Search Documents

By E. S. Bardwell

TWENTY years ago, W. T. Burns, in his paper presented at the Butte Meeting of the Institute, discussed the general scheme then in use in the electrolytic copper refinery of the Anaconda Copper Mining. Co. at Great Falls, Mont., for removing impurities from tank-room electrolyte and controlling its acid content. Since then marked changes have taken place in copper-refinery practice at Great Falls. The electrolytic copper refinery in use in 1913 has been replaced by a new and entirely modern plant capable of producing up to 27,000,000 lb. of cathode copper per month. With the building of the new plant in 1916, the treatment of converter anodes was discontinued. At first the substitution of refined anodes for converter anodes resulted in a decrease in the impurity content of the anodes, but as time went on, changes in metallurgical practices at the smelter resulted in substantial increases in the impurity content of the anodes. The leaching of the silver slime to remove copper before shipment to Raritan for treatment further added to the purification problem. The fact that the scheme of purifying electrolyte which Mr. Burns described in his paper' has proved adequate for our needs under a wide variation of conditions speaks well for its essential soundness.

Jan 1, 1933

By J. E. Johnson

Discussion of the paper of Prof. Henry M. Howe, presented in abstract by Bradley Stoughton at the Cleveland meeting, October, 1912, and printed in Bulletin No. 71, November, 1912, pp. 1181 to 1227. J. E. JOHNSON, JR., Ashland, Wis. :-This material is largely beyond the understanding of most of us practical men, but there are certain portions of that diagram that are of enormous importance. We have been conducting a, very extensive investigation of the quality of charcoal-iron and of all iron, and we have found that this eutectic point here (indicating on chart) is a sort of balance point between the good irons and the bad ones, and I have some slides to show what practical results we have found depending on this point here on the d diagram; and it would be a mistake for men who were -concerned with operation and were getting practical results, to think that these things are of no importance. They are exceedingly hard to understand, and I confess that a great deal of them is beyond me, but there is no doubt that the quality of iron depends to an enormous extent on the location of the carbon content of the iron to the right and left of this point, and the position of that point is dependent upon the silicon, and I wish to make a plea to the practical and operating members of the Institute not to disregard this material on account of its being highly scientific, because I hope to show later that there are matters of the highest practical importance that depend directly upon certain portions of this diagram. HENRY D. HIBBARD, Plainfield, N. J. :-T understand that washed metal, which, is practically pure iron and carbon, does not solidify at one instant, is cast-iron does, but passes through a mushy stage. That has been an enigma as to how and why it did so, and possibly some light might be thrown on it in this connection. BRADLEY STOUGHTON, New York, N. Y.:-If Wittorff's propositions are accepted, washed metal containing more than 4.1 per cent. of carbon would begin to solidify as liquid plus Fe4C

Dec 1, 1912

By Henry M. Howe

(Cleveland Meeting, October, 1912.) Manuscript received Aug. 20, 1912. PROFESSOR RUFF'S most illuminating paper' describing his extremely valuable investigation of the carbon-iron equilibrium diagram assigns definite temperatures to certain very important lines in the diagram, viz.: the eutectic temperature EB (Fig. 1) and the " near-eutectic " temperature E'B', but without indicating clearly enough the reasoning which warrants these assignments. These temperatures are of such primary importance that it is well to indicate the apparent gaps in the reasoning, in the hope that they may be filled in. In his diagram, Fig. 1, the point X and the lines B'D'HI represent the experimental results reached by Ruff and Goecke; 2 AB, AEB, and SE represent the liquidus, the solidus, and the solubility of the carbon in solid austenite in the absence of graphite according to Gutowsky,3 AB being also in accordance with Spiecker's data;4 ,S'E' the solubility of carbon in solid austenite in the presence of graphite according to Ruer and Iljin; 5 GS the line A3 after Goerens and Meyer;6 I and EBC the true eutectic line as inferred by Prof. Ruff from Gutowsky's data. B'D' is the carbon content found in cast-iron which had apparently saturated itself with carbon in the presence of the graphite of the enclosing graphite crucibles in which the determinations were made. From its straightness and its sharp inflection near 6.67 per cent. of carbon, the carbon content of * Professor of Metallurgy in Columbia University. 1 Metallurgie, vol. viii., pp. 456 to 464 and 497 to 508 (1911). 2 Idem, vol. viii., pp. 417 to 421 (1911). 3 Idem, vol. vi., pp. 731 to 743 (1909). 4 Idem, vol. vi., p. 524 (Wüst) (1909). 5 Idem, vol. viii., pp. 97 to 101 (1911). 6 Idem, vol. vii., pp. 307 to 312 (1910).

Nov 1, 1912

By Albert Sauveur

IN an important paper on The Critical Ranges of Pure Iron1 presented at the May, 1913, meeting of the Iron and Steel Institute, Professor Carpenter reports and illustrates the results obtained by him in determining the critical points of some very pure electrolytic iron sheet about 0.01 in. thick. The iron strip was rolled into the form of a cylinder 1; in. long and fn in. wide, weighing 42 g., the thermo-couple being inserted in a small hole in the center. The points Ac3 and Ar3 were readily detected, as well as the point Art, but the point Act was absent from every heating curve. The author concluded from this occurrence that the conception of A2 as an independent allotropic change must be abandoned. At the October, 1913, meeting of the American Institute of Mining Engineers, Dr. G. K. Burgess and J. J. Crowe read a paper on The Critical Ranges A2 and A3 of Pure Iron2 in which they described some extremely important experiments conducted by them at the Bureau of Standards, Washington, D. C. They tested Professor Carpenter's electrolytic iron and detected a sharp and well-defined absorption of heat corresponding to the point Act. Professor Carpenter's failure to detect this point is explained by them on the ground that unless the iron be in a single piece entirely surrounding and in contact with the thermo-couple junction, the thermo-couple will integrate the irregular progress of the heat through the sample and the curves will lose their sharpness. In discussing Messrs. Burgess and Crowe's paper, Professor Carpenter accepted this explanation as probably accounting for the absence of Act in his heating curves.

Jan 2, 1914

By A. Sauveur

In an important paper on The Critical Ranges of Pure Iron presented at May, 1913, meeting of the Iron and Steel Institute…

Jan 1, 1915

By H. H. Taft

IT has long been known that the volcanic area south of Belmont, Nye county, Nevada, had mining possibilities. Some of the old-time prospectors knew that gold existed there. Its remoteness from any source of supplies, its long distances from water, the absence of game, and more, perhaps, the lack of grass for animals to subsist upon, has made this an unattractive region in which to search for mines. The decline of the Comstock mines, the exhaustion of sun-dry large and rich ore-bodies, the high cost of mining, marketing, and particularly the ruinously high freight-charges upon refractory ore that had to be shipped to distant smelters, have kept investors out of Nevada for the last few years; and mining people have hardly yet awakened to the importance of Tonopah, Goldfield and perhaps the newer and less developed districts. The discovery of Tonopah by J. L. Butler, who located the Mizpah Claim in May, 1900, and the fortunes soon realized there, attracted many people. As the boom declined, many went away, some scattering out into the surrounding country, and the population is now about 7,000. In the fall of 1902 a discovery of gold was made 23 miles south, in what is now known as the Sandstorm group, 4 miles NW. of Goldfield. In the winter of 1903-4 the Combination, January and Florence mines were discovered, and shipments of high-grade ore soon followed. In January, 1905, there were 10,000 people in Goldfield. In June, 1904, rich gold-ore was found 85 miles SE., at the foot of the south end of the Kawich mountains, but this discovery was kept quiet until a re-location could be made. On August 10, 1904, the Bullfrog claims, and a month later the Ladd mountain and neighboring claims, were located. The Shoshone group was located September 24. This district is

Nov 1, 1905

By E. W. Parker

THE anthracite region, from which there is produced annually about 80,000,000 tons, or approximately 15 per cent. of the total coal supply of the United States, has a combined area of a little less than 500 sq. mi. The total area occupied by the bituminous fields (including the sub-bituminous and lignite fields) is in the neighborhood of 400,000 sq. mi., so that 15 per cent. of the total output comes from this small area of only a little more than 0.1 per cent. of the coal-producing territory. of the United States. The productive anthracite area is divided into a series of basins or troughs, roughly parallel, extending. in a northeast-southwest direction through the. Counties

Jan 1, 1921

By Robert U. Richards

THE spitzlutte, as described by Rittinger, is an instrument by which sand is sorted in a continual upward-flowing stream of water. Its usual form is that of a pointed box, placed with the point downward, the box receiving its feed of sand and water on one side and discharging its tailings on the other. The sorted product or concentration is discharged at the bottom, at the apex of the box, through which it falls against an upward current -of water, which current serves to effect the sorting and separation. To make this exposure to the quicksand condition still more effective, an inverted dam is placed across the spitzlutte, which forces the sand in its passage from the feed side to the discharge to pass down into the more active portion of the upward-flowing current of water. In the form adopted for small tests in the mining laboratory of the Massachusetts Institute of Technology (see accompanying sketch) the feed is through a central funnel instead of being at the side, and the apex of this funnel, extending down nearly to, the apex of the cone, takes the place of the inverted dam in the large spitzlutte. To keep the experiment perfectly under control a bulb or bottle is attached by a rubber connecting-tube to receive the concentrations, and between this bottle and the apex of the spitzlutte is placed -1 glass tube of conical form, with a side tube for the admission of water. By regulating the admission of water through this tube the concentration may be allowed to go down into the bulb as fast or as slowly as the richness of the ore seems to demand. This spitzlutte may be used on products that have been previously sized to take out the heavy mineral. As, an example of this a test was made upon a' sample of ore where the galena, blende, and quartz were much mixed up, but separated fairly when crushed. The sample of ore was from Newburyport, and assayed 16.50 per cent. lead.. After sizing through a series of sieves the ore yielded concentrations and tailings to the spitzlutte as follows

Jan 1, 1881

By Edgar Bain

THE results are presented of an investigation to discover the fundamental atomic conditions existing in Corson's high-copper alloys hardenable by means of silicide solution and reprecipitation. The study attracted the writer because it offered a very convenient example of the general type of hardenable alloys exemplified by duralumin, Sykes' iron-tungsten and iron-molybdenum alloys and the hardenable lead antimony. SUMMARY OF RESULTS AND CONCLUSIONS In brief, the investigation shows that very perfect crystallinity exists in the solid solutions prepared at high temperature to contain as much dissolved silicide as possible. The lattice parameter of copper is altered very little by the amount of cobalt, nickel or chromium, and silicon dissolved. The hardening by drawing at a temperature to initiate precipitation of excess silicon and metal produces a remarkable change in the atomistics of the solid solution. The old perfection of crystallinity in the solid solution is destroyed and the new atomic spacings are of variable magnitude even in one grain, and a condition in a measure resembling that found in martensite is observed. Were no particles whatever precipitated by the drawing action, still ample causes for hardness would be in evidence in the solid solution alone. The lack of plane continuity in the crystal grains would oppose slip in a very effective manner. Complete annealing restores the crystalline perfection in large measure. In these alloys even the most sensitive methods failed to reveal the characteristic pattern of the silicide precipitated by the anneal, but this was largely to be expected as the amount is relatively small.

Jan 1, 1927

By Bain, Edgar C.

THE results are presented of an investigation to discover the fundamental atomic conditions existing in Corson's high-copper alloys hardenable by means of silicide solution and reprecipitation. The study attracted the writer because it offered a very convenient example of the general type of hardenable alloys exemplified by duralumin. Sykes' iron-tungsten and iron-molybdenum alloys and the hardenable lead-antimony. SUMMARY OF RESULTS AND CONCLUSIONS In brief, the investigation shows that very perfect crystallinity exists in the solid solutions prepared at high temperature to contain as much dissolved silicide as possible. The lattice parameter of copper is altered very little by the amount of cobalt, nickel or chromium, and silicon dissolved. The hardening by drawing at a temperature to initiate precipitation of excess silicon and metal produces a remarkable change in the atomistics of the solid solution. The old perfection of crystallinity in the solid solution is destroyed and the new atomic spacings are of variable magnitude even in one grain, and a condition in a measure resembling that found in martensite is observed. Were no particles whatever precipitated by the drawing action, still ample causes for hardness would be in evidence in the solid solution alone. The lack of plane continuity in the crystal grains would oppose slip in a very effective manner. Complete annealing restores the crystalline perfection in large measure. In these alloys even the most sensitive methods failed to reveal the characteristic pattern of the silicide precipitated by the anneal, but this was largely to be expected as the amount is relatively small.

Jan 1, 1927

By Willard P. Ward

I DESIRE to put on record a fact in relation to the effect of manganese on pig iron which I have never seen in print, and which may, perhaps, not have been observed by anyone except myself. In 1875, while experimenting on the production of ferromanganese in a blast furnace, I had burdened the furnace with a charge which I calculated would produce a 15-per cent spiegel. The furnace was very hot, having been running for a couple of days on a very light burden, and making a burnt iron, preparatory to the experiment I proposed to make to produce a high grade of spiegel. I calculated that the charge would begin to work in the furnace in about twelve hours after it was put on. After the lapse of that time I commenced to watch the furnace carefully, but found little or no. change in the appearance of the metal, although there was a very decided change in the cinder, showing that the manganese ore had come in work. The iron was quite gray and very tough. The character of the cinder was good for a spiegel cinder, and I could not imagine where the manganese was going to, as it (lid not appear to be in either metal or cinder in anything like the quantity indicated by the charge. The next morning things were still in, the same condition. The furnace was very hot, the cinder good, and the iron very gray. I was then sorely puzzled. I had never heard of a gray iron with over 4 or 5 per cent, of manganese, and had no notion that an iron with more manganese could be gray and tough. With great difficulty a piece for chemical test was broken from a pig. Not less than a hundred blows from a heavy sledge were required to break it off I took the specimen to the shop and directed that some borings from it should be brought to the laboratory, as soon as possible, whither I myself at once repaired -to put things in readiness for a rapid manganese determination. In about half an hour the machinist came in with the iron specimen in one hand and about, a dozen blunted and broken steel drills in the other, saying that it was impossible to drill the iron. There were several little depressions in the iron, which was malleable enough to be indented by the point of the drill, but not a chip had been cut from it. The metal was never subjected to further physical tests. A chip broken from

Jan 1, 1882

By J. M. Hartman

ONE of the most important subjects to the blast-furnace engineer is a thorough knowledge of the conditions affecting the temperature in the different portions of the furnace. All efforts to decrease the consumption of fuel and improve the working of the furnace must be based upon it, and I may, therefore, be permitted to place before you the results of my observations, and detail the conclusions to which they have led me. Tracing the thermic conditions from below upwards we have at the bottom of a blast furnace making No. 3 iron a temperature of 2900° F., which increases slightly to a point a little below the tuyeres. In the immediate vicinity of the tuyeres the temperature is somewhat lower, owing to the entering blast ; but a short distance above the tuyeres, where all the oxygen of the blast has been converted into carbonic acid, the highest temperature in the furnace must be attained. This carbonic acid is, however, almost as soon as formed, converted by the glowing coal into carbonic oxide, a process which absorbs heat, reduces the temperature, and provides the active agent for the reduction of the iron ores. In its ascent the hot carbonic oxide gradually parts with its heat, first melting the descending iron and earthy materials, which drop down to the hearth of the furnace. The limit of this zone of fusion is rather sharply defined, and the temperature above this zone suddenly decreases. In its further ascent the hot gas drives off the carbonic acid from the limestone, which causes a further absorption of heat, and finally the gas escapes at the top, at a temperature of about 250° F. if the furnace is working well. These changes of temperature 1 have attempted to represent graphically,

Jan 1, 1880

By Carl Swartz

IN the course of a recent investigation1 to develop a more satisfactory white-metal. bearing alloy, a number of alloy systems were studied. The cadmium-nickel system showed characteristics desirable in a bearing alloy, and the system was therefore studied in some detail. EXPERIMENTAL METHOD Alloying cadmium with nickel offered no serious difficulty. After melting the cadmium, clean sheet nickel was added while the temperature was maintained at 500° to 600° C: A few crystals of ammonium chloride2 served to clean the nickel when the latter did not "wet" readily. In preparing the individual alloys for thermal and microscopic investigation, a "hardener" containing about 10 per cent nickel was used.

Jan 1, 1933

By ROBERT R. ABUOTT

Discussion of the paper of Prof. Albert Sauveur and, G. A. Reinhardt, presented at the Cleveland meeting, October, 1912, and printed in Bulletin No. 71, November, 1912, pp. 1335 to 1341. ROBERT R. ABBOTT, Cleveland, Ohio :-The possibilities of the commercial application of the facts brought out in this paper have been the subject of many experiments by me during the past four years. It will be well to call attention to several facts having a commercial bearing upon the problem: (1) The lower critical temperature (Ac) of a 3.5 per cent. nickel-steel of low carbon (say 0.20 per cent.) content is about 710° C. This is not lowered to any extent by increasing its carbon content by carbonizing,. By this I mean that a steel with Act occurring at 710°, must, after carbonizing, be heated above this temperature before quenching in order to make it hard. This is not contrary to Guillet's diagram, although at first thought it may seem so, due to a confusion of the Act and the Ar1, points. His diagram merely gives the physical condition of the steel upon slow cooling. Now a steel of 1.20 per cent. of carbon and 3 per cent. of nickel (point " B " of the diagram) may be martensitic upon slow cooling from three causes: (a) Its Ar1 point is below the normal atmospheric temperatures. (b) The time element is so important that the change from martensite to pearlite and cementite does not have time to occur even with this slower rate of cooling. We of course know that nickel has the tendency of retarding this change. (c) A combination of the above two causes. (2) With increasing nickel content the ability of steel to absorb carbon decreases. (3) Many carbonized nickel-steels show a martensitic. structure upon air-cooling, and yet ire so soft that they are of no commercial value.

Dec 1, 1912

By Truman Aldrich

THE Clinton group of the Silurian holds the red or fossiliferous ore; its outcrops have been mapped by the State or U. S. Geological Survey. This group is from 100 to 500 ft. thick in Alabama. There are places where it is entirely missing (doubtless islands in the Silurian Sea). In Etowah County, in part of Greasy Cove and vicinity, the upper seam is a coral reef impregnated with the red ore; in this same county, step, faults bring up the ore in three or four separate ridges practically parallel. In Blount County and on Raccoon Mountain, the ore runs under the carboniferous rocks and outcrops in the valleys. In examining some bore holes made in drilling for oil, the writer found the Clinton group continuing under the Warrior coal field wherever the drill has gone deep enough. It demonstrates the possibility of a source for red ore not seriously thought of until lately. The following bore holes made principally for oil, given in Bulletin No. 22, Geological Survey of Alabama, show the presence of either ore or red rock, and suggest vast resources yet to be outlined and developed for ore in the future. The locations are given by number on the map. Hole No. 1 was drilled 6 miles southeast of Russellville, Franklin County, by the Woodward Oil & Gas Co. It passed through 160 ft. of Clinton, but only a small show of red ore was found at 1280-1400 feet. A test well, No. 2, drilled near Atwood, Franklin County, in section 26, township 8, range 14, penetrated the Clinton at 1608 ft. and found 38 ft. of limy red rock, also 8 ft. of some blood red ending at 1820 ft., making 46 ft. of red rock in all. It is reported that some good ore was cut through but this work was done by a churn drill. The hole was continued by diamond drill from here. At least 400 ft. of the formation was classified as Clinton.

Jan 10, 1924

By Alden B. Greninger

THE time-honored description of the growth of metal crystals to form polycrystalline aggregates is. one in which two important steps are considered: (1) nucleation, and (2) dendritic growth, each dendrite starting from its individual nucleus and resulting (upon slow cooling, assuming no solid phase changes) in a macrograin. This picture of the process of crystallization is incomplete, and it is with the hope of enlarging upon it that the following summary and correlation of experimental results are given. MOSAIC STRUCTURE For an exact definition of the mosaic structure of metal crystals-if, indeed, one can be found that will hold for all metals-one must await the results of much experimental work in this field. The writer's investigations have been. confined almost entirely to copper and a few of its alloys. The following terms are proposed, and will be used in this report: Macromosaic Structure1.-Large-scale mosaic structure; more specifically, mosaic structure of a scale large enough to produce splitting of Laue spots on back-reflection Lane X-ray patterns2. The deviation from exact parallelism of neighboring macromosaics varies from a few

Jan 1, 1935

By Alden Greninger

THE time-honored description of the growth of metal crystals to form polycrystalline aggregates is one in which two important steps are con-sidered: (1) nucleation, and (2) dendritic growth, each dendrite starting from its individual nucleus and resulting (upon slow cooling, assuming no solid phase changes) in a macrograin. This picture of the process of crystallization is incomplete, and it is with the hope of enlarging upon it that the following summary and correlation of experimental results are given. MOSAIC STRUCTURE For an exact definition of the mosaic structure of metal crystals-if, indeed, one can be found that will hold for all metals-one must await the results of much experimental work in this field. The writer's investiga-tions have been confined almost entirely to copper and a few of its alloys. The following terms are proposed, and will be used in this report: Macromosaic Structure l.-Large-scale mosaic structure; more specif-ically, mosaic structure of a scale large enough to produce splitting of Laue spots on back-reflection Laue X-ray patterns2. The deviation from exact parallelism of neighboring macromosaics varies from a few

Jan 1, 1935

By S. P. Kinney, A. J. Boynton

THIS paper is not the result of recent research with regard to any particular feature of iron metallurgy, blast-furnace practice or mechanical engineering. It is rather a series of notes with regard to contemporaneous progress in blast-furnace construction and operation, together with some suggestions as to changes and improvements that seem to be desirable. On April 1, 1930, a committee of the Chicago District Blast Furnace and Coke Association presented to that association a rather ambitious study of blast-furnace equipment, together with recommendations for the design of construction of a complete blast-furnace unit for the production of 1000 tons of pig iron per day. This report represented the result of months of study by members of the committee, all of whom were actively engaged in operation or design of blast-furnace plants. It is natural, therefore, that comments on the present state of the art should take this 1000-tons furnace, as recommended by the Association, as a basis for comparison with corresponding features of construction or design. See Table 1. In looking through this report at the present time, the significant difference is in the nature of the gas-washing equipment that today would be specified. In other respects, the installation would remain generally very much as it was outlined in 1930. The change in furnace dimensions would be very slight and of no especial significance, except with respect to changes in stock-line diameter, of which more will be said hereafter. The high-efficiency stove as developed by various designers is thoroughly described within this report, and few changes have taken place since. Blowing equipment shows refinement and reflects the improvement in steam engineering. Hoisting and charging equipment and ladles for the conveyance of iron and slag remain, in general, as they were in 1930. In accordance with a general tendency

Jan 1, 1935

By A. J. Boynton

THIS paper is not the result of recent research with regard to any particular feature of iron metallurgy, blast-furnace practice or mechanical engineering. It is rather a series of notes with regard to contemporaneous progress in blast-furnace construction and operation, together with some suggestions as to changes and improvements that seem to be desirable. On April 1, 1930, a committee of the Chicago District Blast Furnace and Coke Association presented to that association a rather ambitious study of blast-furnace equipment, together with recommendations for the design of construction of a complete blast-furnace unit for the production of 1000 tons of pig iron per day. This report represented the result of months of study by members of the committee, all of whom were actively engaged in operation or design of blast-furnace plants. It is natural, therefore, that comments on the present state of the art should take this 1000-tons furnace, as recommended by the Association, as a basis for comparison with corresponding features of construction or design. See Table 1. In looking through this report at the present time, the significant difference is in the nature of the gas-washing equipment that today would be specified. In other respects, the installation would remain generally very much as it was outlined in 1930. The change in furnace dimensions would be very slight and of no especial significance, except with respect to changes in stock-line diameter, of which more will be said hereafter. The high-efficiency stove as developed by various designers is thoroughly described within this report, and few changes have taken place since. Blowing equipment shows refinement and reflects the improvement in steam engineering. Hoisting and charging equipment and ladles for the conveyance of iron and slag remain, in general, as they were in 1930. In accordance with a general tendency

Jan 1, 1935

By Francis Pyne

THE following notes are presented in an endeavour to point out the disadvantages attending the use of chrome brick in reverberatory furnaces used in the treatment of materials that are too valuable to allow any great amount of metal absorption by the brick. Several years ago one of the large Eastern copper refineries decided to utilize basic material in place of siliceous material in the walls of their reverberatory furnaces for the treatment of very foul blister copper, as the latter rapidly corroded the siliceous linings. It was also expected that there would be less slag formation, with a consequent decrease in the cost of treatment and a reduction in the metal losses. Magnesite brick were first used, but while the corrosive action of the foul material was greatly reduced, and the amount of. slag formed was very much less, the magnesite proved to be unsatisfactory in certain parts of the furnace, because of its tendency to crack and span badly when subjected to the alternate heating and cooling that takes place in a reverberatory refining furnace. This made many repairs necessary, and consequent frequent interruption in the operation of the furnace, which, together with the high price of the magnesite brick, ran up the cost of maintenance to an unreasonable amount. It was, therefore, decided to substitute chronic brick for magnesite brick in the parts of the furnace affected. The results, as far as the reverberatory furnace was concerned, were very satisfactory. The corrosion due to the action of the foul blister was small and the amount of slag formed was no greater than when using magnesite, and the tendency to crack and spill shown by the magnesite was eliminated. Gradually the use of chrome brick was extended to all furnaces; those treating blister copper as well as those melting cathodes, and the results were so satin-, factory that the siliceous roofs were replaced by roofs of chrome brick except in certain places where experience showed a more satisfactory performance on the part of the silica brick.

Jan 12, 1917