbl. per day. Of this amount the Gray County granite wash contributed 9256 bbl. from 57 wells. It has been found in many places that the granite wash carries an additional oil pay from 50 to 75 ft. below sea level. The majority of the wells that were .drilled during the early development in Gray County encountered their oil above sea level. When the wells producing from the upper horizon have become unprofit-

Jan 1, 1934

By Samuel Lasky

THE writer has been interested in determining the mineralogic dis-tribution of silver in the base-metal ore of the Ground Hog mine of the Asarco Mining Co. in the Central mining district of New Mexico.1 This ore consists of a varitextured mixture of sphalerite, chalcopyrite, galena and pyrite in a quartz gangue. The average ore contains 14 per, cent of zinc, 9.5 per cent of lead, 5 per cent of copper, and 10 oz. of silver per ton. Generally it is accepted that the silver in ores of this type, in which no definite silver minerals can be recognized, is contained in the base-metal minerals. Published assays of pure mineral specimens from many dis-tricts show that each of the base-metal minerals contributes in some degree to the total silver content of the ores, and to determine the pro-portionate amount contributed by each mineral in any specific ore it has been the practice to assay selected specimens of each mineral, the purity of which has been determined microscopically. There being no such assays available for the Ground Hog ore, and it being impract cable to have enough of them made to give average results, the writer resorted to a mathematical attack upon composite assays of large quantities of ore and concentrates. Surprising results were obtained, and it is the threefold purpose of this paper to present these results, to describe the mathe-matical method used, and to point out the value of using this method even when assays of pure specimens are available.

Jan 1, 1934

By H. C. Howarth

This paper gives observations on ground movement and subsidence resulting from pillar drawing in the Lick Run section of Montour No. 10 mine of the Pittsburgh Coal Co. adjacent to solid coal owned by the U. S. Bureau of Mines at its Experimental mine in Snowden Township, Allegheny County, Pennsylvania. The investigation was planned by George S. Rice, chief mining engineer of the Bureau, who arranged the necessary cooperation with Dr. L. E. Young, vice president of the Pittsburgh Coal Co. Engineers of that company placed permanent monuments on the surface to establish the boundaries of the coal owned by the Bureau and to furnish fixed points to which all levels were referred. Engineers of the Experimental mine staff then covered the ground under observation with a checkerwork of monuments made of oak pegs set into the ground at 50-ft. intervals in two directions at right angles to each other. The level of these pegs was determined first from Dec. 16 to 30, 1929; again from May 25 to June 5, 1930, and finally from July 7 to 15, 1931. Observations were then discontinued because work had been stopped temporarily in this part on Montour 10 mine. General Conditions Fig. 1 is a plan of the part of the area under observation with which this paper deals. The mine is developed on a room and pillar system known locally as half advancing and half retreating. As Nos. 35 and 36 butt entries were advanced, rooms were driven from No. 36 and drawing of the room pillars followed their completion. When the entries reached the boundary, rooms were started from No. 35 butt beginning at that point, and all remaining pillars were drawn as rapidly as possible. The rib lines of Jan. 9 and June 13, 1930, were surveyed a week to 10 days after the first and second sets of levels were run, and there was some advance in that period. The rib line of July 15, 1931, agrees in time

Jan 1, 1934

By John Devine

DURING the past few years the Bureau of Mines has been studying hydrogen-sulfide corrosion in the petroleum and natural-gas industries. Early work was confined to investigating the various practical, or operat-ing, factors that control the corrosion reaction when ordinary mild steel is being attacked. In that work it was shown that, if the optimum operating conditions1 are obtainable, . gas containing hydrogen sulfide can be handled at ordinary temperatures without serious corrosion resulting. However, frequently "sour" gas must be handled under other than optimum operating conditions, and combating corrosion then becomes a problem to be considered more during the construction of a system than during its subsequent operation. That is, where economi-cally practicable in such cases the designer will seek to substitute for mild steel other steels or ferrous metals more resistant to corrosion by hydro-gen sulfide. The comparative resistance of various commercial steels and alloy steels to corrosion by hydrogen sulfide at elevated temperatures has been reported systematically by previous investigators,2 but similar information for corrosion at ordinary temperatures is available only from scattered sources, representing different methods of testing. As the value of comparative corrosion tests depends largely upon the use of a

Jan 1, 1934

By G. Hileman

At present there are three fields in commercial exploitation in the Argentine. In order of production they are: (1) Comodoro Rivadavia, (2) Salta, (3) Plaza Huincul. Comodoro Rivadavia.—This field is on the coast of the Territory of Chubut in the southern part of the republic. It is being developed and exploited by Yacimientos Petroliferos Fiscales, the government oil entity, and five private companies. The immediate coastal portion of the field, which embraces the original discovery area, is almost completely developed. This is especially true of the original government reserve and one private property. The Y.P.F. has extended its explorations and developments westward to Escalante, 27 km. distant on the state railway. This extension has been the most active and prolific section of the field for the past three years. The government has also continued its exploration drilling in the region lying south of Colonia Sarmiento and east of the Rio Senguer. Private companies have not been able to extend their operations beyond the claims they had in process of acquisition in 1924, when the government reserve areas were established. The Y.P.F. is the major producer of the field and employs modern methods and appliances. Practically all its drilling and pumping operations are electrically operated, from a central power plant. In view of t'he scarcity of water supplies, this facilitates operations and is an economic factor of considerable importance.

Jan 1, 1934

By Leonard Greenburg

The American literature in the field of dust sampling and analysis has been growing rapidly since 1915. Studies made since that time clearly indicate that there are three fundamental factors that determine the production of lung pathology due to dust inhalation. These are the duration of exposure, the chemical and mineralogic nature of the dust and the concentration of the potentially harmful dust. In the estimation of the amount of the potentially hazardous dust there are essentially three problems involved: (I) a determination of the size of the hazardous dust; (2) determination of the most satisfactory apparatus and technique of sampling, and (3) the most satisfactory method of analysis. A complete study of the literature shows that industrial dust varies ill size from ultramicroscopic particles to those larger than lop in diameter and that relatively few particles are found over 5u in diameter. Approximately 10 to 25 per cent are less than 0.811 in diameter and the median size as a rule is found to lie between 1.25 and 1.75 microns. From the point of view of lung pathology the particle sizes of most importance lie between 0.5 and 5u. Of all the instruments for sampling dust the Impinger apparatus is the only one that will sample a larger body of air satisfactorily with small control errors and high efficiency in both low and high concentrations of dust. The Owens apparatus has the advantage of being more portable and somewhat easier to use but it does riot permit of the sampling of large volumes of air. It has a low efficiency in high concentrations of dust and does not permit dust weights to be obtained. From the point of dust-counting technique the author points out that in 1922 considerable experimentation with dark field and oblique illumination resulted in the conclusion that such methods were not readily duplicated from day to day and therefore yielded somewhat erratic results. Changes in the amount of illumination produced variations of several hundred per cent in dust counts obtained with dark ground

Jan 1, 1934

By B. W. Gandrud

In 1927 Coghill showed that there is considerable liberation of the iron oxide mineral from gangue when the Alabama red iron ores are ground to a size suitable for classification and tabling but that there is not sufficient to make possible a high extraction and a high grade of concentrate. Coghill states: High extraction and 'high grade of concentrate cannot be obtained without grinding to a size comparable with flotation or cyanide practice. If grade of concentrate be sacrificed the grinding need not be so severe. These ores slime readily during grinding, and in most instances the slimes are only slightly higher grade than the original ore. Not only do they carry a considerable percentage of the total iron in the ore, so that to reject them as tailings would result in prohibitively low extraction, but they are too fine to be amenable to ordinary methods of gravity concentration. In 1929 the Bureau of Mines published results showing that certain Alabama oolitic iron ores could be concentrated by classification and tabling after grinding, with rubber-covered rods. This selective grinding produced slimes of much higher grade than the original ore, and they were considered part of the final concentrate.

Jan 1, 1934

By George Fowler

THE history of the production of sulfur from salt domes in Louisiana and Texas originated with the operations of the Union Sulphur Co. at Sulphur, La., followed by the Freeport Sulphur Co. at Bryanmound, Tex. Including the present development of Grande Ecaille by the Free-port Sulphur Co., which constitutes the tenth dome to be equipped for the production of sulfur in the Gulf Coast, less than 10 per cent of all the known domes have proved to contain sulfur in commercial quantities. The Grande Ecaille salt dome is in the tidal marsh of the Mississippi River delta in Plaquemines Parish, Louisiana. It is 10 miles southwest of the right descending bank of the river approximately 45 miles below the city of New Orleans and is within 4 miles of the shore line of the Gulf of Mexico.

Jan 1, 1934

By Robert F. Mehl, Dana W. Smith

It has been shown that quenched iron of high purity exhibits a Wid-manstiitten figure much resembling martensite in appearance.1 This figure exhibits a maximum of four directions of the surface traces that characterize it, suggesting that a Fc forms from y Fe in plates parallel to the (III), octahedral, planes in the y lattice. Because of this, and also beqause of the general similarity to the structural appearance of martensite and to the Widmanstatten figure obtaining in slowly cooled hypoeutectoid steels it has been suggested2 that the orientation relationships obtaining between the y and a phases in pure iron are identical with those demonstrated in martensite3 and in slowly cooled hypoeutectoid steels,4 and indeed that the crystallographic mechanism of formation of a Fe from y Fe is the basic mechanism determining the orientation relationships found in both martensite and in slowly cooled hypoeutectoid steels. This paper will present experimental evidence in support of this view. Experimental Technique The analysis of the crystallographic mechanism of transformation in pure metals is beset with special difficulties, though some information

Jan 1, 1934

By Victor Dolmage

Introduction Although gold production in British Columbia began in 1859, lode gold was not produced till 1893. Since 1900, the annual production of lode gold has fluctuated between $3,000,000 and $6,000,000. About twenty per cent of this production may be regarded as by-product gold from base metal mining. The remainder comes from a few isolated but important mines, such as the Rossland mines, Nickel Plate, Surf Inlet, Premier, Pioneer, Bralorne, Reno, and Cariboo Gold Quartz, and from several less important mines. All of the above, except Pioneer, Bralorne, Reno, and Cariboo, are approaching or have reached the point of exhaustion, but all are now receiving, under the stimulus of the. present gold-mining boom, drastic resuscitative treatment, with, in some cases, signs of resurrection. Notwithstanding this, however, the main interest at the present time is focussed on the Bridge River and Cariboo fields. Both these fields were discovered in the early 'sixties' and, apart from the Cariboo placer production, both have from time to time contributed small amounts of lode gold; but it is only during the last two years that the real possibilities of these areas were discovered. and that they began to assume the appearance of goldfields of major importance. The two fields are sorme three hundred miles distant from one another and contain deposits of distinctly different types, so that the only reason for including them in one article is the fact that they came into prominence more or less simultaneously and at a time coinciding with, but not otherwise related to, the present gold-mining boom.

Jan 1, 1934

By C. E. Shoenfelt

The most important development in an otherwise colorless year in Colorado oil fields was the completion, late in 1933, of a well on Iles dome on the western slope of Colorado, with an initial production of 1100 bbl. of oil. No great amount of activity beyond the drilling of necessary offset wells is to be expected as a result of the completion of a well of this size because the acreage on Iles dome is closely held; there is no pipe line outlet other than to the Texas Company's small refinery at Craig, 24 miles distant; and the market for the oil is limited to the demands of sparsely settled communities in the high mountains. The Iles field is located on a large dome developed in Mancos shale of Cretaceous age, and having a closure of approximately 400 ft. The structure is highly faulted on the southern end. The first commercial oil well on this dome was completed by the Midwest Refining Co. on the south side of sec. 22-4N.-92W. Oil of 37" gravity was developed in Nancos shale at 2560 ft. and water in the Dakota sandstone at 2630 ft., but a considerable amount of additional oil was found in the shale immediately overlying the Dakota. From 2630 ft., the well produced in 16 hr. 750 bbl. of fluid, 33 per cent of which was water. Production in the Sundance sand of Jurassic age was found in the Midwest Refining Company's No. 4 Parkinson, in the NE.-SW.-SE. of sec. 22-4N.-92W., and completed on the last day of the year 1926, for an initial production of 450 bbl. of oil at a total depth of 3242 ft. The well completed in 1933 is also a Sundance sand producer and was finished at a depth of 3447 ft. It extends the productive area of the field about a quarter of a mile cast. The Iles field is said to have a potential production of 5000 bbl. of oil a day from its present wells, but for the past year has been held at about 600 bbl. a day. The most important test, stratigraphically, ever drilled on the western slope of Colorado is the California Company's No. 1 Raven, NW.-SE. of sec. 30-2N.-102W., which had an initial production of 138 bbl. of oil of 32" gravity, in the Weber formation (Pennsylvanian) between 5750 to 6315 ft. It mas drilled to 7137 ft., probably near the base of the Weber, and then plugged back to 6315 ft. It was shot quite heavily with nitro-glycerin, but failed to respond. While the well can hardly be classified

Jan 1, 1934

By William Fink

THE investigation of aluminum alloy systems prior to 1923 was severely handicapped by the low purity of the best aluminum available. However, by that time, the electrolytic purification of aluminum had been developed by the Aluminum Company of America,(1)? and in May, 1923, there was initiated, under the direction of E. H. Dix, Jr.,(2) a program for the investigation of the aluminum-alloy systems using the purest electrolytically refined aluminum. Up to the present time 15 papers (a) have been published, most of which are on the binary aluminum systems. Work is now in progress on the ternary systems, and it seems probable that most of the future work will be on ternary systems. Therefore this survey of the binary aluminum systems investigated up to the present time (the sixteenth paper of the series) concludes the first phase of this program. Certain qualitative relations between corresponding portions of the various binary diagrams as well as between different parts of the same diagram were noted some time ago by R. S. Archer.(4) In the present . paper these relations and others are considered quantitatively by the application of well-known thermodynamic equations or by newly discovered empirical relations No attempt will be made here to present the history of the develop-ment of the thermodynamic equations defining equilibrium or their application to metallic systems, although a list of references is given at the end of this paper for those who are interested in this phase of the

Jan 1, 1934

By Frank A. Herald

Through the generous cooperation of the authors of the papers in this chapter, a great wealth of fundamental data is being made available. The same plan of securing and submitting data will be carried forward next year, with the hope of making a nearer approach to complete presentation of the data contemplated in the plan. In order to conserve space, we are printing below such footnotes as have general application to Table 1 in the papers in this chapter: Footnotes to Column Headings Table I In areas where both oil and gas are produced, unless gas is marketed outside the field, such areas are included in column headed "Oil." Manufacture of casinghead gasoline and carbon black is interpreted as outside marketing of gas. Production per acre is determined by dividing into the number of barrels of oil the sum of the number of acres assigned to "Oil" plus such number of acres of the total assigned to "Oil and gas" as represents the portion thereof occupied by oil. Wells producing both oil and gas are classified as "Poducing oil only" unless gas from them is marketed off the lease. W. water; G. gas; A, air. Bottom-hole pressures are preceded by "e." All other figures represent pressures at casinghead with well closed. P, paraffin; A, asphalt; M, mixed. 0 Cam, Cambrian; Ord, Ordovician; Sil, Silurian; Dev, Devonian; Mis, Mississippian; MisL, Lower Mississippian; MisU, Upper Mississippian: Pen. Pennsylvanian; Per. Permian; Tri, Triassic; Jur. Jurassic; CreL, Lower Cretaceous; CreU. Upper Cretaceous; Eoc, Eocene; Olig, Oligocene; Mio. Miocene; Pli. Pliocene. 6, sandstone: SH, sandstone, shaly; Ss, soft sand; H, shale; L, limestone; LS, limestone. sandy; C, chalk; A, anhydrite; D, dolomite; Da. arkosic dolomite; GW, granite wash; P, serpentine. Figures are entered only for fields where the reservoir rock is of pore type. Figures represent, ratio of pore space to total volume of net reservoir rock expressed in per cent. "Par" indicates that the reservoir rock is of pore type but said ratio is not known by the author. "Cav" indicates that the reservoir rock is of cavernous type; "Fis," fissure type. A, anticline; AF, anticline with faulting as important feature; Af, anticline with faulting as minor feature; AM, accumulation due to both anticlinal and monoclinal structure; H, strata are horizontal or near horizontal; MF, monocline-fault; MU, monocline-unconformity; ML, monocline-lens; MC, monocline with accumulation due to change in character of stratum; MI, monocline with accumulation against igneous barrier; MUP, monocline with accumulation due to sealing at outcrop by asphalt; D, dome; Ds, salt dome; T, terrace; TF, terrace with faulting as important feature; N, nose; 9, syncline. Information will be found in text as indicated by symbols: A, name of author, other than above. who has compiled the data on the particular field; C, chemical treatment of wells; G, gas-oil ratios; P, proration; U, unit operation; R, references; W, water; 0. other information.

Jan 1, 1934

NORTH AMERICA Number Members Alaska 36 Canada 323 Mexico 184 Newfoundland 3 United States Alabama 40 Arizona 111 Arkansas 8 California 646 Colorado 171 Connecticut 90 Delaware 21 District of Columbia 136 Florida 16 Georgia 7 Idaho 66 Illinois 208 Indiana 66 Iowa 11 Kansas 24 Kentucky 32

Jan 1, 1934

By K. C. Gilbart, E. Stansfield

This test was described in a privately circulated First Progress Report on Slacking Characteristics of Coal, by A. C. Fieldner and W. A. Selvig, May 1, 1928. The test was described in detail in U. S. Bureau of Mines Report of Investigation 3055, in November, 1930,1 The test, in brief, is conducted as follows: A sample of freshly mined coal consisting of 18 to 25 pieces approximating 1 1/4 cubes, and weighing 500 to 600 grams, is taken for the test. The sample is subjected to a "blank sieving" of carefully regulated degree, on an 8-in. circular sieve with 0.263-in. square openings, and the undersize and oversize weighed. The oversize is placed on a sheet-iron pan and dried for about 24 hr. in an A.S.T.M. air-drying oven maintained at 30" to 35" C. through which a current of warm air is drawn. The sample is then cooled to room temperature and immersed for one hour in water at the same temperature. The water is then drained off, the sample dried for 24 hr. as before, cooled, sieved as before, and the oversize and undersize weighed. The percentage weight of the undersize is calculated and reported as the first cycle loss. This loss, corrected by subtracting the percentage loss in the blank sieving, is termed the slacking index. The sample may be subjected to further cycles of wetting, draining, air-drying, sieving and weighing. In a letter (August, 1932) W. A. Selvig approved making the blank sieving on a duplicate sample, rather than on the sample used for the test, where sufficient coal was available. This method is now employed in these laboratories. The accelerated weathering test has been employed in the laboratories of the Research Council of Alberta, at Edmonton, since 1928, but not found entirely satisfactory as repeat determinations with samples taken from the same or adjacent mines frequently showed serious differences. Differences must be expected as the coal may vary in the seam, and the selection of a typical set of lumps approximating 1 1/4-in. cubes

Jan 1, 1934

By T. E. Brower

SHORTLY after our previous paper on this subject was printed,1 we located a source of uncertainty in the results arising from the unexpected fact that hydrogen slowly reduces silica at 1100° C. in presence of iron, even when the iron nowhere touches the silica, although in absence of iron it does not do so. This difficulty has been completely obviated by the use of a small high-frequency furnace, which enables us to heat the specimen for analysis without raising the temperature of the silica vessel to a point at which any appreciable reduction of the silica glass occurs. We now describe this modification of procedure, and present some typical results obtained with it. For any given specimen of steel, the results are entirely definite and. reproducible, but their precise significance is still open to some question, as indeed is true of all methods of oxygen determination so far developed. Further elucidation of this question is to be expected from the cooperative comparison of the several methods now under way under the auspices of the Bureau of Standards at the instance of the Iron and Steel Division of the A.I.M.E.

Jan 1, 1934

YEAR OF ELECTION 1913. FRANK DAWSON ADAMS, Montreal, Canada. 1921. WILLIAM CUTHBERT BLACKETT Sacriston, Durham, England. 1920. HENRY STURGIS DRINKER Merion Station, Pa. 1933. KARL EIKERS NEW YORK, N. Y. 1922. FEDERICO GIOLITTI Torino, Italy. 1906. SIR ROBERT A. HADFIELD London, England. 1928. JOHN HAYS HAMMOND New York, N. Y. 1921. FRANK WILLIAM HARBORD London, England. 1917. HERBERT HOOVER Washington, D. C. 1905. HENRI LECHATELIER Paris, France. 1931. WALDEMAR LINDGREN Cambridge, Mass. 1921. C. McDERMID London, England. 1913. EZEQUIEL ORDONEZ Mexico City, Mexico. 1911. ROBERT H. RICHARDS Cambridge, Mass. 1934. T.A. RICKARD Victoria, B.C., Canada 1933. ALBERT SAUVER Cambridge, Mass.

Jan 1, 1934

By J. W. Paul, J. N. Geyer

The measures overlying some coal beds are under stress as the result of geological movements after the measures have been deposited. The removal of the coal in the process of mining induces additional stress in the roof measures. Certain types of deposits have sufficient inherent strength to be self-supporting over definite width of excavations and the stress is adjusted without appreciable movement of the material, to the extent that no part of the material will become detached and fall. The uplifting of the measures in geological time often results in the establishing of cleat planes or planes of weakness in the measures overlying the coal, and these planes generally have a uniform direction. Measures that have not been subject to uplifts are less liable to have noncompensable stresses, although any method of underground mining results in the creation of stress in the roof measures over the excavations and over the coal pillars. The immediate roof measures extending upwards of 10 to 15 ft. above the coal give the greatest concern in maintaining the roof in the advance work in the solid coal. Immediate measures which are inherently weak often may be scaled and supported for indefinite periods by a system of timbering. As an excavation is made, the pillars adjoining must take on the weight previously supported by the coal removed. The roof over the excavations takes on the function of a beam and if too weak to be self-supporting it will fail either through shearing or bending, and the roof cavity may enlarge until a self-supporting arch is formed. As support is removed by taking the coal away, there is a tendency for the roof members to move toward the excavation in an effort to compensate the stress; particularly is this true where there is an overlying bed of clay, which may act as a lubricant on which adjoining members may slide. Unstabilized stresses, plus the force of gravity, plus the inherent strength of the material, govern the tendency of mine roof to fall when not supported by timber. In general, most mine roof depresses when the coal is removed. This depression may be in the nature of a sag as the result of bending, or the edges of the pillar may break away or the floor may come up or heave.

Jan 1, 1934