Search Documents

By John J. Howard

WE are approaching the depletion of our principal source of iron ore-the Great Lakes deposits, which have provided 85% of the nation's requirements for the past fifty years. This situation presents a threat to our standard of living and our position as a world power. Fortunately, in the Great Lakes region there are still tremendous quantities of potential iron ore in the form of the low iron, high silica rock called taconite. The economical extraction of iron from taconite will insure our future steel supply if no new deposits available to the United States even in time of war are discovered. In the new process presented here, the iron is reduced directly by a simple method and then separated from the silica. The product is "feed," not to the blast furnace, but to the steel plant. This process requires very much less plant, less physical and chemical treatment, and less fuel and auxiliary materials than the conventional ore concentration-blast furnace method. In my opinion, this new, fundamentally different approach offers a better hope for the solution of the iron ore problem.

Jan 5, 1950

By R. McDonald, B. C. Maine

CERRO'S power system consists of three main hydroelectric plants and one steam-electric plant. These are located on the eastern slope of the main Andes range at altitudes between 12,000 and 13,000 ft. The hydroelectric group utilizes the watershed which extends for about 150 kilometers south from Cerro de Pasco, divided into two catchment areas which lie adjacent. The northern area serve,, the medium-head plant of Mal¬paso. while the southern area provides for the high-head plants at Pachachaca and Oroya. The rainfall, as registered at the rain gauging stations of the three hydroelectric plants, averages 33 in. per year; all these stations are subject to the same hydrological conditions, and the bulk of precipitation occurs between October and March. Rainfall during the dry season, April till September, has no appreciable effect on the total annual runoff. Table 1 shows the storage capacities of the main reservoirs of the hydroelectric group. The four plants are tied together electrically by a 50-kv. transmission network, the system being 3-phase 3-wire at 60 cycles with a solidly grounded

Jan 1, 1945

By F. L. Bosqui

The history of the development of gold metallurgy in South Africa is divisible into two periods: That preceding the introduction of the cyanide process on a commercial scale in 1890; and the 24 years intervening between that important event and the present time. The period between the discovery of the banket reefs of the Witwaters-rand and the year 1890 was one of phenomenally rapid progress in mining development, but of no special interest to the metallurgist, There were no strikingly original or distinctly local advances in the treatment of gold ores. Californian practice in stamp milling and amalgamation prevailed, in mills ranging in equipment between 5 and 50 stamps, operating within a wide range of efficiency, and under the trying conditions usually encountered in a new and isolated field. The somewhat incomplete records of gold output of that period show that in December, 1887, there were 20 producing companies in the Witwatersrand district, of which 10 had mills, the largest being of 20 stamps. The only official record of returns for May, 1887, was that of the Wemmer mine, whose modest achievement with a 5-stamp mill was 100 tons crushed, producing 887 oz. 3 dwt. of gold. This is interesting in view of the present-day output of this field, which in 1913 reached 8,430,998 oz. The rapid increase of production during this first period (May, 1887, to January 1891) is shown by the following summary for the year 1890:1 Total tone crushed.................................. 702,828 Number of stamps in operation....................... 1,046 Average number of tons crushed per stamp per day..... 2.36 Yield of gold, ounces................................ 494,817 Yield per ton, pennyweights.......................... 13.36 The stamp duty ranged between 11/2 and 4 tons per day, according to weight of stamp, fineness of screen, and operating skill available. The prevailing weight of stamp was about 900 lb., and the screen, 900 meshes to the square inch.

Jan 1, 1916

Correspondent. Must be able, by virtue of, connection, to forward; daily ands weekly, reports, outlining, conditions affecting economics of. industries, particularly the metallurgical, and chemical, industries. New. York, City. R-1982. Sales Lubrication. Engineer. Preferably with experience, as chief marines engineer familiar with internal-combustion; engines and, with, automobile trade. Must be good salesman, of goody appearance and be able to meet people. Middle West. Salary $27.00: to start. R71984.

Jan 12, 1919

By Franklin H. Wilson

COPPER base alloys containing various amounts of aluminum and silicon are of considerable commercial interest. In particular the alloy containing 7 pct aluminum and 2 pct silicon shows an attractive combination of high strength with excellent machinability. Since there is no evidence in the literature that this ternary system has been studied since the recognition by Smith1 and Anderson2 in 1939 of the kappa phase of the binary copper-silicon system, the determination of the equilibrium phase relationships in the copper corner of the ternary diagram has been undertaken to provide a fundamental basis for understanding the behavior of these alloys. Phase boundaries have been determined by the micrographic examination of quenched samples while points on the liquidus surface have been obtained by thermal analysis. The results, of sufficient precision to serve their major purpose of providing a guide for the control of the manufacture of alloys in the region covered, are presented in a series of both isothermal and vertical sections of the diagram. ALLOYS STUDIED For a rough survey of the copper corner of this ternary system the compositions listed in Table I (Series A) were employed. The results of examinations of these alloys indicated that the compositions listed in Table 2 (Series B) would further clarify regions of greater complexity. All alloys were prepared as 1200 g heats, the copper being melted under charcoal, and then first silicon and then aluminum were added. An [ ] excess of 0.05 pct silicon was allowed to cover deoxidation losses. Two small bars 5/8 by 1 ¼ by 1 ¼ in. were cast using Durville type casting equipment. Analyses were made on all alloys using material which was hot-rolled at 800 or 850°C and annealed 17 hr at 650°C. All analyses were found to be within 0.05 pct of the nominal composition. Since the precision of analysis is not much better than this, the points of both isothermal sections and vertical sections at fixed aluminum and silicon contents have been plotted at the compositions indicated by the nominal composition.

Jan 1, 1948

By W. F. Vietti

Operations in the Texas Gulf Coast during the past year have resulted in the discovery of a number of new fields arid the extension of a few of the older fields, both by deeper drilling and by lateral development. Twenty-six new oil and gas-producing areas and discovery of oil or gas in three subsequently abandoned wells have been recorded for a total of 29 new fields of varying importance. Producing areas at Conroe, Manvel, Greta, Pierce Junction, Hastings, Dirks, Thompson, Batson, Mykawa, Tomball, Raccoon Bend, Samfordyce arid Saxct have been increased. Primary geologic studies followed by geophysical mapping or straight geophysical prospects have been responsible for the more important discoveries. These methods have further proved themselves an aid to limiting wildcatting to the more prospective areas and as the chief means to discover reserves in the future. Production for 1935 totaled 69,560,720 bbl. as compared with 60,850,-360 bbl. for 1934. Strict proration has been in force throughout the discovery of more oil than has been produced. The number of producing oil wells has steadily increased to 4851 at the end of 1935, as compared with 3858 at the end of 1934, owing in part to the prolonged life, because of proration. There were 1272 oil and gas wells completed during 1935, as against 835 in 1934. New Fields Pertinent data on the new fields discovered during 1935 are includrd in Tables 1 and 2. Hence, the discussion of these discoveries will be limited. Bee County.—Three new gas fields were added during the year: Beeville, discovered by G. M. Church in No. 1 Wood on Sept. 21; Foley, discovered by Dirks Bros. in No. 1 Foley on Feb. 8; and Plummer, discovered by Hynes and Fisher in No. 1 T. M. Plummer on Sept. 2. The three new oil fields added are: Hartzendorf, drilled on fee land by Hartzendorf and completed on May 17 as a small pumper; Ray, southwest of Tuleta, and discovered by Jameson and Burns in No. 1 Campbell

Jan 1, 1936

By W. F. Vietti

Operations in the Texas Gulf Coast during the past year have resulted in the discovery of a number of new fields arid the extension of a few of the older fields, both by deeper drilling and by lateral development. Twenty-six new oil and gas-producing areas and discovery of oil or gas in three subsequently abandoned wells have been recorded for a total of 29 new fields of varying importance. Producing areas at Conroe, Manvel, Greta, Pierce Junction, Hastings, Dirks, Thompson, Batson, Mykawa, Tomball, Raccoon Bend, Samfordyce arid Saxct have been increased. Primary geologic studies followed by geophysical mapping or straight geophysical prospects have been responsible for the more important discoveries. These methods have further proved themselves an aid to limiting wildcatting to the more prospective areas and as the chief means to discover reserves in the future. Production for 1935 totaled 69,560,720 bbl. as compared with 60,850,-360 bbl. for 1934. Strict proration has been in force throughout the discovery of more oil than has been produced. The number of producing oil wells has steadily increased to 4851 at the end of 1935, as compared with 3858 at the end of 1934, owing in part to the prolonged life, because of proration. There were 1272 oil and gas wells completed during 1935, as against 835 in 1934. New Fields Pertinent data on the new fields discovered during 1935 are includrd in Tables 1 and 2. Hence, the discussion of these discoveries will be limited. Bee County.—Three new gas fields were added during the year: Beeville, discovered by G. M. Church in No. 1 Wood on Sept. 21; Foley, discovered by Dirks Bros. in No. 1 Foley on Feb. 8; and Plummer, discovered by Hynes and Fisher in No. 1 T. M. Plummer on Sept. 2. The three new oil fields added are: Hartzendorf, drilled on fee land by Hartzendorf and completed on May 17 as a small pumper; Ray, southwest of Tuleta, and discovered by Jameson and Burns in No. 1 Campbell

Jan 1, 1936

By A. R. Denison

To those who are unfamiliar with the nomenclature in common use among operators in the Mid-Continent, the Permian Basin is a term to cover several widely separated productive areas. It includes what is designated in trade journals as the Panhandle fields, the West Texas fields and Southeastern New Mexico fields. It is bounded roughly on the east by the 100th parallel, on the west by the 104th parallel, 011 thr south by the Rio Grande River and various mountain ranges and on the north by the north line of the Texas Panhandle. The name is derived from the location of these three widely scattered areas which are all in a broad geosyncline which is chiefly filled with sediments of Perrliian age. The oil fields as now known and the approximate limits of the Permian Basin are shown in Fig. 1. The major part of the production in these areas comes from liniestonc or dolomite of Permian age. Some granite wash and sandstone production has been developed but ordinarily does not yield as large initial production as the limestone or dolorliite. Introduction During 1926, Texas, exclusive of the Gulf Coast, produced 12G1SS-1,-952 bbl. of oil,' of which 40,727,877 bbl., or 32 per cent., was produccd in the West Texas Permian Basin. During 1927 the same area in Texas produced 167,176,000 bbl. %f oil, of which 89,779,000 bbL13 or 52 per cent., was produced in the West Texas Permian Basin. The increase in the total production of Texas is about equivalent to the increase in production in the Permian Basin. Since production in the Panliandle District declined during the entire year, the increase of production in the West Texas Permian Basin, and also the increase in Texas, is due to the enormous increase from the West Texas fields. Daily average pl-oduction in the West Texas fields increased from 68,211 bbl. on ,Jan. 8 to

Jan 1, 1928

By A. M. Gaudin, D. W. Fuerstenau

By streaming potential techniques, the zeta potential of quartz has been measured as a function of the concentration of dodecylammonium salts at different pH values. These experiments indicate that adsorption of dodecylammonium ions takes place as individual aminium ions until a certain critical concentration is reached in the double layer, at which point the adsorbed collector ions begin to associate into patches of ions, called hemimicelles. Because of its effect on the surface potential of quartz, pH controls the adsorption of dodecylummonium ions. The value of cooperating agents and nonionized constituents of flotation systems is discussed. Previous investigation of the role of pH in the flotation of quartz with anionic collectors is here extended to an inquiry of how the cationic collector, dodecylammonium acetate, is adsorbed at the quartz-solution interface. The equilibrium amount of this collector adsorbed by quartz has already been measured as a function of concentration and pH. The weight of prior evidence has suggested that two mechanisms are involved in the adsorption of this collector by quartz, but with existing adsorption measurements alone it is not

Oct 1, 1955

By Paul L. Cloke, S. R. Titley

Paul L. Cloke (Assistant Professor, Dept. of Geology and Mineralogy, The University of Michigan, Ann Arbor, Mich.) - On examination of the paper by S. R. Titley it appears that several errors have been made, many of them of a fundamental nature. It is the intent of this discussion to point out these mistakes, and, insofar as possible with the data at hand, to correct them. EXPERIMENTAL WORK Ferrimolybdite-Limonite-Water: The results of the experiments described are very difficult to interpret; nevertheless, a correct interpretation is essential as is pointed out subsequently. The difficulties become apparent when reasons are sought for the continuing change of pH after conversion of ferrimolybdite to ferric hydroxide and for the change of pH when -stirring is stopped after a steady value has been reached. In seeking an explanation of this experiment it seemed to this writer that the (temporarily) steady pH values might be due to kinetic rather than equilibrium effects. In this view the steady pH on stirring might be explained as a balance of several reactions proceeding simultaneously. The first of these would be the slow solution of ferrimolybdite,

Jan 1, 1964

Edward K. Landis, Philadelphia, Pa.: In studying Mr. Thackray's paper it seemed that a critical comparison of the results from different methods therein reported would he of interest. For this purpose Table I. was constructed by grouping all determinations made by the same method, and the probable errors were calculated. In this table all chemists titrating with permanganate (A1, E, F, G and 0 of Mr. Thackray's paper) are classed under "Emmerton;" those titrating with nitric acid (C, J, S and Y) under " Handy; " those re Probable error of result: a, ±0.0008996; b, ±0.000694; c, ±0.00107; d, ±0.00115; e, ±0.000674; f, ±0.000473; g, ±0.003675; h, ±20.0007959; i,0.000.5098; k, ±0.0008303. Probable error of single determination: a, ±0.001989; b, ±0.001547; c, ±0.00214; d, ±0.0023;e, ±0.00116;f, ±0.000818;g, ±0.0019;h, ±0.00224; i, 40.00143; k, ±0.00234. The formula for calculating the probable error is 0.6745, where N (N— 1)' S is the sum of the squares of the differences from the mean and N is the number of determinations.

Jan 1, 1897

By C. T. Eddy, R. W. Drier

The conductivity of copper wire is of prime importance to the electrical industry and consequently to the copper refiner and wire manufacturer. Annealed copper wire has a higher conductivity than hard-drawn wire, but, on the other hand, its tensile strength is lower. In an attempt to explain the reasons for these changes in properry. or at least to study the conditions existing in these two types of copper wire, the experiments described here were carried on in the Department of Metallurgy of the Michigan College of Mining and Technology This is only a partial and preliminary report on some of the work done to date on a research program of which the object is an intensive study of copper. It is offered with the hope that, though it is not an exposition of the most rigorous of conclusions, the general trend of inferences which may be drawn will be of value and interest. Effect of Direction of Drawing In 1928 some work was done on hard-drawn copper wires to determine whether wire drawn in a continuous machine (that is, all in one direction) would have a different structure from wire that had a reversal in direction at some stage of the drawing process. The diffraction spectra obtained indicated that as the position of the wire rayed approached the core, a preferential crystal orientation existed, and that the closer to the core the more preferred was the orientation. The data obtained partly corroborated the work done by Schmidt and Wassermanl and by Clark.2 Samples Used in Following Investigation The samples selected for use in both the diffraction work and the resistivity measurements were taken from Nos. 6, 8, 10 and 12 gage

Jan 1, 1930

By J. B. Mackintosh

In a recent paper read before the Institute on this subject, Mr. G. C. Stone advances the theory that the precipitate obtained in Williams's volumetric process,* by treating the boiling nitric acid solution of a manganese salt with potassic chlorate., is not pure MnO2, but approximates more closely to the composition 10Mn02+Mn0. As the value of the process and the accuracy of our results mould depend on whether this was the fact or not, it has seemed to me, and to my associates, Messrs. Beebe and Colly, that it would be of importance to ascertain the composition of this precipitate, and to that end the following series of experiments has been undertaken. These experiments have all been based on the considerations: 1. That potassic permanganate has an oxidizing power equivalent to five atoms of oxygen for every two of manganese : K2Mn2O8 = K2O.2MnO.O5. 2. That the same amount of manganese in the state of binoxide has an oxidizing power equivalent to two atoms of oxygen : 2MnO2 = 2MnO.O2. If, then, we take a solution of permanganate of unknown strength, and reduce it to binoxide, the oxidizing power of the binoxide formed will be equal to that of two-fifths the quantity of the permanganate solution originally taken. While, if the precipitate obtained should not be binoxide, but an indefinite or definite mixture of bin- and mon- oxides, then, from its equivalent oxidizing power, we can calculate its composition. The analyses were conducted in the following manner. The amount of permanganate taken was decomposed with hydrochloric acid, and concentrated to expel the excess of water. Sufficient excess of nitric acid was now added, and the solution was boiled until all the hydrochloric acid was destroyed. The manganese was then precipitated by potassic chlorate; and, after standing some time to cool—a pre-

Jan 1, 1884

By H. M. Chance

Jan 1, 1887

By J. W. Fuller

During the last 20 years, experimenters have sought to utilize pulverized coal in boiler plants, but refractory and slag troubles have usually overbalanced any gains in efficiency that were obtained. These troubles were caused largely by lack of knowledge of the proper methods of coal preparation and lack of a furnace design that would be adaptable to pulverized-coal firing. When this information was obtained the results were so much more satisfactory that within the last 5 years approximately 80,000 boiler hp. have been installed in plants ranging from 1000 to 20,000 boiler hp. rated capacity, which are operating at an average of 174 per cent. of rating. The first commercial installation for burning pulverized coal under boilers was made at the Parsons shops of the M. K. & T. railroad in 1916. Of eight 250-hp. water-tube boilers thus fitted, six were operated continuously for 3 yr., when oil was obtained at such a low price as to warrant its use. Since then, this plant has been operating alternately on pulverized coal and oil, depending on prevailing prices. Recently gas has been discovered within 20 mi. of this plant, so that this fuel may also be used. No change in furnace design was made when changing from pulverized coal to oil or gas firing. One of the principal advantages of this method of firing is its ability to utilize, with maximum efficiency, all kinds of coal, irrespective of quality or physical characteristics. In fact, quite a number of the present pulverized-coal installations were made for the purpose of utilizing coal that could not be marketed or used in any other way. The Pacific Coast Coal Co., at Black Diamond, Wash., formerly hand fired the boilers, using approximately 125 tons of prepared steam-size coal daily. The Black Diamond steam coal is the best quality of coal for steaming purposes mined in that section of the country, and there is a market for all that can be produced. By installing a pulverizing plant, it was possible to substitute coal of a size and quality that had no commercial value, so that 125 tons daily of the best steam coal formerly used at the mines is now marketed. This company made a similar installation, at its New Castle, Wash., mine with approximately similar results. Each of these two plants has a

Jan 1, 1922

There are no data available of shipments until 1858, and then estimated, when railroad service became available. By reason of the iron made in the county, and the large amounts of salt, the tonnage used locally was quite large, and until 1860 these are estimated from the probable salt and iron production. After 1859 the tonnages shipped by rail are partly estimated from the total shipments of the Allegheny Valley Railroad from four counties. It is believed that the totals are below, rather than above, what the actual production was. While considerably larger for the same years than figures given by the Secretary of Internal Affairs and census reports, these data were always too low in those days, partly because of lack of facilities for gathering them, but more largely because of the strong prejudice existing in those days against furnishing such data to any governmental agencies.

Jan 1, 1942

DAVID COLE, El Paso, Tex. (communication to the Secretary*).-I notice that Mr. Callow takes issue with me on certain points I have previously contributed to the Transactions, to which I would reply," follows: Regarding Inspiration: Mr. Callow's refinements of gravity methods made just prior to the advent of flotation in the Southwest were intended to be "down to the minute," and the gravity flow sheet for Inspiration ores which he brought out at that time as the result of his long campaign of experimentation was regarded by many as the last word in wet gravity work. It was not my intention to discredit this good work when I said "he threw small light on slime treatment." "Minus 200 mesh," however, is not-synonymous with "real slime." Repeated treatments by gravity methods will result in the separation of a very large percentage of the fine granular sulphides in sizes much finer than 200 mesh, and I gave him due credit for that. The unavoidable losses were made in the "real" slime which I mentioned in the paragraph which Mr. Callow refers to as my "statements on page 1871," and his refinements were not a solution to the real slime problem. Mr. Callow has called attention to the fact that his best results at. Inspiration in gravity work were obtained on Joe Bush ore and he presents a table, viz., Tests Nos. 68, 71, and 72; and asks us to consider and compare these results, made by his refined gravity methods, with the results obtained by 'flotation methods in the new mill "during the second half of 1915 and in August, 1916." Of course, the ore treated in the mill was what happened to come. It was not Joe Bush ore, and therefore comparisons of detail figures are of no value. In regard to Mr. Callow's statements under the caption "Callow Cells Adopted by Arizona Copper Company," I would say that those who have been sufficiently interested to read my paper and' his remarks will notice what amounts to contradictory statements on several points which I cannot allow to go unchallenged. I was on the ground constantly and got my information at first hand. The tables published were the official ones. Mr. Callow was not present during the period and has been laboring under that disadvantage.

Jan 5, 1917

By T. W. Turk, S. A. Sass

The engineering design features of the Mercur Gold Project crushing, grinding, carbon-in-leach, bullion, reagent, and tailings disposal areas will be discussed in this paper. It presents a project overview and flowsheet along with selected design alternatives. During the engineering phase, major equipment and other features were selected to optimize design within site location and environmental constraints. The project team organization and procedures were adopted to accommodate multiple optimization studies while maintaining a tight engineering schedule.

Jan 1, 1985

By Jose J. Alvarez

INTRODUCTION Three major plant processes control the operation of an uranium in-situ mining plant. The three pro- cesses can be categorized under the headings termed - loading, elution, and precipitation. All other plant processes can be subcategorized under the three. In summary, the major plant processes in an in- situ mining plant are: 1. Loading - The pregnant lixiviant from the production wells flows through anion exchange resin. The loading of the uranium on the resin is controlled by various factors, such as ionic strength, valence, functional groups, etc. 2. Elution - The uranium-loaded resin is eluted or regenerated by a high ionic strength solution. The effluent from this process is de- fined as the pregnant eluate. 3 . Precipitation - The pregnant eluate is chemically treated to attain the ideal conditions for the selective precipitation of uranium. Hydrogen peroxide is an ideal precipitating agent under URI conditions. This paper will focus on two of the major processes: Elution and Precipitation. For the most part this relatively young industry has grasped the theory of liquids in continuous flow for the operation of the plant processes. Likewise, URI initially operated the two processes of interest with the continuous flow concept. Through laboratory experiments, URI has abandoned the flow concept and operates the eluting and precipitating stages with the concept of "batch" processes, whereby only one container, filled to a specified liquid level, is chemically processed. In the proceeding sections, each of the two pro- cesses will be discussed in detail, from their original concept to the present operation. Only minor modifications were made to the existing circuits for the conversion with results revealing a greater level of efficiency. ELUTION The elution of the loaded anion exchange resin can be achieved by different high ionic strength solutions. The selection of constituents is dependent on both chemical and economic considerations (i .e., efficiency and operating/restoration factors). The regenerant presently used is a pH 9.5 solution containing 7.5% NaCl and a mixture of Na2C03/NaHC03. expressed as 2.5% NaHC03.

Jan 1, 1980

By A. G. McGregor

In the Southwest a number of large steel chimneys discharge the gases from the copper smelting furnaces. Some of these chimneys show no deterioration after twenty years, others show serious deterioration after tour years service. A steel stack 20 ft. 7 1/2 in. (6.3 m.) in diameter by 279 ft. (85 m.) high, Fig. 1, used for roaster gases only at the Calumet & Arizona Mining CO.'s plant at Douglas, Ariz., was lined to the top of the bell, 40 ft. (12 m.) from the base, with common brick placed on end, lime and cement mortar being used. Above the bell, the stack was lined with 4 by 4 by 8 in. (10 by 10 by 20 cm.) building tile, placed on end, making a lining 4 in. thick. The tiles were supported on angle-iron rings, riveted to the inside of the chimney shell about 15 ft. apart. The chimney was put into use in the latter part of 1913. Four years later, considerable of the tile lining had fallen out of place and mixed with the flue dust at the bottom of the chimney. About a year later, several flattened and distorted sheets were noticed about 70 ft. below the top. An investigation showed that the thickness of these distorted sheets, which was originally 1/4in., had been reduced one half. One year later, or six years after the chimney went into use, several holes, a square foot and smaller in area, appeared near the flat spots mentioned. Soon afterwards, when the chimney was shut down for repairs, it was found that most of the tile lining for 125 ft, from the top had fallen out and that the steel shell was badly corroded; but where the lining remained in place the shell was undamaged. Many of the tile which had fallen out of place had become soft or flaky and crumbling, others were as hard as when placed. An analysis of the hard and soft tile is shown on p. 2. These tile were of a fireclay mixture, and were used instead of common brick on account of lower cost. The steel forming the upper 125 ft. (38 m.) was replaced, and a 4 1/4-in. (11-cm.) lining of firebrick was laid from the bell to the top of the chimney. The brick were placed on end, wedges and keys being used. After each 8-in. course was set, keyed, and wedged tight, thin mortar —water, sodium silicate, and finely ground silica—was worked into the joints. The brick are supported on angle-iron rings, every 15 ft., in the same way as the tile.

Jan 1, 1921