Search Documents

By Nathaniel Arbiter

THE separation of minerals by flotation can be regarded as a rate process, with the extraction of any one mineral determined by its flotation rate, and the grade of concentrate by the relative rates for all the minerals. So regarded, the significant variables for the process are those that control the rates. These variables are of two types, the first describing the ore and its physical and chemical treatment prior to flotation and the second characterizing the separation process in the cells. This paper will examine the variation in rates for a group of separations, will show that a simple rate law appears to govern, and will consider the relation of the control variables to the rates. The use of rate constants for evaluation of performance and efficiency will be discussed. Flotation involves the selective levitation of mineral and its transfer from cell to launder. The flotation rate is the rate of this transfer. It may be defined by the slope of a recovery-time curve for any cell in a bank, or at any time in batch operation. The objective in flotation rate study is an equation expressing the rate in terms of some measurable property of the pulp. This can be either the concentration of floatable mineral in weight per unit volume1,2 or a relative concentration, which will be a function of the recovery." A rate equation for an actual flotation pulp will contain at least two constants, both to be determined from the data. One of these, the initial concentration or proportion of floatable mineral, is not necessarily equal to the feed assay because of nonfloatable oversize or locked particles." The other, a rate constant, is a measure of proportionality between the rate and the pulp property on which the rate depends. The value of the rate constant will be determined by the values of all variables which control the process and will be changed by significant changes in any of them. It is, therefore, a direct measure of performance. Where recovery or grade change continuously with flotation time, the rate constant will be independent of time and will characterize the entire course of the separation. Development of Rate Equations Rate equations can be developed either by analysis of the mechanism of the process or by direct fitting of equations to recovery-time data. Sutherland's attempt by the first method' suggests that the effect of particle size variation on the rate complicates the derivation of a simple equation applicable to an ore pulp. A further problem with an ore is the concentrate grade requirement, which usually involves a variable rate of froth removal. Thus the final rate for any cell may depend on the froth character and froth height, as well as on the pulp composition. This does not imply that each cell cannot reach a steady state2 in which the rate will depend ultimately on pulp composition. The second method is the fitting of rate equations consistent with the necessary boundary conditions* to experimental recovery-time curves. On the assumption that under constant operating conditions the flotation rate is proportional to the actual or relative concentration of floatable mineral in the pulp, a generalized rate equation may be expressed as follows: Rate = Kcn [I] where K is the rate constant, c is some measure of the quantity of floatable mineral in the pulp at time t, and n is a positive number. In previous rate studies, the value of n has been taken as 1, either by direct assumption," or as a result of the hypothesis that bubble-particle collision is rate determining.' A first order equation results, which after integration in terms of cumulative recovery R, leads to Loge A/A-R = Kt [2] The quantity A is the maximum possible recovery with prolonged time under the conditions used. No conclusive proof for the validity of this equation in flotation has been advanced. The evidence cited in its support consists entirely in the demonstration that it appears to apply to a limited number of recovery-time curves."' It will be shown subsequently that this procedure is not sufficient to establish the order of a flotation rate equation. The possibility that the equation may be of higher order therefore requires examination. If, in particular, the exponent in eq 1 is assumed to be 2, then after integration there results R = A2Kt/1 + AKt [3] with K again a rate constant and A the maximum proportion of recoverable mineral. Eq 3 may be

Jan 1, 1952

By E. R. Gilliand

KNOWLEDGE of a large number of the physical properties of the hydro-carbons is needed in the calculations and studies of the production engi-neer. Since experimental data on these properties of the individual hydrocarbons and their mixtures are available in only a relatively few cases, the use of generalized correlations has developed rapidly in the last decade. The generalized correlations of pressure, volume and tem-perature,1 enthalpy,2 and vapor-liquid equilibria,3 on the basis of reduced temperatures and pressures, are outstanding examples of this develop-ment. These correlations are of particular value to the production engineer interested in the flow of oil and gas, since they can be used to estimate the conditions under which segregation of phase occurs as well as the composition, density, and relative amounts of the phases. The technique of using the correlations for such estimations has been pub-lished by several authors¢ and will not be considered in this paper. A considerable number of data have been obtained to check the generalized pressure-volume-temperature correlations, and these data indicate that such relations hold for the hydrocarbons from methane to heptane. The accuracy is in general sufficient for most engineering pressure-volume-temperature calculations. However, the generalized correlations of enthalpy and vapor-liquid equilibria have in general been developed from the pressure-volume-temperature correlations by thermodynamic calculations and only a relatively few experimental data are available to check these calculated values. The calculations from pressure, volume and temperature are likely to lead to large errors in the calculated enthalpy and vapor-liquid equilibria. Thus, in the calculations of enthalpies, slopes of the pressure-volume-temperature data are employed, and determining these slopes can greatly amplify small errors in the original data. In the vapor-liquid equilibria calculations it is necessary to have the pressure-volume-temperature relations of the mixture, and the latter properties are much more questionable than those of pure components, particularly in the critical region. A considerable number of experimental data are

Jan 1, 1939

By S. R. Gilbert, B. W. Burdett, L. B. Hales

INTRODUCTION Grinding control is currently the most publicized control subject in the minerals industry. Flotation control and crushing control have not as yet received the widespread attention that grinding control has. The primary reasons for this center around the high capital costs of grinding equipment and the fact that grinding mills are extremely energy inefficient. In a typical copper concentrator the energy requirements to produce refined copper are: (1) [Total Energy Distribution (Kw-hr/yr)% Mining4.93 x 10817.3 Mineral Processing9.63 x 10833.8 Smelting and Refining13.9 x 10848.9] One-third of the total energy used for copper production is for mineral processing. A breakdown of this energy into process steps yields the following: [Total Energy Distribution (Kw-hr/yr)% Crushing and Grinding 7.16 x 108 73.8 Flotation 0.76 x 108 7.8 Materials Handling1.78 x 10818.4] It is apparent that comminution operations consume the largest proportion of the energy for processing and as such become major targets for improvements in efficiency. Unfortunately, size reduction operations are extremely inefficient. Efficiency calculations for grinding range from less than 0.1% to about 5% depending on the evaluation criteria. Because of these facts, grinding control has been and will continue to be the focus of major process control efforts. OBJECTIVES AND BENEFITS All benefits achieved as the result of grinding control are derived from modifying physical parameters which improve the efficiency of the grinding mill. These parameters include:

Jan 1, 1982

By C. Y. Garber

INTEREST in the sink-float plant operated by the Bunker Hill and Sullivan Mining and Concentrating Co. at Kellogg, Idaho, has recently been largely confined to the preparation and reconditioning of the galena medium. The ease and simplicity of these operations in a concentrating mill producing a flotation lead concentrate has occasioned enough comment to possibly justify a brief description. The subject is treated from an operator's viewpoint only; no effort is made to discuss it in a theoretical or technical manner. PLANT AT BUNKER HILL The sink-float plant was first operated in December 1941. It was built for the purpose of quickly increasing mill capacity with a minimum of equipment. It is the Huntington-Heberlein sink and float process using galena medium and is operated under license from the Sink and Float Corporation, New York city. Lead-zinc ore from the Bunker Hill mine is crushed to pass a I ¼ in. screen and is then washed and separated on vibrating screens having 4 ½ -mm. round-hole punched plates. The screen undersize, amounting to 33 per cent of the mill feed, is dewatered and sent to the ball-mill feed bins. The screen oversize goes to the sink-float plant, where 50 per cent of the oversize, or 33 per cent of the mill feed, is float, which is washed free of medium and disposed of as a tailing. The remaining 33 per cent is sink, which, after washing, is crushed to pass ½ in. and joins the undersize in the ball-mill feed bins. The normal plant capacity is 1700 tons per 24 hr. About 1.7 per cent of the total lead and 3.3 per cent of the total zinc in the feed is lost with the sink-float tailings (Table I). [ ] More complete descriptions of the H. H. process and Bunker Hill sink-float plant are given in several recent articles, listed at the end of the paper. PREPARATION OF MEDIUM AT BUNKER HILL New medium is made intermittently as needed, usually 12 hr. in each 6 days. A small portion of the current mill flotation concentrate is diverted from the lead cleaner cells to a I-in. vertical sand pump and pumped to a two-cell 36-in. Agitair flotation machine, where NaCn and ZnSO4 are added and the grade of concentrate is improved as shown in Table 2. Tailing from this operation is returned by gravity

Jan 1, 1946

By L. W. Kempf

THE thermal conductivity of aluminum alloys is of considerable indus-trial importance. This is particularly true in such applications as inter-nal-combustion engines where one of the principal reasons for the use of these alloys is their relatively higher thermal conductivity. In spite of this, relatively few data are available on this property of aluminum alloys. The reason is quite apparent to anyone concerned with the experimental determination of coefficients of thermal conductivity. This determina-tion, in theory so simple, is extraordinarily difficult if any high degree of precision is desired. Fortunately, some knowledge of the relationship between thermal and electrical conductivity has permitted the use of the latter physical property in deducing the relative approximate-values of the thermal conductivities of metals. This fundamental relationship, first enunciated by Wiedemann and Franz1 as applying to metallic elements and later shown by Lorenz to be applicable to alloys within a single metallic system, has been of service in the industrial application of metals. The electrical conductivity of metals is easily determined with a high degree of precision. It is, therefore, quite desirable that the relationship between electrical and thermal characteristics be determined with the greatest possible precision. This paper presents data on the electrical and thermal conductivities of a number of aluminum alloys of industrial importance.

Jan 1, 1937

By John E. Greenawalt

IN view of the increasing importance of sintering in the beneficiation of iron ores preparatory to their reduction in the blast furnace, the writer believes the time is opportune for an up-to-date, though brief, discussion of the general principles of sintering, as well as a discussion of some of the recent improvements and mechanical developments that have been achieved in this industry, its present trend and future possibilities. Sintering may be defined as the agglomeration of fine mineral particles into a porous mass by incipient fusion caused by heat produced by combustion within the mass itself. Its first successful application, in connection with the treatment of lead sulphide ores, was known as blast roasting. In this process the lead ore was given a preliminary roasting in a reverberatory furnace, then the incandescent mass was placed in a pot with a grate and an updraft was applied. It was soon discovered that much better results were obtained by cooling the ore, moistening it, and placing it upon an incandescent layer of fuel previously placed upon a grate, and then applying the updraft blast of air until the charge was converted into a more or less agglomerated mass. The fuel was sulphur and the process one of desulphurization, which was the principal result desired. The application of downdraft to blast roasting brought about tremendous results; the principal one of which was that an air blast of high pressure could be passed through a layer of ore deposited on a grate, or on a grate covered by a layer of coarse porous material, without disturbing the particles of which the layer is composed. The air current is compelled to filter through the pores of the charge, whereas if the same pressure applied to the top surface of the charge with downdraft were applied to the bottom of the charge with updraft, the charge would be completely blown off the grate. Another result of reversing the draft was that a flame could readily be applied to the top surface of the charge for igniting purposes. Broadly speaking, sintering is an oxidizing or burning process, therefore the volume of air passing through the charge is the all-important factor. The air must pass through with sufficient rapidity to produce the necessary heat, and if sufficiently rapid the combustion will take place

Jan 1, 1938

By U. N. Bhrany, J. H. Brown

The specifications of particle size and the size analyses of fine particulate materials are commonly presented without reference to the method of analysis. A review of the various sizing methods showed that not only numerous sources of error but also different definitions of size are inherent in the different methods of analysis. To demonstrate the magnitude of the differences, the size distribution of finely crushed quartz was determined by several methods and a critical comparison of the data was made. The experimental results show that appreciably different sizes may be indicated for one material depending on the method of analysis used. However, because the shape of the size-distribution curves determined by almost all methods was constant, a simple method of correlating the data can be adopted. Conversion factors, by which size measurements obtained by the various methods can be expressed in terms of screen sizes, are presented as one method of correlating the results. The size analysis and size specification for lump rock and particulate materials are commonly determined with sieves. With this technique, a readily reproducible measure of size can be obtained merely by duplicating the sieve and the method of sizing. For coarse materials, such as lumps of rock, this technique is simple and effective; and the effects of properties of the material such as shape can be readily understood. When very small particles are measured, however, problems arise. Small sieves are difficult to construct and maintain, and size separations become sensitive to the analytical procedure. For these difficult sieve sizes and for still smaller particles, other sizing techniques employing, for example, sedimentation, light extinction, and microscopy are available. Because these various techniques employ different size-dependent responses to establish size, it is appropriate to examine them and the results they yield to determine their relationships and their limitations. The 'size' of a lump of rock is not an absolute quantity, but rather a defined one. With sieves, for example, the size is defined by the screen aperture and no reference is made to the actual shape of the particle, yet only in the case of a regular body, such as a sphere, can the sieve size be related to the shape. With sieves only two dimensions of a particle can be suggested, for the sieve makes no allowance for the length of the particle passing through an aperture. Similarly, the size of a particle as measured by a microscopic technique need bear no relationship to the size as measured by sieves, although the two measurements may be related statistically. In fact, all size measurements are not only defined quantities, but the measurement actually used is an average that is controlled by the technique employed and, hopefully, reflects the property in question of the material of interest. Many papers have been presented that illustrate the accuracy of the various sizing methods, present the advantages of certain methods for certain applications, and describe the principles and technical backgrounds of the various methods. No attempt will be made in the present paper to review this work in detail. The purpose of the present paper is to examine the limitations of the various sizing methods, and to demonstrate experimentally the relationships between the sizes as determined by the various methods of analysis. It is hoped that by recognition of these limitations and relationships, errors in size specification and size interpretation will be avoided. REVIEW OF SIZING METHODS In specifying the size of a material, not only the data but also the method of measurement must be stated. For example, the size of the lump might be specified as its total volume, as its weight, as its maximum dimension, or as its minimum cross section; any one definition might be completely satisfactory for one problem, but not for another. In practice, size analyses follow four basic approaches and fortunately, the conditions that restrict each approach are known. These major methods include: 1) screen analysis, 2) direct measurement of particle dimensions, 3) determination of equivalent spherical size in response to fluid flow, and 4) specific surface determination. Each of these techniques has several variations and some new techniques are also available, but the limitations of the basic techniques are sufficient to permit at least a preliminary

Jan 1, 1962

By Gary Mack, Donald E. Scheck

To help management keep key machines or processes at peak efficiency, draglines, shovels and blast hole drills have been equipped with micro- processor based monitors. The dragline and shovel monitors log operating data such as energy consumption, bucket position, cab rotation and the clock time of key elements within each cycle as well as maintenance information. The output consists of maintenance and management data such as the number of cycles, average cycle times, swing angles, depths and downtimes. Extensive experimentation has shown that a blast hole monitor can provide a log of each blast hole in which the hard to fragment zones are indicated. The logs are printed on board the drill and serve as the design basis for decked loading of the holes. As a result, the fragmentation of the overburden has improved markedly with the same or less powder consumption. This material is based upon work supported in part by the Bureau of Mines, Department of the Interior under Grant Numbers G1115006, G1115391 and G5 105006. Any opinions, findings, and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of the Bureau of Mines, Department of the Interior.

Jan 1, 1983

By A. A. Berk

LAGGING technology and the slow spread of in- formation have been the chief obstacles to wide- spread participation in minimizing the industrial pollution load. These obstacles can be conquered by fact finding and applied research. There are three aspects to the problem of waterborne mineral waste: to minimize the amount being thrown away, to select a predisposal treatment, and to use the products or dispose of them.

Jan 7, 1957

By I. Iwasaki, K. A. Smith

A chlorination process for treating metal sulfide flotation concentrates is described. The process involves low temperature chlorination combined with sulfur chloride conversion and selective oxidation to produce nonferrous metal chlorides, elemental sulfur, and ferric oxide in a three-stage fluidized bed reactor. The nonferrous metal chlorides may then be water leached, separated, purified, and electrolyzed to recover copper and nickel metals and to regenerate chlorine for recycling.

Jan 1, 1986

By W. M. Courtis

SINCE many accounts of Silver Islet Mine, in Lake Superior, have already been published, it is supposed that the members of the Institute are familiar with the location and character of the mine. To many, however, a more detailed account of the ore and manner of treatment will, perhaps, not be without interest. The extraordinary richness of Silver Islet ores, and the difficulty of obtaining correct assays, induced the company to erect smelting works of its own. Accordingly, ground was broken at Wyandotte, Mich., early in the spring of 1871, and the work was pressed on with indomitable energy by Captain E. B. Ward of Detroit, as president, and Mr. Thomas MacFarlane, the discoverer of Silver Islet, as superintendent of the works, so that by July 1st the works were so far completed that smelting could be commenced. The intention was to work Western ores with those from Lake Superior, since the latter contain but a small amount of lead. Hence the capacity of the works is much larger than is needed for the present yield of the mine. The planned process for treatment was smelt-

Jan 1, 1874

By Reinhard H. Wöhlbier, George E. Aiken

8.4-1. Introduction. Surface excavating is done on a continuous basis with a variety of machines: 1) trenchers and ditchers, 2) conveyor loaders, and 3) bucket-chain and bucket-wheel excavators (BWE). Surface mining is largely concerned with only the last two machines, although some of the conveyor-loader types are often used by contractors for large-scale earth moving, and the engineer should be aware of these applications. This chapter is restricted to discussing only the bucket-chain and bucket-wheel excavators, with special emphasis on the latter. Continuous digging with these machines is done with evenly spaced buckets on either an endless chain or on the periphery of a rotating wheel. The bucket-chain digger excavates and takes the material away from the face with the same machinery. In the case of the bucket-wheel machine, these two functions are separate. Material is excavated by the buckets and then flows in a continuous stream via a transfer to the conveyor system, which in turn transports it to the discharge point. The wheels are adjustable in rpm and mount 6 to 12 buckets of equal size. These are classed as celled, cell-less or semicelled. The digging ladders for both types can move simultaneously in the horizontal and vertical planes. Ladder control is similar to the operation of a bucket-line dredge. Some ladder-chain and all wheel excavators have the machinery deck and superstructure mounted on a 360° swing circle. Other chain diggers have a fixed machinery deck. Deep and/or high cutting models are available in either the chain or wheel excavator. Machines can be either rail or crawler mounted, but the trend is to crawlers. Many of the early ladder-chain, rail-mounted excavators have been converted to crawlers. These may be either of the single or bogie type, depending on the conditions. Large machines usually use the superior three-point crawler suspension.

Jan 1, 1968

By T. Sterry Hunt

Coal was mined to a small extent near Tuscaloosa, in Alabama, and even carried by boats to Mobile, half a century since. Professor Porter, and later, Professor R. T. Brumby, occupied themselves with the geology of the region, and it attracted the attention of Sir Charles Lyell, who visited the region. It is, however, to the two reports of Professor Tuomey, geologist to the State, published in 1850 and 1858, that we owe our first clear notions of the geology of the Alabama coal-fields. His work has, since 1873, been continued by Professor Eugene A. Smith, of the University of Alabama, whose valuable reports, extending and completing the work of his predecessor, are indispensable guides to the geological inquirer in this region. Mention should also be made of the investigations of Mr. R. P. Roth well, who, after three years of studies in parts of the Alabama coal-fields, read before the American Institute of Mining Engineers, in October, 1873," a brief but important paper calling attention to the great deposits of coal and iron-ore found therein, and to their economic importance. Having myself, during the past year, had an opportunity of visiting parts of this region, I have thought that it might be well to bring before the Institute some facts and considerations as to its geological and geographical characteristics, and the value of its mineral resources. It is well known that the great Appalachian coal-basin, stretching southwestward through western Tennessee and the northwestern corner of Georgia, terminates in northern Alabama in what is called the Warrior coal-field, which there covers an area of about 5000 square miles. Lying along the southeast border of this, but

Jan 1, 1883

By R. W. Jenkins, O. J. Neslage

The need for mechanized drilling to decrease mining costs has resulted in the development of the jumbo from column-and-bar drill by carriages to hydraulically controlled jib jumbos. Resultant savings from use of improved jumbos have been high enough usually to retire the machinery investment within a year. The jib jumbo unit has been and applied successfully to development and preparation programs and flat-bedded stopes, and an even wider field is forecast for its use.

Jan 5, 1950

By Stephen Birch

THE Chitina mining district of Alaska is located at the headwaters of the Chitina and Copper Rivers. At present, the only producing min-ing properties are the mines of the Kennecott Copper Corpn. and the Mother Lode Coalition Co., which are situated 196 miles from Cordova the port of entry. The first claims, later acquired by the Kennecott Mines Co. and afterwards transferred to the Kennecott Copper Corpn., were discovered in 1900. The Copper River & Northwestern Ry., which connects the mines with tide water at Cordova, was completed in the spring of 1911. Contemporary with the construction of the railroad, aerial tram equipment was brought to the mines by pack train and a tramway, 3 miles, long, connecting Bonanza mine with the proposed railroad terminal, was finished, enabling shipments of high-grade ore to be made immediately on the completion of the railroad. A mill to treat the lower grade ore was begun the same year. The Kennecott company's holdings consist of 111 mineral claims. The Mother Lode Coalition Mines Co., which is controlled by the Kennecott Copper Corpn., owns 73 claims adjoining the Kennecott holdings. All data on operations and geology refer equally well to the Mother Lode property.

Jan 1, 1924

By E. A. Sondreal, F. I. Honea

This paper summarizes coal and ash properties and pilot-plant results for 14 key lignite seams in the Gulf Province, as well as discusses operating problems encountered by the four Texas power plants burning Gulf Province lignite. Some consideration is given to the design of future power plants.

Jan 1, 1982

By William Newnam

LEAD mining has been carried on in several localities of the Province of Ontario in a desultory fashion for the past 60 years, but up to 1916 the results have not been of much commercial importance. The most ambitious attempt was made by an English company which built a lead smelter at Kingston, Ont., in, 1880 to treat the- concentrates from the . Frontenac mines. After 2 years' operation, the -mines and smelter were abandoned. Up to the present there has not been a sufficient tonnage of concentrates to support an efficient blast-furnace plant and as most of the small properties could not stand the heavy freight and duty charges into the United States, development work has not been carried on in a, systematic manner. As a result the industry has languished. At, Galetta, in southeastern Ontario, considerable prospecting and development work has been carried on in the last few years and very promising deposits of galena have been found in the Chats Island group by the James Robertson Co., Ltd., of Montreal. The chief deposit is that of the Galetta or Kingdon lead mine, located on Chats Island in. the Ottawa River about 5 miles east of the town of Arnprior. The rocks' in the vicinity of -the mines consist of an interbedded series of crystalline limestone and biotite gneisses of pre-Cambrian age. This part of the Ottawa River' basin has been severely faulted and the overlying Paleozoic limestones may be seen in normal fault contact with the pre-Cambrian. The vein filling is of a very well-marked fault fissure type and although the amount of displacement is not determined at this point, similar faults to the east show a displacement of from 1,500 to 1,800 ft. which would indicate mineralized fault fissures of considerable depth. The ore occurs in a highly crystalline calcite with some barite and fluorite, and the galena occurs more or less richly disseminated in clusters and crystal aggregates. In some cases these masses weigh several hundred pounds. Small amounts of sphalerite are occasionally encountered.

Jan 4, 1917

By F. R. Conley, D. A. Shock

Solution Mining or In Situ Mining has received increasing attention in the past few years because the method offers many attractive possibilities in improving the environmental impact of mining as well as making feasible the extraction of deeper, low value mineral reserves. This paper will present a brief discussion of the state-of- the-art in total solution, slurry removal and leach mining. As a specific example of utilizing the method, the uranium deposits of South Texas are a type of deposit which have been considered as being amenable to in situ leach mining. The parameters required to ascertain the amenability of such a deposit for in situ leaching will be presented. The hydrological factors which influence the production of the soluble uranium solutions and their displacement will be discussed. Mathematical models for various well configurations have been made and graphic representation will be presented to illustrate various production methods and the resultant hydrological consequences.

Jan 1, 1974

By C. H. Benedict

A 2000-ton ammonia leaching plant has been operated by the Calumet & Hecla Mining Co., at Lake Linden, Mich., continuously since February, 1917, except from April, 1921, to April, 1922, during the period of depression following the European war. A description of this plant has been published elsewhere.1 The present article describes the operation of the plant since 1917, together with the changes in the process that have been found to be advisable. In spite of the fact that this plant and that of the Kennecott Copper Co. are the only ones operating on a large scale, much experimental work is being done on ammonia leaching, as applied to oxidized ores, and it is to aid these and other investigators that this detailed account is given. For the benefit of those unfamiliar with the original article, the following short abstract is given, as an introduction to a more detailed account of the process and the reasons for present practice.

Jan 2, 1924

By David M. Hopstock, Jose M. Pastrana

The magnetic susceptibilities of samples of natural and artificial hematite and natural goethite were measured by the Faraday method as functions of temperature, magnetizing field strength and particle size. All samples showed weak ferromagnetism, which accounted for large variations in susceptibility with field strength and particle size. Three probable sources of the weak ferromagnetism were identified.

Jan 1, 1978