Search Documents

By Sven Fornander

L. F. Reinartz (Armco Steel Corp., Middletown, Ohio) —I would like to know, in the practical application of the Kalling process, what kind of a lining was used, how thick was the lining, and how much metal was treated at one time? S. Fornander (author's reply)—The rotary furnace is lined with a course of fireclay bricks 6 in. thick. This course is backed by 5 in. of insulation. The furnace has a capacity of about 15 tons. Mr. Reinartz—How was the ladle preheated? Mr. Fornander—As pointed out in the paper, the furnace was heated by a gas flame in the beginning of the experiments. During these first tests, however, the desulphurization was inconsistent. We think that this was due to the fact that iron droplets sticking to the furnace walls were oxidized by the gas flame. Now, the furnace is operated without preheating of any kind, and the results are much better. T. L. Joseph (University of Minnesota, Minneapolis, Minn.)—I might add one comment. This furnace was heated with a flame and for a time they had a little difficulty due to some residual metal in the rotating drum that would oxidize in between treatments and they found therefore, that it was very essential to drain the drum completely of metal so that they would not build up any ferrous oxide between treatments and they eliminated some of their erratic heats by maintaining those more reducing conditions. It was interesting to watch this operation. As soon as the drum started to rotate there was considerable flame, at least, at the time I saw it, that came out around the flanges, indicating there was quite a little pressure on the inside of the drum. W. 0. Philbrook (Carnegie Institute of Technology, Pittsburgh)—Is the reaction slag in the Kalling process liquid or solid, and how is it separated from the metal? Mr. Fornander—In the process there is no slag in the usual sense of the word. The lime powder does not melt during the treatment. After the treatment the lime is still in the form of a fine powder. It is separated from the metal by means of a piece of wood of suitable size placed within the furnace before it is emptied. D. C. Hilty (Union Carbide & Carbon Research Laboratories, Niagara Falls, N. Y.)—Dr. Chipman has given us some of his ideas in connection with a specific effect of silicon and silica on sulphur elimination and how silicon might interfere with desulphuriz- ing in the blast furnace. I wonder if he would like to elaborate on the possibility of a similar effect of silicon in the Kalling process? J. Chipman (Massachusetts Institute of Technology, Cambridge, Mass.)—Silicon does not interfere with the Kalling process. Anything that has strong reducing action is good for desulphurization. In these tests where the temperature was low compared to blast furnace temperatures, the silicon that is in the metal is a better reducing agent than the carbon. At high temperatures, carbon is the better. It is not the silicon in the metal that interferes with desulphurization, it is the silica in the slag. Mr. Joseph—I might add that the metal that was tapped from the drum after desulphurization was really at quite a low temperature. It was not measured, but I think it was well under 1300 °C, probably 1200" or a little above that. That was one of the difficulties, and I think there is no question about the fact that the Kalling process—in that it affects desulphurization between powdered lime, solid and liquid iron— is a reaction definitely between the solid lime and the liquid iron. E. Spire (Canadian Liquid Air, Montreal, Canada) — This Kalling process seems very interesting to us and after all it is only a mixing action that is taking place between the iron and the slag. We have attempted to do the same thing in another way. We have placed at the bottom of the ladle a porous plug through which we injected an inert gas. It can be nitrogen or argon. This plug is placed at the bottom of the conventional ladle and gas injected through the plug. That has appeared in our patent. To define this new type of treatment, I use the word gasometallurgy. I do not know if you like it, but it is a way of defining methods of treating metal using gases. What we do is exactly what is done in the exchange process in another way. We have a porous plug at the bottom with a high lime slag on top of the metal. Using this method, we have very good agitation of metal and slag, and with a small flow of gas, we can achieve a very strong agitation. For instance, in the 500 lb ladle, we use only 5 liters of gas a minute. We have an agitation compared to very rapidly boiling water in a pail. Moreover, the agitation can be controlled to create any amount of mixing desired. In a few minutes, with this method, the sulphur dropped from 0.58 to 0.11. These results have been improved since, and we have obtained results like 0.08

Jan 1, 1952

By Frank B. Fulkerson

A study was made of the economic factors governing silver production in the Coeur d'Alene mining region, Shoshone County, Idaho. Production and other stat is tics were analyzed. It was concluded that if the rapid rise in the open market price that began late in 1961 continues, greater production of the metal should result from the mines of the Coeur d?Alene mining region. However, the output gain would be limited, because costs per ton of ore mined would rise as production was increased beyond optimum capacity. Advances in the price of silver would increase the size of ore reserves by making some marginal ore profitable to mine and by providing greater funds for exploration projects that might result in discoveries adding to the reserves. Output of byproduct silver from lead-zinc mining probably will not exceed 4.3 million ounces, judging from the trend from 1946 to 1961.

Jan 1, 1964

By Alan D. Randolph

A size-continuous breakage kinetics model is postulated and supported with industrial grinding mill data. It is shown that the kinetics parameters include the effects of possible classification at the mill discharge. The estimated parameters were then used in the kinetics model to simulate the behavior of a grinding mill in open and closed circuit operation. While material transport and kinetics are equally important in open loop operation, heavy recycle in a closed circuit make a detailed material transport model for grinding unnecessary. In effect, the mill can be considered as backmixed. This suggests that more emphasis should be placed on modeling the classifier than the development of detailed material transport models for the grinding mill.

Jan 1, 1975

By Wilbur Duvall

For the past few years the Applied Physics Branch has been investigating, experimentally, the generation and propagation of strain waves in rock. The strain waves are produced by the detonation of a high-velocity explosive placed in a small cavity in the rock. Strain gages mounted in the solid rock surrounding the cavity record the strain pulses propagated away from the source. Numerous strain-wave pulses have been recorded in different rock types for charges of various sizes and at various distances from the charge. Analysis of these records has shown, among other things, that there is associated with the strain wave a characteristic pulse width. This pulse width is dependent on the rock type and the distance from the shot point to the gage, but is apparently independent of the charge size or cavity size. Recent published literature on the generation and propagation of seismic waves indicates that the characteristic frequency of an elastic wave generated by an explosion should be a function of the rock type and the size of the cavity created by the explosion. Also, internal friction in the rock would produce absorption of high-frequency components, which would cause continual change in the shape of the pulse propagated. As the experimental data did not verify the expected change in pulse shape with cavity size, a theoretical investigation of the vibrations associated with a spherical cavity in an elastic medium was undertaken.

Jan 1, 1950

By B. J. Dotson, P. N. McCreery, R. L. Slobod, James W. Spurlock

A core sample porosity-check program is described. A number of laboratories participated in the investigation. which comprised measuring the porosities of ten selected natural and synthetic core samples. Each laboratory employed its own method (or methods) of measurement. These included Boyle's Law, water-saturation. and organic liquid-saturation methods. Brief descriptions of each method employed are given. In general, the measured values of porosity were in reasonably good agreement. The average deviation of porosity from the mean or average values for the group of samples was ±0.5 porosity per cent. The results of these measurements are compiled and presented in graphical form. INTRODUCTION The total hulk volume of a porous sample is derived from the sum of two volumes, the void or pore volume and the solid* or grain volume. The porosity fraction of the sample ia defined as the ratio of the void volume to the bulk volume. Multiplication of this ratio by 100 converts it to per cent porosity. Distinction is made here between total porosity and effective porosity. The former includes all void space, both interconnected and isolated, while the latter includes only those voids which are interconnected. This investigation deals only with the effective porosity.* Numerous cases had been noticed in which there appeared to be appreciable discrepancies in the results of porosity determinations on core samples. Differences of as much as 10 to 15 porosity per cent had been reported from different sources

Jan 1, 1951

By P. N. McCreery, R. L. Slobod, B. J. Dotson, James W. Spurlock

A core sample porosity-check program is described. A number of laboratories participated in the investigation. which comprised measuring the porosities of ten selected natural and synthetic core samples. Each laboratory employed its own method (or methods) of measurement. These included Boyle's Law, water-saturation. and organic liquid-saturation methods. Brief descriptions of each method employed are given. In general, the measured values of porosity were in reasonably good agreement. The average deviation of porosity from the mean or average values for the group of samples was ±0.5 porosity per cent. The results of these measurements are compiled and presented in graphical form. INTRODUCTION The total hulk volume of a porous sample is derived from the sum of two volumes, the void or pore volume and the solid* or grain volume. The porosity fraction of the sample ia defined as the ratio of the void volume to the bulk volume. Multiplication of this ratio by 100 converts it to per cent porosity. Distinction is made here between total porosity and effective porosity. The former includes all void space, both interconnected and isolated, while the latter includes only those voids which are interconnected. This investigation deals only with the effective porosity.* Numerous cases had been noticed in which there appeared to be appreciable discrepancies in the results of porosity determinations on core samples. Differences of as much as 10 to 15 porosity per cent had been reported from different sources

Jan 1, 1951

By A. Bauer, P. Katsabanis, M. Baker

The principal results from a previous paper presented at the Mine Accident Prevention Association of Ontario in 1987 (1) are summarized here. It is concluded that the selection of explosives in mines experiencing sulphide ignitions is important and that aluminized explosives should be avoided because of their high explosion temperatures and their detonation products. The duration of the blast is also critical and secondary blasts should not occur simultaneously with nearby priroa:ty blasts.

May 11, 1988

By D. P. Cook, Jr. Jones, J. W. Evans

"It is well understood that the flow of liquid metal in the mold region of the caster can have a notable effect on the overall solidification behavior in aluminum casting. Previously our group has characterized the flow in this region in a water model, using particle imaging velocimetry. The flow is significantly effected by the specifics of the metal delivery system, i.e. size, geometry, used during the cast. Recently, a time-averaged 3D numerical model of the nozzle, combo bag and sump region has been developed and its validity has been tested against experimental data. Cases were run for different sized combo bags as well as bags that have been blocked. The results show satisfactory agreement between the physical and mathematical models.IntroductionFluid flow in the head of an electromagnetic (EM) and direct chill (DC) ingots has been shown to effect many important operational parameters in casting such as superheat removal, macrosegregation, and hot cracking. In most cases, the momentum from the nozzle is diverted through a delivery system, termed a bag, before entering the mold region. The specific geometries of the bags vary widely from operator to operator suggesting that there is still some question as to the optimum design. Several studies [1-4] have been carried out to understand the influence of the bag design on the solidification behavior in aluminum casters. However, few studies have included numerical prediction with experimental verification. Unfortunately, many simulations performed today still do not accurately account for flow in such devices.Recently, a significant amount of research into the flow resulting from the metal delivery systems has been performed [5-8]. The thrust of this research has been characterizing the flow patterns by a measurement technique called particle imaging velocimetry. The data suggest that the method of delivering the liquid metal into the mold region has a significant effect on the nature of the flow. With a detailed set of quantitative data from a water model, a numerical model can be built and tested to ensure a proper description of the flow field. The computational model used in these simulations is designed to represent a laboratory scale model of a caster in which ""combo"" bags were simulated. Such bags are mostly of a densely woven glass fiber material with low permeability to aluminum. ""Windows"", of a more open weave, permit aluminum to flow from the bag, see Fig. I. Two different sized combo bags were modeled, each subject to two different outflow window configurations. Furthermore, experimental measurements [9] were used to validate the model results."

Jan 1, 1999

By Bee R. Waples

The Alimak Raise Method, using the Alimak Climber, has proven to be a new, fast, economical method of driving raises at the Iron King Branch of Shat-tuck Denn Mining Corporation, Humboldt, Arizona. In order to fully understand the advantages of the Alimak Raise Climber, it is necessary first to describe the basic machine; second, a complete cycle of a raise round; and third, the comparative speed and economy afforded the Iron King operation by using this new method of raise driving: The raise climber consists of a drilling platform that is raised and lowered by two 6.2 HP compressed air motors that drive, through roller chains and sprockets, three worm gears connected in series. These worm gears drive pinions which grip the guide rail rack. The worm gems are fitted with spring-loaded couplings so that a small rock or grit in the rack is passed over without disturbing the mesh between the rack and the other pinions.

Jan 1, 1961

By W. H. Oppelt

Two exploratory gasification tests of limited duration were made at Dorsten, Germany, with strongly coking Pittsburgh-bed coal in a fixed-bed pressure gasifier. To reduce the coking potential of the Arkwright coal, ash refuse from the gasification of Dorsten (Leopold) coal was added to pre-pare coal-ash mixtures which contained nominal 20 and 30 percent ash. Major adjustments in principal process variables at the beginning of the tests resulted in wide fluctuations in the rate of gas production, gas offtake temperature, and carbon dioxide concentration in the gas. Because a higher steam-to-oxygen ratio was used in the tests than was customarily employed for routine operation, gas production was lower, and steam and oxygen consumption was higher. After operating conditions had been improved, the results of a 12~-hour-long test with addition of sufficient ash refuse to obtain a nominal 30 percent ash concentration indicated that operation is feasible with this coal mixture using a fixed-bed pressure gasifier equipped with a rotating coal distributor and stirrer.

Jan 1, 1966

By X. Q. Jiang, D. L. Chen, A. Macwan

Magnesium alloy, being the lightest structural material, is a key candidate for the weight reduction of vehicles in automotive and aerospace industries; however, its poor corrosion resistance and difficulty in deformation are limiting factors. Recently, Al/Mg/Al tri-layered clad sheets have been considered as a promising alternative to improve corrosion resistance and formability of magnesium alloys. The aim of this study was to evaluate the tensile properties and deformation behavior of Al/Mg/Al tri-layer clad sheets at different temperatures and strain rates. Yield strength (YS), ultimate tensile strength (UTS), and elongation were observed to increase with increasing strain rate. With increasing temperature from -40°C to 180°C, YS and UTS decreased and ductility increased. The strain hardening exponent evaluated via three equations of Hollomon, Swift, and Afrin et al. showed a similar increase with increasing strain rate and decreasing temperature. The fractographic analysis showed partial debonding of clad interface especially at higher strain rates and lower temperatures.

Jan 1, 2015

By B. A. Cheadle, C. E. Ells

The ovienlalion of hexagonal a-zirconium crystals in cold-drawn Zircaloy-2 tubes and in both as-extruded and heat-treated Zr-2.5 wt pcl ND tubes has been rrleasured using the inverse Pole - figure technique. Uniaxial tensile tests hare shown the Zil-caloy-2 lubes to hare highly anisotropic delovmation behavior whereas heat-treated Zr-2.5 wt pet Nb tubes are nearly isotropic. The differences in deformation behavior are esbloilred from the observed textures of the tubes and the deformation systems known to oberate in Zircnloy,-3. Tlze vcsults lead to an explanation of the high effective strain observed fov the Zircaloy-2 tubes as colrzpared to heat-treated Zr-2.5 wt pet Nb tubes in restrained end burst tests at 280° to 300°C. THE Zr-2.5 wt pct Nb alloy. when heat-treated by quenching from temperatures slightly below the (a + p)/P phase boundary and subsequently aged at 500°C, has tensile strengths about 50 pct greater than those of 20 pct cold-worked Zircaloy-2.1 There is. therefore, considerable economic incentive to replace Zircaloy-2 with the heat-treated alloy in some in-reactor applications. notably for pressure tubes, where strength is a prime consideration. However, when tubes of Zr-2.5 wt pct Nb fabricated by extrusion and drawing prior to heat treatment were subjected to restrained end-burst tests at 300 °C. the circumferential expansion at failure1 was only 10 pct that of cold-drawn Zircaloy-2 tubes in similar tests,2 Table I. This result was not predicted from uniaxial tensile tests in which the ductilities of the two materials were quite similar.' The detailed requirement for ductility in pressure vessels has not yet been worked out. and comparison of the relative merits of cold-worked Zircaloy-2 and heat-treated Zr-2.5 wt pct Nb is further complicated by the irradiation-induced loss of ductility and hydrogen pickup each will undergo from reactor operation.1,2 Nevertheless. the lower circumferential expansion of the Zr-2.5 wt pct Nb tubes raised serious doubts on the advisability of their use as pressure vessels. For this reason, an investigation to determine the cause for different ductilities of the two types of tubes was initiated. From a study of uniaxial tensile properties and crystallographic orientation it has been found that the fracture ductilities observed in the burst tests can be related directly to the crystallographic ori-

Jan 1, 1965

By R. Le Houillier, M. Nagamori, A. J. Plumpton

"Asbestos tailings are a typical serpentine, hydrated magnesium silicate, consisting primarily of antigorite and a few per cent of unrecovered fibrous chrysotile. Serpentine has unique thermal properties in that dehydroxylating calcination at 600 to 700°C can yield an amorphous structure in which most of the magnesia is liberated from the original bond with silica and thereby rendered chemically active. The amount of activated magnesia can be quantified by selective leaching with a weak acid such as acetic acid. Many wet processes have been proposed to extract magnesium values from calcined or natural serpentine by means of various acids and their ammonium salts. However, the only commercial utilization of serpentine today is found in dry processes such as the production of fused magnesium phosphate fertilizer .The present paper reviews the physical chemistry of the dehydroxylation of serpentine and also various wet and dry processes in order to provide a general view regarding the chemical transformation of asbestos tailings.IntroductionAsbestos fibers occur as chrysotile, a serpentine polymorph. More than nine tenths of mined asbestos ores consist of massive non-fibrous serpentine, primarily the polymorph antigorite, which is discarded as tailings after fiber extraction. For many decades, Quebec has been a world leader in the production of asbestos fibers. As a result, colossal tonnages of serpentine have accumulated in the Eastern Townships'!'. Serpentine minerals are found in large deposit s throughout the world, and significant research efforts have been made, particularly in Japan and the USSR, mostly in attempts to extract magnesia or silica values therefrom.The chemical composition of serpentine (chrysotile and antigorite) may be ideally expressed as Mg3Si2Os(OH). which corresponds to 43.7% MgO and 43.3% SiO2. Natural serpentines often contain a few per cent of Ni-bearing magnetite and other impurities, and therefore the contents of MgO and Si02are both about 40% or lower.Serpentine (hydrated magnesium silicate) is virtually insoluble in weak acids. However, when dehydroxylated under controlled conditions, ' it decomposes to an amorphous state in which magnesia and silica are isolated and chemically active. The magnesia so activated is readily soluble in weak acids. This unique characteristic of calcined serpentine, first discovered in 1942,, has since attracted the attention of innumerable researchers, whence many processes have been proposed."

Jan 1, 1980

By Fiona M. Doyle Garner

Hydrolytic stripping is the process whereby metal ions in a loaded solvent extractant are hydrolyzed by water, typically at 130°-200°C. Equilibrium hydrolytic stripping tests were done on Versatic 10 solutions of various metals, singly and in mixtures. Single solutions of Fe, Xi, Cu, Mg, and tin precipitated aFe2O3, Ni(OH)2, Cu2O, Mg(OH) 2 and ?Mn203 respectively during hydrolytic stripping. Several mixtures containing iron precipitated magnetic spinel ferrites, MFe20k. No other combinations formed mixed oxides. It is proposed that the mechanisms for hydrolysis and precipitation of iron in loaded Versa- tic solutions are similar to those for iron hydrolysis in aqueous solutions. Infrared spec- troscopy and thin-layer chromatography of mixed iron-nickel Versatic solutions indicated that mixed Fe-Ni carboxylate complexes are present. Magnetic interactions between mixed complexes are thought to be responsible for the homogeneous nucleation of magnetic mixed oxides in preference to single oxides.

Jan 1, 1984

By James J. Olson

Most of the projects summarized in this paper are funded under the Bureau's Advancing Metal-Nonmetal Mining Technology program and are part of the in situ mining subprogram coordinated at the Twin Cities Mining Research Center. The borehole mining equipment development, are funded under the Bureau's Advancing Coal Mining Technology program, is essential to accelerated progress in the in situ mining subprogram, and some background information on this concept is included. The initial in situ research projects started in 1971-72 evaluated some of the critical blast design parameters for in situ copper leaching. The present program totals about $1 million per year and includes in-house research projects, cooperative (cost-sharing) projects with the mining industry, and research and development contracts. Two different extraction schemes are under investigation: inplace leaching and borehole (or slurry) mining. The principal commodity targets for the Bureau's improved in situ mining technologies include copper, uranium, coal, and potash. The principal purpose of this paper is to inform the mining industry of the in situ mining research recently completed or being conducted by the Bureau of Mines. In addition the Bureau encourages companies to comment on the program accomplishments as well as offer suggestions on top priority research needs for improving the in situ mining methods described in this paper.

Jan 1, 1977

By Henry D. Hibbard

FoR the purpose of this paper all steels will be divided into two divisions: effervescing and non-effervescing. This classification must be borne in mind as many statements true of one class are not true of the other. Effervescing steels are those that effervesce in the molds, evolving large volumes of gases, which, in escaping, throw up a spray of molten particles of steel and give the molten metal in the mold a rapid stirring or churning motion until it freezes, or, at least, becomes pasty. This effervescing is intended and proper. The non-effervescing steels, which are more or less completely killed, should not evolve any gas in the molds. Many years ago the writer proposed to call the two classes of steels evolution and solution steels, depending on whether the gases in the molds are evolved or kept in solution, but the present designations seem to be better. In America, at least, effervescing steels are in a class by themselves and include steels made for boiler, tank, and ship plates, steel pipe, soft steel wire, wire nails, soft machincry, and structural steels. These are all low-carbon steels with less than 0.40 per cent. carbon. Low-carbon steels that do not effervesce but are killed are also made; these are put into ingots and castings but they are outside the limits of this paper. Effervescing steel is cast into square or slab molds, either top or bottom poured. Most of the effervescing steel is made in the basic open-hearth furnace, though much is made by the acid Bessemer process and a little is made in the acid open-hearth. It is not made by the crucible or electric-furnace process, in which the volume of gas required cannot be produced in the steel and evolved from it in the mold to give the effervescence desired. Any steel ingot, when teeming is finished, is of a certain size or volume. Leaving out of consideration for the moment the lessening of the over-all dimensions of the ingot through contraction due to loss of heat, the ingot may be placed in one of three classes: It may settle, losing in volume; it may stand, that is, remain constant in volume; it may rise, that is, increase in volume. Into which class an ingot will fall depends on the gases of the steel,

Jan 1, 1920

By Daryl R. Tweeton

The Bureau of Mines compared the effects of surface-active chemicals with those of pure water on core-drilling rock and quartz and on fracturing quartz. The objectives were to determine if these chemicals assist drilling and fracturing rock and if zeta potential control was worth pursuing as an aid to full-scale drilling. All tests were performed in the laboratory. No significant increases in drilling rates were observed for 1 to 10,000 ppm Al (as AlC13) in flushing water when drilling microcline or serpentine, nor for 1 to 10,000 ppm Mg (as MgC12) when drilling microcline or marble. Concentrations near those giving a zero zeta potential were used with microcline and marble. A 50-pct rate increase was observed for drilling quartz in 0.7 ppm oleylammonium acetate, which gave a zero zeta potential. No other concentrations were used. No effect on the tensile strength of quartz thread in 0.05 ppm Al (as AlC13), which gave a zero zeta potential, was observed. Heptanol, reported to give a zero zeta potential on quartz, did not reduce the bending strength of small quartz bars below that in water. These generally negative findings raise doubt about the usefulness of zeta potential control in full-scale drilling. The results, combined with previous studies, suggest that controlling physical properties of drilling fluids, such as viscosity, is more useful than altering chemical properties.

Jan 1, 1976

By L. B. Pankratz

Enthalpies above 298° K were determined for AgCl to 1,400° K and for AgBr to 1,200° K by copper-block calorimetry. Before melting, both compounds showed an anomalous increase in enthalpy, an effect associated with lattice defect formation. Melting points and heats of fusion were determined to be AgCl, 730° K and 2,945 cal/mole; AgBr, 700° K and 2,030 cal/mole. The heat capacity of both melted substances decreased with increasing temperature over the entire liquid ranges investigated. Enthalpies above 298° K were tabulated at even temperatures along with derived entropy increments and heat capacity values. The results were combined with the known heats of formation and entropies at 29.8.15° K to obtain values of the heats and free energies of formation over the range 298.15° to 1,400° K for AgC1 and 298.15° to 1,200° K for AgBr.

Jan 1, 1970

A viable mineral economy requires availability of economic deposits, the ability to exploit then, and protection of the public interests. The natural tendency for each party to consider only its legitimate self interests can lead to delays or deadlocks. The paper presents several viewpoints aimed at opening each party's mind to the ideas of others in the hope that negotiations might be more reasonable and rapid for the benefit of all concerned.

Jan 1, 1979

By O. E. Traubert, E. G. Meiter

"The daily sulphur dioxide content of the air at the Pittsburgh Experiment Station of the Bureau of Mines was determined by the bureau at the request of Committee D-14 on Screen Wire Cloth of the American Society for Testing Materials. The purpose of the investigation was to obtain data for use in interpreting screen-corrosion tests, which were conducted simultaneously and jointly by the committee4 and the Bureau of Mines.The plan of this investigation was to vale a single daily determination, except on Sundays and holidays, of the SO2 content of the atmosphere at the exposed test screens, for a period of one year. No attempt was made to determine the variation during the day, except on the morning of October 15, 1927, when the atmosphere was unusually foggy and smoky. Several successive determinations were made on that date to ascertain changes in the SO2 as the morning haze disappeared. The SO2 determinations were begun on August 11, 1927, and continued until August 11, 1923.The Pittsburgh Experiment Station of the Bureau of Mines is in a residential section about 3 miles east of the central or main business district of Pittsburgh. The air samples were taken on the roof above the auditorium, or the center rear of the building. This was also the place where the pieces of screen were placed for corrosion tests. The height was about 40 feet above ground level. Several sources of considerable smoke and SO2 are in the immediate vicinity. On the west side of the building is a double-track railroad. At this point the line runs through a cat below the level of the building. The horizontal distance between the air-sampling position and the railroad tracks is about 400 feet, and the vertical distance about 70 feet. The power plant for the Bureau of Mines building is about 250 feet south of the sampling position. About 1000 feet farther to the south are the engineering building and power plant of Carnegie Institute of Technology, and 1/4 mile to the southwest is the power and heating plant for the Carnegie Library and Museum. About 1 to 2 miles farther south and also southwest"

Jun 1, 1930